Final Report Summary - ACQWA (Assessment of Climatic change and impacts on the Quantity and quality of Water)
Executive Summary:
The overarching goal of the ACQWA project has been to quantify and assess the consequences of shifts in temperature and precipitation patterns, as well as the behaviour of snow and ice in many mountain regions in a changing climate. These transformations of mountain climates and environments are expected to change the quantity, seasonality, and possibly also the quality of water originating in mountains and uplands. As a result, shifts in water availability will affect both upland and populated lowland areas. Economic sectors such as agriculture, tourism or hydropower may enter into rivalries if water is no longer available in sufficient quantities or at the right time of the year. The challenge of the ACQWA project has thus been to estimate as accurately as possible future changes in order to prepare the way for appropriate adaptation strategies and improved water governance.
The project identified the vulnerability of water resources in mountain regions such as the European Alps, the Central Chilean Andes, and the mountains of Central Asia (Kyrgyzstan) where declining snow and ice are likely to strongly affect hydrological regimes in a warmer climate. Model results were then used to quantify the environmental, economic and social impacts of changing water resources in order to assess how robust current water governance strategies are and what adaptations may be needed in order to alleviate the most negative impacts of climate change on water resources and water use.
Current generations of state-of-the-art models were applied to various interacting elements of the climate system, that include regional atmospheric processes in complex terrain, snow and ice, vegetation, and hydrology. Results from the suite of models have enabled to project shifts in water regimes in a warmer climate in the case-study regions mentioned above. Observations, targeted models, and methodologies from the social sciences were applied to the impacts analyses on sectors such as tourism, agriculture and hydropower which will certainly be strongly impacted upon by changing water regimes. The results from these different approaches then served to develop a portfolio of recommendations for adaptation and updated water governance strategies.
In summary, the ACQWA project explored specific science and policy themes that included in particular the following elements:
• An integrated modelling approach to enable accurate projections of changes in the seasonality and quantity of runoff in the river basins under scrutiny in the ACQWA project;
• Identification of key economic impacts on a number of important water-relevant sectors, including the possible compounded effect of changing frequencies and intensities of extreme events;
• An assessment of the possible rivalries between economic actors that will be faced with changing water resources;
• A portfolio of possible water governance strategies to alleviate future problems of water allocation and use, of relevance to upcoming revisions of the EU Water Framework Directive.
Project Context and Objectives:
Beniston and Stoffel (2014) have emphasized the fact that mountains are recognized as “sentinels for environmental change”, in the sense that they exhibit dynamics in physical and biological systems that are often more readily identifiable than in other geographical entities of the globe (e.g. Loarie et al., 2009; Engler et al., 2011; Gobiet et al., 2014). The study of cryospheric, hydrologic, geomorphic and socio-economic change in sensitive mountain regions enables to further our understanding of how an important part of the terrestrial environment responds to, is affected by, and may adapt to rapid and sustained changes in temperature and precipitation regimes (Beniston, 2009; IPCC, 2013).
The importance of mountain regions as a provider of numerous ecosystem services was already recognized at the United Nations Conference on Environment and Development (UNCED, Rio de Janeiro, Brazil, 1992); mountain regions were included under Agenda 21 of the UNCED conference, which mentions under its Article 13 (UN, 1992) that “Mountains are important sources of water, energy, minerals, forest and agricultural products and areas of recreation. They are storehouses of biological diversity, home to endangered species and an essential part of the global ecosystem. From the Andes to the Himalayas, and from Southeast Asia to East and Central Africa, there is serious ecological deterioration. Most mountain areas are experiencing environmental degradation.”
Among these visible impacts, changes in glacier length and volume are perhaps the most spectacular manifestations of climate impacts in mountains; currently and with few exceptions, mountain glaciers from the equatorial to the high latitudes are experience glacier shrinkage (Paul, 2011; Bolch et al., 2012), highlighting the fact that this is a global phenomenon. Shifts in mountain snow-pack behavior in the past decades has also been observed, with collateral impacts on the timing of snow-pack melting and thus of surface runoff (e.g. Beniston, 2010), and also an influence on the start of the vegetation period for certain mountain plant species (e.g. Gottfried et al., 2012).
More subtle changes are reported for mountain ecosystems, in part because of the longer timescales involved in biological systems compared to mountain cryospheric systems, for example, and also because certain species are more adaptable than others (Dubuis et al., 2013), thus resulting in greater difficulties in attributing cause-to-effect relationships of climate change. In mountains, the transitions between biological entities and vegetation (ecotones) can occur over short distances, contrarily to what takes place in lowland plains (e.g. Gosz, 1993). Many changes in vegetation patterns related to sustained shifts in environmental conditions can be identified in these ecotone transition zones, as shown for example by Gottfried et al. (2012) at the boundary between high alpine vegetation and the snowline.
Finally, mountains are also the locale for numerous socio-economic activities, in particular tourism, agriculture, industry, mining, and energy (hydropower). These sectors are all sensitive to climate change, since climate exerts the essential controls on the availability of snow for ski tourism (e.g. Uhlmann et al., 2009; Morrison and Pickering, 2013), or of water for mountain agriculture, hydropower, and for mineral exploitation (e.g. Finger et al., 2012; Fuhrer et al., 2014; Gaudard et al., 2014), for example. However, it should be emphasized here that low priorities to sustainable land-use and natural resource management in many mountain regions in the world implies that changes in forest resources, mountain agriculture and water resources are driven not only environmental change but also by economic and demographic factors (Beniston, 2003).
It is in this context of rapid dynamics of environmental and social change in mountain regions that the ACQWA project (Assessing Climate impacts on the Quantity and quality of Water) was developed, in response to the first call for climate-relevant projects under the 7th European Union Framework R&D Programme (EU-FP7). The large integrating project, coordinated by the University of Geneva from 01.10.2008 – 31.03.2014 under EC contract nr 212250, brought together a consortium of 30 partners in 10 countries and on three continents for a total of 37 different research, public, or private research entities representing over 100 scientists.
The overarching goal of the project was to assess the vulnerability of water resources in mountain regions where snow and ice represent a major input of water for rivers originating in mountains, and where declining snow amounts and receding glaciers in a warmer climate are likely to have profound impacts on hydrological regimes. Future shifts in temperature and precipitation patterns, and changes in the behavior of snow and ice in many mountain regions will change the quantity, seasonality, and possibly also the quality of water originating in mountains and uplands (Sorg et al., 2012; Immerzeel et al., 2013). As a result, changing water availability will affect both upland and populated lowland areas (Hill et al., in press; Sorg et al., in press). The challenge of the ACQWA project has thus been to estimate as accurately as possible future changes in water availability, and the impacts these changes may impose on a range of water-dependent economic systems (Fig. 1).
Figure 1: Structure of the ACQWA project and its main components
The flow diagram in Fig. 1 illustrates the broad structure of the ACQWA project. Current generation of state-of-the-art models (e.g. Themessl et al., 2010; Gobiet et al., 2014; Heinrich et al., 2013) were applied to various interacting elements of the climate system, that include regional atmospheric processes in complex terrain, snow and ice, vegetation, and hydrology in order to project shifts in water regimes in a warmer climate in mountain regions as diverse as the European Alps, the Central Andes of Chile and Argentina, and the mountains of Central Asia (Kyrgyzstan). Observations, targeted models, and methodologies from both the social and the natural sciences were then applied to conduct analyses of climate impacts on sectors such as tourism, agriculture and hydropower which could be strongly influenced by changing water regimes. Because these economic sectors and other water-dependent industries may well enter experience conflicts of interests and rivalries if water is no longer available in sufficient quantities or at the right time of the year, a further goal of the ACQWA project was thus to define a portfolio of proposals to pave the way for appropriate adaptation strategies and improved water governance (Hill and Allan, in press; Hill and Engle, 2013). These are designed to help alleviate the more negative impacts of climatic change on water resources and to reduce the risks of conflict between the economic actors most affected by these changes.
The focus on water has been the key element of the ACQWA proposal, because it is an essential resource for human populations, animal and plant communities. Water is relevant in every aspect of mountain systems, in the physical, biological and socio-economic systems. It directly influences the energy supply (hydropower), tourism (snow, water usage, glaciers), forestry and agriculture (productivity changes with changes in water supply, need for irrigation) and services from natural and semi-natural ecosystems. Changes in any of these compartments resulting from shifts in temperature and precipitation regimes are expected to result in feedbacks on water availability.
Achieving sustainable water use poses particular challenges for policy making because of its nature as a public good and because it often has both upstream/downstream and trans-boundary/transnational characteristics. Any changes in climate affecting precipitation and the behavior of snow and, where relevant, glaciers, will have a major influence on the seasonality, amount, and quality of surface runoff. The main changes are thus expected in the surface water systems, which have been comprehensively analyzed in the ACQWA project, in order to quantify the changes affecting the streamflow regimes, which may lead to uneven temporal distribution of the resource throughout the year. In this respect, the project has mainly concentrated on surface waters, both because mountain regions are in general depending on them rather than on groundwater, and given that the configuration of topography and landscape that do not allow the presence of sizeable alluvial deposits necessary for groundwater systems to develop.
A changing climatic regime could alter the frequency and the magnitude of a wide range of geomorphic processes related to extreme precipitation events that could in particular increase the severity of floods and debris flows (e.g. Stoffel and Beniston, 2006). Extreme precipitation events would in addition contribute to larger rates of erosion, discharge and sedimentation. A further factor responsible for decreased slope stability in a warmer climate is the reduced cohesion of the soil through permafrost degradation, particularly in the higher elevations in the Alps. Deglaciation can in some instances lead to problems of water accumulation behind unstable moraines that, if they fail, result in intense flooding and debris flows referred to as glacier-lake outburst floods (GLOFs) that put communities and infrastructure at risk. Similarly, hanging glaciers also pose a threat that is taken seriously in zones where glaciers are located above villages and major communication routes (Funk, 2006). In their retreat, these glaciers reveal a large quantity of unstable rubble and, sometimes ice that could result in severe down-slope flow of material. Given changes in the distribution of population with more extensive and intensive land use, particularly from tourism, any increase in the number and intensity of natural disasters could have proportionally higher human costs. By taking into account the impacts of extreme events within the hydrological cycle, the ACQWA project has acknowledged the fact that resulting natural hazards could contribute to the disruption of access to, and use of, water for many economic purposes.
Shifting precipitation patterns by season and sharply curtailed glacier mass in the mountains will lead to modifications in hydrological regimes and will also mean glaciers will no longer feed water into river catchments at a time of the year when precipitation amounts are low and the snow-pack has completely melted. These changes will have significant impacts on several key socio-economic sectors in mountain regions, particularly since these are also subjected to various other forces that influence their viability. There will in addition be cascading effects on downstream areas. Climatic changes will affect overall land use patterns, which in turn feed back into effects on water and carbon fluxes. Mountain agriculture has been under pressure from lower-cost production in lowland areas. Potential increases in drought conditions will only serve to increase its vulnerability. Forests’ crucial role in protecting against erosion and protecting biodiversity and water storage are potentially threatened. As a result of shifting seasonality of precipitation and glacier melt, the reduction in capacity to store water could also diminish the potential for hydroelectric production just as European-wide efforts are being initiated to reduce dependence on fossil-fuel based energy sources. Not only mountain regions, but the entire European electrical grid could thus be affected. Some rivers, such as those that flow from the northern part of the Alps, may dry up partially or completely towards the end of the summer; this is already the case in the Mediterranean mountains, where the drought situation may well worsen. Although the energy potential of the Alps has by now largely been exploited, this is not the case for other regions of the world where this usage has barely been tapped (Romerio, 2002). By investigating the challenges of managing crucial but often limited water resources in many of the mountain regions that have been studied, the ACQWA project has aimed to better understand social and economic issues such as changes in social arrangements among mountain populations and their downstream neighbours, energy production in regions where water is underused as a means of helping abate greenhouse-gas emissions, and environmental issues such as the future evolution of water supply for use in domestic, tourist, or agricultural sectors.
Climatic change leading to shifts in hydrological regimes has the potential to increase competition over water that will be available at different times and in different quantities. Water is difficult to allocate because of its public good features, which are aggravated by upstream potential to capture the resource and by the fact that flowing water may cross internal and international borders. Changing land cover and land use will generate significant shifts in the amount and seasonality of water resources. For example, deforestation causes an increase in the average annual discharge, and an acceleration of runoff during rainstorm events, also enhancing erosion and downstream sediment supply. Changing social patterns and economic incentives have resulted in major land-use changes in many mountains of the world and, in some instances, have exacerbated the risks associated with excessive runoff and erosion potential.
The AQCWA project has sought to address such issues, particularly in regions with significant social change, for example resulting in land abandonment as farming loses attractiveness and spontaneous vegetation colonizes previously-managed terrain. Conflicting water use, for example between agriculture and hydropower, or between hydropower and tourism, as the resource diminishes through reduced precipitation in some areas and glacier retreat in others, has also been investigated within the ACQWA context (Hill et al., 2014). This has been achieved to acknowledge the fact that new water resource management is not just a matter of adjusting to shifts in the physical environment but is also associated with social changes generated by changing types and levels of water use and new market conditions affecting the distribution of the resource. These different issues have been tackled in the ACQWA project through a number of work packages (WPs), namely:
• WP1: Coordination and administration
• WP2: Climatic and socio-economic drivers of change
This WP provided a quantitative description of the primary (or direct) drivers such as climate change, and of the indirect socio-economic drivers, with relevance to mountain water resources.
• WP3: Modelling climate change impacts on water resources, including extremes.
The focus of this WP was to develop the climate scenarios at the regional and local scale and to model the effects of climatic change on water resources at the basin scale, using state-of-the-art cryosphere, biosphere, and hydrological models.
• WP4: Impacts on natural and socio-economic systems, adaptation strategies, and policy issues
The impacts, adaptation and policy WP investigated the manner in which changing water resources and water use may influence a range of sectors. It was also designed to consider a portfolio of adaptation and other response strategies, and to revisit current water governance issues with a view of improving water resource management.
• WP5: Dissemination, outreach and training
This WP was dedicated to project-specific workshop activities, publications, exchange of young scientists within the network, and public and stakeholder outreach activities.
Project Results:
The ACQWA S&T description will focus on the principal results emerging from the project, pertaining to the following elements:
• The ACQWA case-study regions
• Regional climate change in the ACQWA-case study areas
• Climate impacts by the middle of the 21st century on natural environmental systems that determine water availability: snow and ice; hydrology; extreme events
• Impacts on socio-economic systems (hydropower, agriculture, tourism) and semi-managed systems (forests, aquatic ecosystems)
• Lessons to be learned from the non-European case-study areas
A more complete description of the scientific results, including policy-relevance and possible adaptation strategies, is available in a 100-page ACQWA Science and Policy Brief (Beniston et al., 2013), and a shorter focus is provided in the introduction to the Special Issue of Science of the Total Environment dedicated to the project by Beniston and Stoffel (2014).
Case study regions used in the ACQWA project
Fig. 2 shows the main case-study areas investigated in the context of the ACQWA project. The Rhone and Po river basins in the European Alps have been used as a common “test ground” for model investigations, where the different methodological approaches have converged to the basin scale through appropriate up- or down-scaling techniques. Both basins represent ideal case-study areas, as they comprise all the elements of the natural environment that have been modeled (snow, ice, vegetation, hydrology) and have a wealth of data to enable models to be validated. At the same time, these are highly regulated watersheds, where economic activities related to hydropower generation, irrigated agriculture, and tourism take place in the context of a climate that is at the borderline between Mediterranean and Continental, and are therefore particularly vulnerable to climatic change (Beniston, 2003). The boundaries of the Rhone catchment study-area include the alpine segment, running from the Rhone Glacier in Central Switzerland to Lake Geneva. The boundaries of the Po case-study area used in the ACQWA project do not extend as far as the Adriatic Sea, for reasons of data access and hydrological model constraints. The investigations have thus focused more on the flows from the Alps of Piemonte and Val d’Aosta, with the “ACQWA Po” boundary that is limited to Cremona, on the Po River south of Milan and the western segment of the basin (Coppola et al., 2014). Regional climate model results, however, cover the entire basin as illustrated in the map on Fig. 1.
Figure 2: The principal ACQWA case-study regions: the Swiss Rhone and Italian Po basins, the Chilean Aconcagua basin, and the Kyrgyz catchments of Central Asia
Some of the methodologies developed in the intensive investigations of the European alpine catchments have been applied to the Aconcagua Basin in Chile, where receding glaciers already today pose a genuine threat to water availability (Pellicciotti et al., 2014). Investigating the coping strategies of Chilean economic sectors affected by changes in the quantity and seasonality of water resources can help highlight the types of problems that could arise in the Alps in coming decades (Hill et al, 2014). In Central Asia (Kyrgyzstan), on the other hand, the same processes of ice-mass wasting in the headwaters of the Syr Daria or Amu Daria rivers involve much larger glaciers (Sorg et al., 2012). During the 21st century, the meltwaters from the Tien Shan could potentially represent a source of economic opportunity, for example through the development of hydropower as a source of foreign revenue, but also a risk in view of the political instability and rivalries between different independent states of former USSR (Sorg et al., 2014).
Other research-specific case-study areas, not shown in Fig. 2 comprise the Aragón Basin in Spain for interdisciplinary investigations pertaining to agriculture and energy in a context of changing land-use and climate (Lopez-Moreno et al., 2014); and French Pyrenean watersheds for aquatic ecosystem studies in a hydrology, habitat and biota framework (Khamis et al., 2014a). These are located in the Cauterets region in the vicinity of the French Pyrenees National Park. By analogy with the other non-Alpine case study regions, some of the issues addressed in the Pyrenees in today’s world are likely to be those that will arise in the European Alps in tomorrow’s world.
Regional climate change in the ACQWA-case study areas
The high-resolution simulations carried out within the EU-FP6 ENSEMBLES project (www.ensembles-eu.org) formed the basis for the focused modelling work and climate impacts assessments within ACQWA. Two principal simulations were chosen from the ENSEMBLES multi-model dataset (namely ICTP_RegCM and MPI_REMO, both driven by ECHAM5-r3) and used by all partners. In addition, several impacts studies used the entire ENSEMBLES dataset in order to identify more completely climate-induced uncertainties. The IPCC (2001) A1B greenhouse-gas scenario through to the mid-21st century was applied across all the individual case studies to have a common scenario referenced period for all projections and impacts studies. The 2050s was set as the principal time horizon for the project, as this is a period in the future which is not too far removed from the time-scales typical of forward economic planning and decision-making. Using the ENSEMBLES simulations from a set of 22 high resolution regional climate models (RCMs), climate data in the regions of interest was compiled up to the year 2050. The ENSEMBLES RCMs were used with a horizontal grid spacing of 25 km, and the lateral boundary conditions were provided by eight different global climate models (GCMs). The restriction of using only the A1B emission scenario, rather than other scenarios or a range of emission futures, is of minor importance since the uncertainty due to the choice of emission scenario remains fairly small in the first half of the 21st century (Gobiet et al., 2014).
Results by 2050 using the multi-model mean climate change signals exhibit stronger warming along the Alpine ridge, especially in summer. The high sensitivity of the Alps becomes even more evident in the rather small Rhone case-study region located in the Valais region of south-central Switzerland, with projected median warming over 1.5°C in winter (DJF) and close to 2°C in summer (JJA). Warming is projected by all models of the dataset and for all seasons; the uncertainty of the projected changes is larger in summer and autumn than in winter and spring (Gobiet et al., 2014). In-depth analyses of this data suggests that the choice of the GCM that drives the RCM initial and boundary conditions has by far the largest effect on the total uncertainty, contributing more than 75% to the overall variance in most cases (Im et al., 2010).
Figure 3: Alpine-scale precipitation change in winter and summer by 2050.
(Source: A. Gobiet, University of Graz, Austria).
The massive presence of the Alpine ridge as a dividing feature between Mediterranean and Atlantic or Continental climates clearly influences the spatial distribution of precipitation and the projections of change, as seen in Fig. 3. In the Rhone catchment, the summer decrease and winter increase are small through to 2050, although these changes may well amplify into the second half of the 21st century as shown by a number of earlier studies (e.g. Beniston, 2006). In the Po catchment, temperature and precipitation changes are somewhat more marked than in the Rhone catchment to the north. Between the decades 2001-2010 and 2041-2050, increases of temperature according to different RCM simulations range from 2 - 3°C, and the variation of mean annual precipitation ranges from 1-10%, mainly in the winter and early spring period. Accelerated melting periods, earlier in the year, and likely increases in summertime evapotranspiration will inevitably counter the influence of the larger amounts of summer precipitation on river discharge that are projected for the region.
In the Andean zone of Central Chile, the Aconcagua catchment is projected to experience warmer winters and decreasing precipitation which, as in other mountain regions, will affect the behavior of the mountain snowpack and lead to changes in the timing of snow and glacier melt (Pelliciotti et al., 2014).
In the Central Asian republic of Kyrgyzstan, available climate simulations project by 2050 decreases in summer precipitation by around 5% and increases in winter precipitation around 8%. Temperature increases of between 2.5 and 4.5°C are projected for all seasons in the region. Overall, extreme events will tend to increase, in particular at both ends of the moisture spectrum with more summer droughts and winter or spring flood events (Sorg et al., 2012).
Snow and glacier response to climatic change
• Remote-sensing of changes in snow cover
Remote sensing provides a unique opportunity to address the question of snow cover regime changes at regional scales. Since the availability of daily optical satellite data at the end of the 1980s (NOAA-AVHRR), methods have been developed to compute changes in the surface area of snow cover (SCA) and snow cover duration (SCD). The main parameters analyzed are the timing and duration of the melting season under current and future climate conditions. In this context, a remote sensing database of snow cover dynamics over a time period of 10 hydrological years (2000-2010) was compiled (Dedieu et al., 2014). It focused on the four watersheds of the ACQWA project, namely the upper Rhone (5’300 km2) in Switzerland, the upper Po (37’800 km2) in Italy, the Aconcagua in Chile (5’800 km2) and the Syr Darya source region in Kyrgyzstan (110’000 km2). The satellite data were provided by the MODIS Terra MOD-09 images and the MOD-10 snow products (NSIDC).
Figure 4: Annual standard deviation of snow cover in the Po Basin
The results rely on previous studies already conducted in different regions to reconstruct time series of snow cover at the regional scale and to analyze snow regime trends under the current climate change context. The specific added values of work within ACQWA include a novel and original method for climate impact detection, as well as scientific results for regions with crucial lack of information (Kyrgyzstan, Chile). Significant progress has been made in using statistical tools to assess the interannual variability of snow cover. Maps of standard deviation highlight sensitive regions where strong temporal and spatial variability becomes a significant proxy of climate change related to recent changes in temperature, as shown for example in Fig. 4.
• Changes in snow cover in the alpine part of the Rhone catchment
Numerical simulations to assess the future course of seasonal snow cover under the influence of climate change were performed during the ACWQA project using the detailed snowpack model CROCUS coupled to the land surface model ISBA. In contrast to snowpack components of most land surface schemes within GCMs/RCMs and hydrological models, CROCUS explicitly accounts for internal processes occurring within the snowpack such as compaction, phase change/refreezing and snow metamorphism. During the ACQWA project, the model was driven by atmospheric fields from the RCM REMO at a horizontal resolution of 10x10 km² and run for a century, from 1950 to 2050. Regardless of the metric used to characterize the time evolution of snow conditions, the main outcome from the ACQWA-based study is that the seasonal snow cover in Alpine regions is most likely to decline within the next decades (Fig. 5), essentially as a result of increases in temperature (e.g. Beniston, 2012). This leads to a shift in the snow/rain partitioning towards relatively more rain and less snow precipitation. Such changes are seen to occur particularly at mid-altitude locations, between 1000 and 2000 m elevation above sea-level, which are most sensitive to air temperature fluctuations around the freezing point. The results obtained using the detailed snowpack model Crocus are consistent with the conclusions drawn from other impact models that use simpler snow schemes.
Figure 5: Upper left map mean snow water equivalent (SWEmax; mm); right: altitudinal profile of SWEmax for 1950-1969 period. Lower left: map of changes in SWEmax changes (mm) between 1950-1969 and 2030-2049;right: changes in altitudinal profiles. Dots in the right-hand figures refer to individual simulation grid points, whose average value over a given altitude band is provided as bars.
• Glacier changes in the Rhone and Po catchments
Understanding the glacier response to climate change in mountain regions is extremely relevant for water resources management. Future glacier response is also important to identify the boundary conditions for the evolution of the glacier hazard potential. Climate modifications can indeed influence stream flow regimes, increase downstream landslide and flood risk, and have an impact on hydropower production and other water uses, which strongly depend on melt water. The modelling of glacier response to climate change was carried out at the glacier scale by means of state-of-the-art continuous mass balance models for six different glaciers characterized by different morphological and surface characteristics and representative of the greater Alpine region. The models developed and implemented within the ACQWA project accounted for, in a spatially distributed fashion and at sub-daily temporal scales, accumulation and ablation processes as well as glacier evolution (Finger et al., 2012). The results thus allowed to quantify how glaciers tend to modify in response to changes in climate in order to reach the equilibrium with the current climate. Changes of the mass budget of a glacier and of glacier geometry (surface area, length and volume), were modeled at the highest level of detail in relation to knowledge of glacier behaviour and to the need of long-term modelling under stochastic climate scenario forcing.
The predictions of ice volume evolution suggest a progressive glacier retreat for the period 2001-2050 and a related ice volume reduction (see Fig. 6 for the Rhone Glacier, source of the Rhone River, in south-central Switzerland). Regardless of glacier features and future meteorological conditions, at no time in the coming decades are stationary conditions for ice volume attained, for any of the glaciers investigated. Ablation is generally sufficient to melt the entire amount of snow accumulated during the winter season over the remaining glacierized area and does not lead to changes in the mass balance trend.
Figure 6: 3-D representation of the extent of the Rhone Glacier between 2010 and 2050
(Source: P. Burlando, ETH-Zurich, Switzerland).
An interesting result concerns the behaviour of glaciers and/or glacier areas, which are covered by a thick layer of debris. Their response to climate change shows a slower negative mass balance in comparison to debris-free glaciers. The debris cover tends to insulate the ice beneath, thus preserving the ice. The thin ice thickness located at high elevation and directly in contact with the atmosphere tends to disappear, leading to a scenario in which most of the remaining glacier surface is located at low elevations and is not exposed, but covered by a mantle of rock debris. This result points to the importance of including debris-covered areas in realistic simulations of future evolution of glaciers.
No significant variations in the mean spatial snow melt rates over the simulated decades were observed. However, the variability introduced by spatial stochastic climate forcing allows for spatial changes to occur. Some simulations showed that in some decades high snow melt rates at low elevations could compensate for low rates at higher elevations; in addition, higher snow melt rates were obtained at high elevation, due to increased temperatures in the future, whereas no snow melt occurred at low elevations as a result of the earlier disappearance of snow and the higher fraction of liquid precipitation events. Snow melt contribution to the runoff hydrograph, as well as its variability, thus reflect the characteristics of the meteorological forcing, indicating the importance of accounting for climate scenarios that are localized, spatially variable and highly resolved in time (Fatichi et al., 2014).
A progressive shift and change in shape of the ice melt rate frequency distribution was identified across all melting seasons. This is due to new melt events during times of the day that in the current and past climate did not generally lead to melting. Thus, anticipated ice exposure in spring generates low melt rates, as temperature is not high enough to cause higher melt rates. In summer, more frequent melt occurrence at night is related to higher night temperatures or when the solar radiation contribution is still low, such as during early mornings and late afternoons. Low melt increase is also observed in autumn in some decades and can likely be associated to melt at night. Higher frequency of high ice melt rates was also observed in the last two decades investigated (2031 to 2050) both in spring and summer, due to significantly higher temperature and reduced elevation range of the reduced area of glacier cover. In general, the overall ice melt rate is more variable than snow melt due to the transition of most of the investigated glaciers to small scale glaciers. In this respect, the timing of the transition and, more generally the impact of climate forcing, is significantly influenced by the glacier morphology, the initial ice thickness distributed over a large elevation range and by the intensity of the local climate change signal, which can be different from the large scale patterns simulated by GCMs and RCMs. The ice melt contribution to runoff in glacierized catchments tends accordingly to disappear gradually. An exception is represented by the Aletsch Glacier, the largest alpine glacier situated in south-central Switzerland, where no large changes for the simulated decades (2001 to 2050) are observed in the ice melt hydrograph.
The gradual disappearance of ice volume induced by climate change may confine glaciers to higher elevations causing a possible reduction of the speed of glacier retreat. A key role in this respect is played by the winter accumulated precipitation, which can mitigate (high accumulation) or enhance (low accumulation) the glacier retreat and the backward shift of the accumulation area.
• Stability of glaciers
Three different types of instabilities can be identified according to the thermal properties of the interface between the ice and the bedrock (Failletaz et al., 2012). If cold, the maturation of the rupture is associated with changes in surface velocities and the seismic activity generated by the glacier which, if known, can have predictive value. For the other types of instabilities, water plays a key role in the initiation and the development of an instability leading to rupture. If the ice/bed interface is partly temperate, the presence of melt-water at the interface reduces its basal resistance, which enhances the instability but also renders its prediction difficult. The third type of instability concerns steep temperate glacier tongues which experience enhanced basal motion during the summer melt period. Although instabilities of this nature are still difficult to forecast, a novel numerical model that includes water flow in a sub-glacial drainage network has been developed that has predictive capability. In the context of climate change, the stability of some alpine glaciers may be affected in the near future due to changes in the thermal regime at the ice/bedrock interface. Although some presently-hazardous glaciers may present a lower risk in the near future because of their retreat, some others may evolve towards a critical situation and present a genuine hazard for communities and infrastructure lower down in the valleys. A timely identification of such transitions towards potentially critical hazards is today a challenge that the ACQWA project has contributed to improving.
The modelling techniques developed and used in the ACQWA project represent a step forward with respect to existing literature as they allowed to highlight some important aspects of the impact of climate change on future glacier evolution (Failletaz et al., 2012). The key elements of the added value include: (i) the identification of the important role played by distributed in space and highly resolved in time localized scenarios, which explicitly model the internal variability of future climate, in modulating the glacier evolution trajectory jointly with the influence of glacier morphology, altitudinal range and ice thickness; (ii) a quantification of the evolution trajectories of different representative glaciers by means of statistical descriptors of ice- and snow-melt changes following ensemble based simulations by means of advanced mass balance glacier models; (iii) a decomposition of the mechanisms by which glacierized catchments will change their runoff regime, also showing relative changes of the runoff components and how these depend on the local climate and the glacier configuration (size, aspect, shape, etc.); (iv) the quantification of the role of debris cover in modifying the glacier retreat pattern in space and time, thus highlighting the importance of including this component in glacier evolution assessments for more realistic predictions.
Hydrological response to climatic change
• Distributed basin-scale responses to climate scenarios (Rhone Catchment)
The physically based distributed modelling methodology, as one approach to assessing changes in hydrology, allowed to model the response of the (upper) Rhone river to a high number of stochastically downscaled ensemble members at hourly temporal resolution and with a spatial discretization of 250x250 m, also keeping the highest detail of representation of anthropogenic disturbances, such as hydropower, water abstraction and irrigation. This level of detail is unprecedented, particularly for studies in Alpine regions, and represents a new basis, as compared to existing studies, for investigating options for adaptation policies. The results show how climate change effects on stream flow propagate from high elevation headwater catchments to the river in the major valley, highlighting the damping effect of the river network on the mean and on the extremes, even when the latter show large increases in the upper river reaches. The simulations also indicate that changes in the natural hydrological regime imposed by the existing hydraulic infrastructure are likely larger than climate change signals expected by the middle of the 21th century in most of the river network (Fatichi et al., 2014).
Results suggest that internal (stochastic) climate variability is a fundamental source of uncertainty, typically larger than the projected climate change signal. Therefore, climate change effects in stream flow mean, frequency and seasonality are masked by possible natural climatic fluctuations in large part of the analyzed regions (Fatichi et al., 2014). Simulations also identify regions where strong precipitation increase in the February to April period leads to flow larger than natural climate variability during the melting season. Despite the strong uncertainties induced by stochastic climate variability, an elevation dependence of climate change impacts on stream flow could be identified, with a severe reduction due to the missing contribution of water from ice melt at high-elevation and a damped effect downstream. The presence of reservoirs and river diversions tends to decrease the uncertainty in future stream-flow predictions that are conversely very large for highly glacierized catchments. Despite uncertainty, reduced ice cover and ice melt are likely to have significant implication for aquatic biodiversity and hydropower production. A decrease of August-September discharge and an increase of hourly-daily maximum flows appear as the most robust projected changes for the different parts of the Rhone catchment. This will have a significant impact on hydropower production and management. Water abstraction and irrigation needs, as described by the available data, will likely be marginally affected, as they represent a very small fraction of the use of the available water resources. However, while local changes may be of some relevance, it is unlikely that major changes in total runoff for the entire upper Rhone basin will occur in the decades up to 2050.
• Distributed basin-scale responses to climate scenarios (Po Basin)
The most common approach to assess the hydrologic impacts of global climate change involves the use of climate models to simulate climatic effects of increasing atmospheric concentrations of greenhouse gases, and hydrological models to simulate water-related impacts of climate change. River discharges and their temporal distributions are strongly affected by high mountain areas that are particularly sensitive to global warming. Therefore the quality of hydrological impact investigations, even for larger catchments, depends on the capability to model those specific processes in mountains.
Comparison of parsimonious and fully physically based models showed that simpler models, despite their approach for computing evapotranspiration based only on temperature, are sufficiently robust and accurate to perform hydrological impact investigations of climate change for alpine river basins investigated in the ACQWA project (Ravazzani, 2013). The bias resulting from the approximation of the method implemented to compute evapotranspiration is lower than uncertainty associated with different climate models, however.
Impacts of climate change on hydrological processes were assessed by comparing hydrological model results for the upper Po river basin driven by two different regional climate models (REMO and RegCM3) for the decade 2041-2050 with respect to the decade 2001-2010. Increase of temperature ranges between 15.2 and 17.5%, while variation of mean annual precipitation ranges from 1.1 to 9.6%. Precipitation increase is mainly concentrated in the period from January through March, and causes an increase of snow water equivalent during this period, followed by an accelerated melting period in the mountains. The rise in temperature causes an increase of actual evapotranspiration, mainly in the summer period that counteracts the influence of the larger amounts of summer precipitation on river discharge. Impacts of climate change on flow duration curves of mountain tributaries of the Po River, as provided by different regional models, exhibit a general decrease of discharge for high durations (low flows) and an increase in discharge for low durations (high flows). The two climate models yield different results for the impact on flow duration curves in Po the Po river basin. In particular, the REMO model data results in an increase in discharge for both low and high durations, while use of the RegCM model data yields a decrease in discharge for both low and high durations except for the period ranging from 12 to 58 days.
• Modeling changes in the Po, Rhone and Kyrgyz catchments with the CHyM model
In order not to depend on one single model, the effects of predicted climate changes on Hydrological cycle of Po and Rhone catchments have been studied by forcing the CHyM hydrological distributed model with 8 different climate scenarios simulated by RegCM and REMO regional climate models. A further simulation with the RegCM model has also been used to simulate the future hydrological scenario on the Kyrgyzstan catchments. The future discharge trend has been analyzed for the whole drainage network, obtaining a map of such effects for different seasons (Coppola et al., 2014).
Downscaling of climatic scenario at hydrological scale can be considered an important challenge after the downscaling from GCMs to RCMs models. The results show that the simulated effects of climate change on the hydrological cycle appear to be considerably different for the Po and Rhone basins and also for different regions within the same basin. The same conclusions can be reached using the 8 different simulated climatic scenarios. The decrease of water resources is shown to be more critical for the entire flood plain during the fall season, leading to a loss of 200 m3/sec at the outlet of the Po River. The decrease of flow discharge is estimated to be more than 50% of the seasonal average for a large portion of the drainage network (Im et al. 2010). The effects of climate change on the hydrological cycle appear less evident in the high part of the Rhone valley and, more generally, in the higher part of the Alpine region (Fig. 7). The situation is quite different for Kyrgyzstan, where, for a large portion of simulated domain, increases in discharge during winter months and decreases of water availability during summer are observed.
Figure 7. 30-year mean discharge difference (%) in the Po basin (left) and Rhone basin (right).
Extreme events
Extremes of heat and moisture can have a strong bearing on the rate of snow melt, glacier retreat, and thus on water regimes in mountain watersheds (floods and droughts and associated impacts on a number of water-dependent economic sectors). Geomorphic hazards can add an extra burden on local economies by damaging buildings, disrupting communicating routes and potentially rivers themselves (silting, damming, etc.). A focus on extreme events and natural hazards was thus an important focus of the ACQWA project.
• Novel statistical methods to deal with extreme events
Weather and climate extremes affect societies and ecosystems and sometimes induce fatalities and large financial losses. To reduce the impact of such events and to improve risk assessment studies, it is essential to obtain accurate statistical features of extremes. In the framework of the Extreme Value Theory, a new method has been established, based on Generalised Probability Weighted Moments and Kernel regression (Naveau et al., 2011).
Novel statistical methods were proposed to analyze extreme events, to develop new algorithms to implement such methods and to apply those approaches to hydrological and climate data. Among many byproducts of this project, three specific topics were highlighted: spatial clustering of weekly maxima of hourly French rainfall, non-stationary analysis of heavy precipitation in Switzerland, and the study of global changes in seasonal extreme precipitation (Kallache et al., 2011). For the first topic, one of the main objectives of statistical climatology is to extract relevant information hidden in complex spatial-temporal climatological datasets. To identify spatial patterns, the most well-known statistical techniques are based on the concept of intra and inter clusters variances (like the k-means algorithm or EOF’s). As analyzing quantitatively extremes like heavy rainfall has become more and more prevalent for climatologists and hydrologists during the last decades, finding spatial patterns, simple and fast clustering tools tailored for extremes have been lacking. In this context, a novel algorithm based on multivariate extreme value theory was proposed. Comparing with classical clustering on weekly maxima of hourly precipitation recorded in France (Fall season, 92 stations, 1993-2011) shows that other patterns, specific to extremes events, were missing when employing traditional approaches.
The second example dealt with the inference of high rainfall return levels in a situation of non-stationarity in space and time. A new estimation technique of such high return levels based on semi-parametric approaches within the mathematical framework of Extreme Value Analysis was implemented (Smith et al., 2013).
The third example focused on exploring how anthropogenically-warmed climate is expected to sustain a larger increase of precipitation extremes. Here, new evidence was provided on global seasonal precipitation extremes for the 21st century, using 8 new high- resolution global climate model simulations. In the mid and high latitudes of both hemispheres, a significant intensification of extremes is evident in all seasons at the end of the century (Toreti et al., 2013).
The method provided an accurate estimation of heavy, extreme daily precipitation by statistically modeling exceedances above a high threshold and handling for spatio-temporal non-stationarities. The method is fast (no optimization is required) and flexible (non-parametric) and can be applied to large data sets from catchment area series to global climate models. Importantly, the method is computationally inexpensive (Naveau et al., 2011).
A similar approach was applied to high resolution CMIP5 Global Climate Models. Results show a remarkable intensification of extreme precipitation events at the end of the 21st century (especially under the RCP8.5 scenario) in all seasons. A parallelized R package of the method has been developed and is available for use by hydrologists, climatologists, engineers. This part of ACQWA research enabled the development of a state-of-the-art fast and flexible methodology that takes into account spatio-temporal non-stationarities in the statistical modeling of the daily precipitation exceedances distribution. The method can be also applied to large data sets. The new methodology allows in a computationally inexpensive approach to accurately characterize the statistical features of extreme precipitations, such as the return values illustrated in Fig. 8 for Switzerland (Stucki et al., 2012).
Figure 8. 50-year return levels of precipitation (mm) for Switzerland, 1961-2010. Spatial-temporal differences are evident with the north-south gradient and the tendency towards higher return levels in the 1980s as well as the recent decade over the Upper Rhone valley.
• Hot spells in the Rhone valley
In order to achieve a better understanding of the potential impacts of climate and climate change over the target areas, temperature extremes carefully and accurately studied. Here and for the first time, warm spells both in winter and summer were identified and analyzed for the last six decades (1951-2009) over the Upper and Southern Rhone valley by applying a recently developed approach (Toreti et al., 2012).
Figure 9:Intensity of winter and summer warm spells from 1951-2009 in the Upper Rhone Valley. Grey dots are associated with values of intensity calculated for each grid point. The black solid line represents the median of the values.
The approach is based on the identification of consecutive sequences of days (with a certain tolerance) with daily maximum temperature above a station-specified threshold. An increase in the intensity and frequency seems to have affected the two areas in the last 30 years in both seasons, especially after 1990 (Christidis et al., 2012). Furthermore, the higher intensity of the warm summer spell in the Southern Rhone valley, especially at the end of the period, can be highlighted. Finally, the 1983, 2003 and 2006 summer events are clearly visible in both valleys (Fig. 9).
The application of the state-of-the-art methods to recently released data sets gave the possibility to identify, estimate and analyze for the first time warm spells over the target areas both for winter and summer. The approach can be easily adapted for other regions of the world and it is available to the climate community (e.g. Hegerl et al., 2011).
• Hydrologically-relevant geomorphologic hazards
Changes in temperature and precipitation are likely to have a range of secondary effects on the occurrence of natural hazards, in particular also in mountain environments. However, while theoretical understanding exists for increased mass-movement activity as a consequence of predicted climate change, impacts can hardly be detected currently in observational records. One of the most obvious consequences of climate change at higher elevations is the glacier downwasting and related formation of ice-marginal lakes, ice avalanches and gravitational processes originating from the debuttressing of previously glacierized walls and hillslopes (Jomelli, 2012; Stoffel and Huggel, 2012). Glacier downwasting is likely to promote many rock slope failures at rather short future time scales, probably in the order of decades. Important effects of climate change on slope stability are also related to the warming and thawing of permafrost. Slopes currently underlain by degrading permafrost will probably become less stable at progressively higher altitudes with ongoing climate change. The probability of rock instability and the incidence of large (>106 m3) rockfalls will likely increase in a warming climate. A large number of recent slope failures have been documented in permafrost areas, related to increasing temperatures.
The impact of climate change on hydrologically-relevant geomorphologic hazards was further investigated by the development of (i) the slope stability component of the distributed hydrological model used for climate change impact analysis and (ii) an advanced slope stability model, which can simulate the effect of climate change on the occurrences of shallow landslides at high resolution in time and space.
The soil slip simulation tool that was implemented in the distributed hydrological model during this study is based on the concept of the infinite slope model (Tiranti et al., 2013). This model is aimed at exploring the ability to reproduce the slope response using a coarse terrain resolution at large catchment scales (river basin mesoscale, of the order of hundreds of km2) and thus for large scale simulations that are aimed at predicting hazard changes at the regional scale.
The second model was developed to investigate more in detail and at finer spatial scales the stability of slopes and their dependence on the detailed description of the involved hydrological and geotechnical processes. The HYDROlisthisis model is aimed at fine spatial scale resolution (of the order of the meter) and thus suitable for local scale (i.e. hill-slope and catchment scales of the order of 10 km2 at most) investigations of slope stability response to climate change forcings. The model consists of an hydrological and a geotechnical component, which are coupled together. The hydrological component takes into account the 3D variably saturated flow through soil, surface runoff, and hysteresis of the Soil Water Retention Curve (SWRC), topography-dependent solar radiation, potential evapotranspiration and root water uptake. The geotechnical component, which is based on a multidimensional limit equilibrium analysis, considers simple earth pressure conditions acting on the lateral sides of the soil column. These forces are computed using the coefficients of active, passive and at rest pressure, which are computed taking into account unsaturated conditions (Stoffel et al., 2014).
Both models showed, when applied to case studies, to be able to capture the observed slope instabilities at the scale for which they were developed. In particular, they showed significantly-better performance than statistical models, especially because they explicitly account for the spatial and temporal variability of the soil water dynamics, which is a key variable in the triggering of shallow landslides. This result is particularly valuable in the context of analyzing the shallow landslide hazard potential due to climate change, as they can explicitly account not only for the changes of the climatic forcing, but also for the consequences that this can have on the soil water dynamics and the related soil-vegetation medium response.
Climate and land use impacts on water availability and management: The case of the Aragón river basin, Pyrenees
Water availability is probably the main constraint for the development of modern agriculture, industry and tourism in the Ebro valley. The work developed within the ACQWA project provides robust information to water managers and policy makers concerning the response of river flows under projected environmental change. The study provides advance warning concerning limitations to maintain current water demand even if the regulation capacity is increased in headwater areas. Moreover, the study suggests that controlling land cover may be a mitigation strategy to minimize the probable reduction of available water resources (Lopez-Moreno et al., 2014).
A warming between 1-2°C and a decrease of precipitation between 5-20% is projected for the Spanish Pyrenees according to climate model simulations using the IPCC A1B scenario for the 2021-2050 period. A noticeable increase of warm events during winter months is also expected in the region. In the Pyrenees, an increase of 1ºC implies a 20% decrease in snow accumulation at 2000 m above sea level. This sensitivity increases at lower altitudes and decreases at upper elevations.
According to observed land cover change and the analysis of current aerial photography and remote sensing data, a scenario of land cover for the middle of the 21st century in the Pyrenees has been developed. Shrub areas are projected to evolve toward pine forests, with a 100 meters rise in the tree line and an increase of the shrub areas in the subalpine areas.
Increased vegetation in the basin could decrease annual stream flow in the Upper Aragón river basin by 16%, mainly in early spring, and autumn. Projected climate change could decrease annual stream flow by 13.8%, mainly in late spring and summer. Combined effects of forest regeneration and climate change may thus to reduce annual stream flows by 29.6%. Simulating the management of the main reservoir of the region using the modeled hydrological data, it is likely that serious difficulties to meet the current water demand, based on its current storage capacity (476 hm3) will emerge. If the current project to enlarge the reservoir to a capacity of 1059 hm3 is completed, the potential exists to apply multi-annual stream flow management, which will enhance the capacity to maintain the current water supply. However, under future climate and land cover scenarios, reservoir storage will rarely exceed half of the expected capacity, and the river flows downstream of the reservoir may be dramatically reduced (Lopez-Moreno et al, 2014).
Impacts on the hydropower sector
• The Swiss Rhone catchment
This research provided an analysis of the hydropower future in the context of climate change, opening of the electricity market to competition, decarbonisation of the energy system and, in some countries such as Switzerland and Germany, even phasing out of nuclear energy. The case study of the hydropower installation of Mattmark, located in the Rhone catchment represents the focal point of the analysis conducted in the context of ACQWA (Gaudard and Romerio, 2014). In addition, the study encompasses Swiss and European dynamics, in particular those related to energy policy and markets. The first part of the research, devoted to Mattmark, provides quantitative results, whereas the Swiss and European analyses are mainly qualitative.
The main interest of this study was to highlight the link between climate change, electricity markets and energy policy. In particular, it shows how electricity generation is affected by climate change, the opening of the electricity market to competition, as well as the development of micro and super-grids, new storage technologies and intermittent energy resources such as solar and wind energy. These factors also influence the added value created by hydropower, which represents an important source of revenue in mountain regions. The analysis takes into consideration the electricity demand, which is affected only modestly by climate change. The impact of its variation on the wholesale power market is estimated by means of econometric tools. Microeconomic models and techniques based on operational research are used to simulate the markets’ behaviour, hydropower reservoir management, as well as electricity generation.
Figure 9: Multiple constraints on electricity markets
In Fig. 10 (Gaudard et al., 2014), red boxes represent the drivers that are transforming the electricity system, which includes centralized and decentralized generation; storage services; consumption; supply and demand; flows of power through the electric lines (plain lines) and flows of information, notably price signals (dotted lines). Spot, future, balancing and ancillary services markets will determine the value of hydropower. Climate change will affect hydropower because of the possible reduction in surface water flows and seasonal shifts in water availability. The technological, economical and behavioural changes in the electricity system are, however, expected to exert a stronger impact on hydropower.
The outcome of the research conducted on the Mattmark site provides key information to decision-makers about the factors that will determine the future output and value of hydropower. Despite the margins of uncertainty, public bodies and private or public companies may use this information in the definition of their policies and strategies. The particular case of the renegotiation of the concessions for water rights, which represents a topical issue in several Alpine regions, is particularly relevant. In the case of Switzerland, most of the concessions will come to an end between 2030 and 2060. It should be mentioned here that, thanks to its flexibility, hydropower and reservoirs play a very important role in ensuring the security of electricity supply and the network’s stability in Europe.
• The Po Basin
Climate change can influence hydropower production in two ways: directly, through changes in precipitation and, as a consequence, in inflows; and indirectly through the electricity load because energy consumption varies with air temperature. This could be very important in determining the management of hydropower reservoirs and may cause conflicts with concurrent water uses.
In particular, reservoir management is aimed to provide the water resource when it is needed, transforming the natural regime, with its modulation across the year and its random fluctuations, in a regulated and more useful flow.
The main results from the ACQWA focus on the Po Basin have been summarized by Maran et al. (2014) as follows:
1. A large local variability in electricity production, which follows an analogous rainfall pattern: a reduction of 10% is estimated in the Val d’Aosta while a 20% increase is expected in the Toce valley;
2. The monthly modulation of power production throughout the year is expected to change: in both areas, the greater variability in summer inflows will affect the filling of hydropower reservoirs and, in some years, it will not be possible to use all the storage capacity.
3. the interannual variability of production is projected to increase.
All these effects are mostly the consequence of greater variability in river flows and the decrease in snow fall. The effects on hydropower production of variations in energy prices, i.e. the other major forcing factor, were also studied. The two Italian small case studies are very similar: in general, larger volumes of water are stored in winter and greater quantity of water are used in Spring, corresponding to a higher energy production in this period (Fig. 11).
Thanks to the collaborations of the ACQWA project, universities, research centers and private companies have had the possibility to deal with a high quantity of data produced by climatological and hydrological models at large spatiotemporal scales. These data allowed the implementation of detailed management models, simulating the management of hydropower plants. Large areas of north-western Italy were analyzed and an assessment of the impacts of climatic change on the water and electricity sectors in these zones has been made available.
Figure 11: Frequency distribution of yearly production for the Toce (left)
and the Valle d’Aosta (right) networks
However, new storage capacity implies the construction of new dams and reservoirs that could modify heavily the natural landscape of the Alpine region and are not easily accepted by local communities. This potential for conflict calls for the development and implementation of methodologies and processes to increase the adaptive capacity of water governance and management systems. ACQWA results could be used as an important aspect to be considered in decision making procedures regarding the development of new infrastructures for the water sector.
Impacts on the agricultural sector
• The Swiss Rhone catchment
In the Swiss Rhone catchment, irrigation has historically been important. Today, about 11’000 ha of land is irrigated with over half occupied by grasslands used for livestock production, followed by orchards and vineyards. This allocation of irrigation reflects the economic importance of the production of livestock and permanent crops, while arable crops play a minor role (Fuhrer and Jasper, 2012).
Over the past decades, a trend was identified towards a higher frequency of extreme droughts on timescales shorter than 2 months, whereas for longer timescales, no clear indications of a change over time could be found. For the future, the mean of several agro-climatic indices even suggest a shift towards warmer and rather wetter mean conditions during the growing seasons, but an increase in risks caused by high temperature (i.e. heat stress). The thermal growing season becomes longer, with potentially positive effects on pasture and livestock production, most pronounced at mountain sites, whereas at the valley bottom, a trend occurs towards increasing risks of frost in permanent crops due to an asynchronous change in the beginning of the growing season and late frost occurrence, and in heat stress for livestock (Fuhrer et al., 2014).
With increasing temperatures, water consumption through crop evapotranspiration increases, thus leading to additional irrigation needs to maintain optimal yields. This concerns much of the lower part of the valley, but also the pasture-dominated south-facing slopes, especially on soils with low water holding capacity. Simulations reveal a moderate average increase in water requirement for irrigation in 2021-2050 relative to 1981-2009. In the currently driest areas, additional potential water requirements from 1981-2009 to 2021-2049 would range between 0 and +200 mm per average growing season, depending on the climate scenario. During extremely dry years, such as 2003 or 2011, the increase would be much higher and could exceed the water availability in surface waters of smaller catchments with a nival runoff regime. An example would be the catchment of the Sionne (27.7 km2) where water is drawn through water channels (‘Suonen’) for grassland irrigation, and where discharge is controlled by snowmelt and rain. A much reduced area could be irrigated under these low-flow conditions, which may become more frequent in the future. At the catchment-scale, the estimated mean total water requirement is 32 × 106 m3 per year (1981–2009), and a 45% increase during the 2003 European heat wave in the driest area of the catchment. With an extreme scenario, the increase is about 44% by 2050, which is consistent with the values of 2003 as shown by Smith et al. (2012).
Coping with climate change in agriculture requires knowledge of possible trends in agro-climatic conditions, with a focus at the smaller scales such as catchments or sub-catchments where decisions are taken. At those scales, risks from water shortage during parts of the season are likely to become more important in the long-term future, particularly in view of a drastic decrease in glacier volume and snow cover. The methodology developed in this project can help to identify emerging water conflicts. Given the importance of grazed pastures in the mountain zones for their economic, ecological and cultural role, it can be expected that agricultural policy will continue to support the maintenance of this extensive form of agricultural production, which requires irrigation. In other parts of the valley, maintenance or even expansion of the production of high-value crops such as grapevines and fruit trees, or even vegetables, would continue to use water for irrigation. However, the amount is much smaller than for pastures, given the smaller surface area concerned, and the shorter length of the irrigation period. Moreover water resources at the valley bottom are much larger and may not be subject to the same extent to inter- and inner-annual variability on the timescale considered because the dependence on rain and snow is less than in sub-catchments with a nival regime. And, technical measures to optimize and control irrigation, or building reservoirs would be easier to implement.
Thus, in the shorter-term, the demand for water for irrigation will increase and put pressure on smaller rivers in catchments with little or no water supply from glaciers, particularly in years with limited precipitation and snowmelt during springtime. In the sub-catchments concerned, water management will play an even more essential role under future climate conditions than in the past. This improved water management should include both regulations regarding the allocation of water to different users of the same source, installation and management of reservoirs, and technical measures to improve the efficiency of irrigation by avoiding losses of distribution systems, evaporative losses, and excessive runoff due to over-application of water (Smith et al., 2012).
An important result of this assessment is that until 2050 major agro-climatic risks under climate change will likely be caused by high temperatures rather than by increased drought. Situations with high temperatures may have negative effects on both crop and livestock production. Measures to cope with this risk could include the use of additional water for cooling purposes, shifting crop cultivation windows, and shifts in cultivar selection in both arable crops and permanent crops (grapevines, fruit trees). In addition, extending the grazing period and shifting of grazing zones to higher and/or cooler parts of the catchment could help to avoid heat stress in animals. However, such shifts in intensification of grassland use may have negative consequences for other ecosystem services such as biodiversity, soil carbon storage, and nitrogen retention. Thus, coping with heat stress in livestock production will require careful consideration of the sensitivity of alpine grassland ecosystems to intensification (Fuhrer et al., 2014).
• The Po Basin
A basin-wide enquiry for the Po Basin has looked into the optimal policy that might facilitate adaptation to climate change in agriculture, in particular policies that can help to smooth water input fluctuations. The different roles of farmers and the public sector in adapting to these changes have been addressed. The first and most fundamental level of adaptation to climate change in agriculture occurs at the level of the local farmer. Farmers undertake strategies to adapt to the form of climate change that they are able to foresee, through observation of the recent trends in indicators such as average temperatures and average precipitation. However, they can do little to respond to the greater uncertainty inherent in climate change. Farmers’ adaptation to expected climate change will often take the form of investment in assets to shift water temporally, using locally appropriate water storage techniques (Bozzola and Swanson, 2014). It might also be possible for water management to be pursued through more efficient irrigation practices. Because of the low costs and relative abundance of the water resource in the Po basin, farmers have traditionally relied on inefficient irrigation methods, which are still one of the main causes of waste of fresh water resources. Local farmers adopt strategies to cope with expected climate change, but the important question for policy makers concerns the role of governance in supporting adaptation.
A key question here is the role that remains for policy in light of local adaptation. Evidence is given by this study that there are some impacts of climate change that farmers’ adaptations do not reach at present. Despite the farmers’ observed investments to adapt to mean changes in climate, variability in climate continues to impact crop yields. Table 1 shows the results of a simple correlation analysis of the relationship between variability in temperature, precipitation (during the months March to August) and agricultural yields in the Po basin.
Log
Yields Precipitation anomalies Temperature anomalies
Maize -0.16*** 0.02
Barley -0.28*** -0.15***
Soft Wheat -0.43*** -0.1*
Sugar Beet -0.21*** -0.2***
Soya -0.23*** 0.08
Table 1: Correlation between crops yields and standard deviation of precipitation and temperature. Significance Levels of the correlation coefficients at: ***1%; **5%; *10%.
The negative correlation between agricultural yields and observed anomalies indicates that agricultural production has suffered when unanticipated climatic variability occurs. In short, farmers’ adaptations appear to be based on individual (farmer-based) forecasts deriving from current trends in observed weather patterns, but are not responding to the increasing variabilities (Bozzola and Swanson, 2014).
Impacts on mountain tourism: the case of Valais, Switzerland
This part of the ACQWA project has investigated the climate-related and socio-economic drivers of winter tourism in the Swiss canton of Valais, as well as its expected water consumption in the future. Winter tourism is amongst the main economic sectors within the canton. The most significant finding is the industry’s exposure to reduced snow cover. A report by the OECD published in 2007 depicted Valais as showing little vulnerability to the reduction of snow cover. However, instead of approaching the tourism industry at the cantonal level, a more regional/local approach of the canton (48 resorts grouped into 17 regions) has been adopted. By comparing two average winters (2004-06) with a snow-poor winter with average economic conditions (2006-07) and a snow-rich winter at the beginning of the current economic crisis (2008-09), the study reveals high disparities between the resorts. This is an indication that vulnerability assessments and, more particularly, adaptation measures should be considered at a much lower level than has been done so far (Schaub and Andonova, 2012).
Valais seems to be more exposed or vulnerable to climate change than previously expected. Although the canton witnessed a smaller reduction in skier days than other regions during snow-poor winters, some resorts were strongly affected. And just in removing the biggest resort of the canton (Zermatt) from the statistics, the impact of the exceptionally warm winter of 2005-06 appears to be much bigger: Instead of losing only 4.9% in the number of skiers, the canton would have lost 7.8%. If the hotel sector proved less exposed to a snow-poor winter than the cable-car companies, the accommodation sector is facing serious difficulties in some regions when longer trends and socio-economic factors are taken into account. The canton could consider some support to the hotel sector so that they reach a “critical mass” of tourists, a strategy adopted by Austria, for instance.
The study can be seen partly as a “wake-up call” in many respects. The main political implications of these findings are that policy-makers need to take into account the big disparity existing between regions, and how vulnerable small and medium size resorts can be to the changes that are expected to impact tourism in years to come. The main challenges are to find ways for increasing cooperation amongst and within resorts and touristic regions, improve the promotion of the canton in general (the label “Valais Excellence” and the website of “Valais Tourism” constitute a good start in this respect), and more vitally, the ability to consider alternatives to skiing during warmer winters, especially for more vulnerable regions
.
For the regions themselves, it would be advantageous to build further winter-sport infrastructure exclusively in snow-reliable regions (thus increasing their center of gravity) and to adopt a more integrated structure. The target – the type and origin of population they would like to attract for winter holidays – should be chosen carefully. More generally, stakeholders should be aware of potential changes in the tourists’ preferences –a major unknown for the next decades. The future water consumption of the tourism industry can only be estimated. It can be said that the current coordination between the hydroelectric sector and the cable-car companies – in which the first sells water to the second for artificial snow-making purposes – should be maintained and perhaps even increased in some cases. More studies should be led to determine whether the current law on artificial snow in the canton should be improved in the future. In addition, the type of tourists (according to: age, country of origin, and the social class) they expect to attract should be considered. The preference of a high social class can change within a small time frame, and it is uncertain whether the middle class would continue its skiing activities at the same level as it does currently, or even if it can afford this time of recreation in the future.
Impacts on mountain forests
Climate change has the potential to substantially alter the provisioning of essential ecosystem services, and mountain regions are likely to be both particularly vulnerable and heterogeneous in their response to climate change. To date, few studies have attempted to quantify the impact on mountain ecosystems under a wide range of climate scenarios, including novel “2° scenarios” that are based on the assumption of an early stabilization of greenhouse gas concentrations in the atmosphere. Similarly, although mountain ecosystem services are crucial both locally (e.g. forest-mediated protection of human infrastructure from avalanches or rockfalls) and regionally (e.g. impact of ecosystems on runoff generation), there is a scarcity of studies focusing on shifts in ecosystem services.
A systematic assessment of climate change impacts on mountain forest properties and the ecosystem services they provide was performed in the context of the ACQWA project (Wolf et al., 2012). The focus was on five ecosystem services (carbon storage, runoff, timber production, diversity, and protection from natural hazards); four regionally downscaled climate scenarios that cover a wide range of possible future conditions were used; three complementary, state-of-the art models of forest dynamics that were used to simulate a wide range of forest properties were employed; and the simulations in two climatically contrasting case study areas (catchments) of the European Alps at several spatial scales, from the stand to the entire catchment, were performed.
Figure 12: Changes in forest-derived avalanche protection in the Saas valley (part of the Rhone catchment) under four climate scenarios as projected using the forest landscape model LandClim. Changes are shown as normalized difference compared to 2010 levels.
The results suggest that the sensitivity of mountain forest ecosystem services to a 2 °C warmer world depends heavily on the current climatic conditions of a region, the strong elevation gradients within a region, and the specific ecosystem services in question (Fig. 12). Model projections show that large negative impacts will occur at low and intermediate elevations in dryer/warmer regions. Here, relatively small climatic shifts result in negative drought-related impacts on forest ecosystem services (Manusch et al., 2013).
In contrast, at higher elevations, and in regions that are initially cool-wet, our simulations suggest that forest ecosystem services will be comparatively resistant to a 2 °C warmer world. It was furthermore found that considerable variation exists in the vulnerability of forest ecosystem services to climate change, with some services such as protection against rock-fall and avalanches being sensitive to 2 °C global climate change, but other services such as carbon storage being reasonably resistant. While these results indicate a heterogeneous response of mountain forest ecosystem services to climate change, the projected substantial reduction of some forest ecosystem services in dry regions suggests that even a 2 °C increase of global mean temperature cannot be seen as a universally “safe” boundary for the maintenance of mountain forest ecosystem services (Elkin et al., 2013)
The results achieved in the ACQWA context rely on research activities that were begun more than a decade ago and benefited considerably from the funding made available in ACQWA in the sense of a synthesis and further development of existing approaches. Beyond this, the added value of the project was the availability of state-of-the-art climate scenarios and the comprehensive focus of the study, as ACQWA has always emphasized cross-sectoral activities rather than isolated impact assessments of individual components of mountain systems.
Impacts on mountain aquatic ecosystems
• Mountain lakes in the Gran Paradiso National Park, Valle d’Aosta, Italy
During the ACQWA project, measurements of the physical, chemical and biological parameters in 20 high-altitude, ultra-oligotrophic lakes in the Gran Paradiso National Park (PNGP) were undertaken. All these lakes are characterized by extreme seasonality; most are naturally fishless, while some were stocked with fish about forty years ago (Tiberti et al., 2012). Both fishless and stocked lakes were sampled in order to determine the effects of introduced fish and of the interplay between the introduction of alien species, strong seasonality and shifts in environmental conditions. Among the different measurements, zooplankton densities at different depths in order to gain insight on their vertical distribution, their migratory behavior along the water column and the response of zooplankton to vertical abiotic gradients (e.g. light intensity and UV gradients). The PNGP measurements were complemented by the development of deterministic one-layer models to simulate the dynamics of high-altitude lake ecosystems, in order to estimate the response of the lake ecosystems to climate and environmental change (temperature, duration of ice cover, etc). These observational and modeling studies have enabled some of the first quantitative determinations of the effects of introduced fish and of the interplay between stocking and environmental change (Tiberti et al., 2013).
• Impacts on Pyrenean aquatic ecosystems
In mountainous river basins, the water source balance (i.e. the contribution of rainfall, glacier-melt, snowmelt, and groundwater to flow) is extremely sensitive to climate change. Projected warming is likely to alter stream flow quantity/quality regimes and modify in-stream physico-chemical habitat. As a result, biodiversity patterns of alpine running water organisms are highly likely to be altered, as vulnerable species must either adapt physiologically and/or genetically or migrate to more suitable habitats to persist (Finn et al., 2013).
As part of the ACQWA project an inter-disciplinary approach was adopted to investigate alpine stream ecosystem responses to climate change. Both the drivers of change (i.e. climate) and responses to change (i.e. hydrology – habitat – biota) were quantified using a combination of scenario simulation, space for time substitution surveys and in-situ experimental work. Macro-invertebrate community structure and function, population genetics, habitat characteristics and water source contributions were recorded at 26 sites representing a gradient of glacial influence, across five river basins in the French Pyrenees. Contemporary river flow, habitat, taxonomic and genetic patterns were related to both glacier cover and melt-water contribution to bulk discharge. Downscaled climate data was then used to drive (i) a watershed simulation (TOPKAPI) of river flow and glacier/snow melt dynamics and (ii) a water temperature regression model. Temperature and hydrological projections were then linked to observed patterns to predict future hydro-ecological change (Khamis et al., 2013a).
Further work, to improve understanding of finer scale hydrological and biological process , investigated; (i) the response of an alpine spring community to the introduction of a large bodied invertebrate predator (predicted to expand its range) and (ii) the inter-annual variability of the heat budget and thermal dynamics of a glacier-fed river reach.
Until now application and testing of the water source contribution approach as a hydro-ecological management tool, has been limited to single river basins. Results from this study highlight the utility of this approach for both management and prediction of climate driven ecosystem change. Furthermore, knowledge transfer between project partners enabled the successful completion of watershed simulations for a glacierized river basin in the relatively data poor Pyrenees, where until now few simulations have been conducted. This provided a unique insight into future hydrological and physico-chemical habitat change and the potential implications for biodiversity. Quantification of melt-water contribution enabled the identification of climate sensitive macro-invertebrate taxa which are expected to exhibit considerable range contraction as glacier retreat (Khamis et al., 2013b). Trait profiles of glacial stream taxa were also recorded and were shown to have the potential to act as indicators of changing water source dynamics (i.e. reduced glacier contribution to discharge), particularly when comparing sites from multiple biogeographic regions. Improved process understanding from experimental work and high resolution hydro-meteorological observations could prove valuable for further development of deterministic temperature models and species distribution models.
The Pyrenees represent the southern limit of contemporary glaciation in Europe, and only small cirque glaciers remain of the extensive ice cover which existed during the Little Ice Age. Thus, findings here can be viewed as a future analogue for mountain ranges which currently have significant glacier cover but which will shrink in the future. The successful application of the water source approach in a changing cryosphere across multiple river basins highlights its potential for use in future alpine conservation planning. Furthermore, the coupling of hydrological simulation and space for time surveys has a potential to inform future baseline conditions and could prove valuable for determining appropriate climate adaption conservation strategies (Khamis et al., 2014b).
Lessons learned from non-European regions
• Central Andes, Aconcagua Catchment, Chile
Chile represents a highly contrasting case to the European basins, from both a physical and a governance perspective. Water rights are a marketable commodity with minimal environmental regulation and no sectorial prioritization. Government institutions are highly centralized, with limited agency and capacity of water managers at the regional level. Increasing drought periods (from a combination of reduced summer runoff and altering precipitation patterns) are likely to compound current issues relating to the overuse of surface and groundwater from both legal and illegal abstractions in the Aconcagua Basin (Hill and Allan, 2014).
Major challenges from the governance perspective relate to the accuracy and applicability of monitoring data; lack of available, accurate, systematized and accessible information on water rights, water judgments, water market and prices, health and availability of water resources; climate data and uncertainty calculations are not used in planning; constricted agency and capacity of technical experts at the regional level; lack of trust and agreement on scientific studies and hydrological data that blocks concensus building; lack of a formal flexible conflict resolution mechanism.
A common thread across the different case studies are the challenges or opportunities presented by regional networks, actors and the differing levels of trust that can be capitalized for more integrated and longer term planning of water resources management. In the Aconcagua, there are a number of institutions that could be fostered to improve adaptive capacity. The Mesa del Agua is a collaborative attempt to strengthen inter-sectoral cooperation for the Aconcagua Project that should be made more inclusive with a broader, climate-related, goal orientation. While there is potential flexibility through the water rights system, denoting flexibility and autonomy to react quickly to changing conditions in the river, improvements to trust between river sections and more expedient and flexible conflict resolution mechanisms could greatly enhance its role in adaptation to drought conditions (Hill and Allan, 2014).
• Water governance in the Mendoza watershed in Cuyo, Argentina
During the ACQWA project, an analysis of the institutional environment in Argentina, its attributes and functionalities was undertaken, against the requirements to adequately respond to expected impacts on water resources and governance (Girón et al. 2014). Further, research focused on how the inherent constraints resulting from a dysfunctional institutional system have a potential to undermine adaptive capacity and adaptation actions to climate change. During the last three decades a weak institutional context created problems of credibility, coordination and cooperation and has had severe impacts on the quality of public policies. A weak institutional context is defined by low enforcement of the rules and/or by an application that is broadly discretional and one in which institutional change is frequent and radical. Hence, Argentina’s public policies tend to be unstable, poorly coordinated, weakly enforced, and highly rigid, and the decision making process lacks credibility. This persistent environment of instability, policy volatility, political short-termism and precarious enforcement pervades the water governance regime across scales and in the case study area, even in a formally highly decentralized federal system (as in other regions and sectors affected by the impacts of climate change). The decision making context, including drivers of water use, in a complex governance landscape, reinforces informational and cognitive, financial, social and cultural barriers to implement adaptation efforts and enhance adaptive capacities to improve the resilience of the social-ecological system. The scientific evidence that under projected climate change conditions the water resources of the Mendoza region could be reduced in a significant way thus requires the adoption of adaptive governance principles whose application is obstructed by rigidity, lack of cooperation and coordination, serial replacement (as opposed to predictable law and institutional structures) and fragmented or truncated learning processes and knowledge creation and dissemination.
• Climate change impacts on Kyrgyzstan
Within the ACQWA project, the Syr Darya river basin in the Tien Shan ranges have been chosen as a pilot area for a specific case study. The goal is to illustrate the impact of climate change on a transboundary river, where complex responses result from asymmetric power relations and less robust institutions. Kyrgyzstan has the role of a water tower as many rivers – such as the Syr Darya – originate within the country’s territory. The difficulty for the Central Asian states is to apply the principles of equitable use of water and to agree on a balanced reservoir management, which would allow the generation of energy in winter – benefiting upstream countries such as Kyrgyzstan – and irrigation for large-scale agriculture in summer – benefiting downstream countries such as Uzbekistan. Current challenges in the water operating regime are likely to be exacerbated by climate change impacts as water shortages during summer become more frequent with expected decreases in summer precipitation and reduced glacial meltwater releases due to smaller glacier volume (Sorg et al., 2012).
The Tien Shan mountains are located in the center of the Eurasian continent; the distance to the nearest sea exceeds 1500 km. Arid and semi-arid climate conditions result from atmospheric flows coming from the Atlantic and the Arctic, which lose major parts of moisture over the European territories and Siberia. Research within the framework of the ACQWA project and based on previous studies under the UNFCCC have demonstrated that the most probable climate change scenario fits the emission scenario B2-MESSAGE. By 2050, Kyrgyzstan will thus experience an increase in average annual temperature by +2.5°С and an increase in annual precipitation sums by +2.5%, with increasing amounts in winter and decreasing amounts in summer.
Sorg et al. (2012) show that these climate changes may have various impacts, in particular water shortages exacerbated by a decrease in glacial meltwater releases in the long-term due to reduced glacier volume. Glacier shrinkage is most pronounced in peripheral, lower-elevation ranges near the densely populated forelands, where summers are dry and where snow and glacial meltwater is essential for water availability. Under the high greenhouse-gas emissions SRES A2 scenario, 28-35% of today’s glacier volume in the Syr Darya catchment may melt by 2050.
Furthermore, shifts in seasonal runoff maxima have already been observed in Kyrgyzstan, and summer runoff will probably further decrease if precipitation and discharge from thawing permafrost bodies do not compensate sufficiently for water shortfalls. Runoff may decrease by 15%. Currently, however, river runoff in the formation zone of the Syr Darya basin is increasing due to intensive glacier melting. This trend is likely to continue up to 2025. After 2025, however, contribution of glaciers to river runoff in the headwaters of the Syr Darya will decrease irreversibly; as a consequence, the total hydropower potential of the Syr Darya basin may decrease by up to 15%.
Potential Impact:
The ACQWA project was formulated in response to the first call for climate-relevant projects under the EU 7th R&D Framework Programme (FP7). The philosophy of the project was based on the need to accurately assess the vulnerability of water resources in high-elevation, mid-latitude populated mountain regions. In such regions, declining snow and ice in a warmer climate are likely to strongly affect hydrological regimes, in terms of quantity, seasonality, and also quality. As a consequence of changing water availability, both upland and populated lowland areas will be affected. Rivalries and conflicts of interest may emerge as economic sectors such as agriculture, tourism or hydropower compete for water that may no longer available in sufficient quantities or at the right time of the year for these sectors to function. The challenge for the ACQWA project was thus to estimate as accurately as possible future changes in order to prepare the way for appropriate adaptation strategies and improved water governance. The project has enabled a suite of state-of-the art models to be applied, adapted, or developed to address many of the issues related to a changing physical world and to the socio-economic impacts that these changes will inevitable generate. Model results have also been used to assess how robust current water governance strategies are and what adaptations may be needed to alleviate the most negative impacts of climate change on water resources and water use.
Introduction
As the evidence for human induced climate change becomes clearer, so too does the realisation that its effects will have impacts on socio-economic systems and terrestrial ecosystems with multiple implications for society. Mountains are recognised as very sensitive physical environments with local populations that are highly exposed to rapid changes in the resource base on which their economic livelihoods are dependant. Moreover, policy priorities for alpine regions are often set by downstream actors, leading to trade-offs across different policy contexts that have the potential to be further exacerbated by changes in the hydro-climatic environment. While governance is well recognised as a core issue in current water resource related challenges, to date there is still a paucity of information on how adaptable water governance regimes in mountain areas could be to hydro-climatic changes impacts and the socio-economic impacts that these changes imply.
Large integrating projects generally represent a step forward in furthering our understanding of various complex processes and interactions between environmental, economic, social, and technological systems. The ACQWA project is no exception to this rule, and the five years of research has indeed enabled a number of issues to be refined and clarified, but has also identified problem areas that would need to be addressed in future investigations of this nature.
In January 2011, the ACQWA project organised a workshop in Riederalp, Switzerland, where over 25 EU projects focusing on water resources and water management were represented. Institutional and financial obstacles to data access for use in modelling exercises were identified, and gaps in scientific knowledge that contribute to uncertainty were highlighted. A working paper was subsequently published in 2012 in Environmental Science and Policy to report on the main conclusions of this crucial meeting. The discussions summarised in the paper have identified a number of sectors where these gaps often represent barriers to successful research outcomes, and suggested ways and means of alleviating some of these difficulties. A major issue that has been raised is that of data for research purposes. Policies aimed at ensuring free and unrestricted access to data, especially those generated by the numerous research projects that focus on issues of water availability, quality and management have been recommended. Implementation of the recommendations formulated in the Environmental Science and Policy paper may help pave the way for a more rapid and efficient production of research results that are of importance for policy guidance at the local, national and supra-national (EU) levels.
Socio-Economic Implications
ACQWA utilised advanced modelling techniques to quantify the influence of climatic change on the major determinants of river discharge at various time and space scales, and analyse their impact on society and economy. The main focus was to develop continuous transient scenarios from the 1960s up to 2050. By focussing on developing scenarios of change up to 2050 for a set of river basins, the project aims to develop climate information downscaled to temporal and spatial scales that are more useful to the challenges decision makers face (Beniston et al., 2011). This shortened modelling horizon allows for a more realistic assessment of the potential impact on the governance and socio-economic system components.
Key findings for different socio-economic sectors and themes
The following sections highlight the key findings from the ACQWA project with regards to socio-economic implications in key sectors across the case studies, namely, hydropower, agriculture and tourism (Fatichi et al., 2013a; Fatichi et al., 2013b; Fuhrer and Jasper, 2012; Fuhrer et al., 2013; Gaudard et al., 2013; Gaudard and Romerio, 2013; Gaudard et al., 2014; Gobiet et al., 2014; Hill Clarvis et al., 2013b; Stoffel et al., 2013; Toreti et al., 2013). In mountain regions, climate change takes on particular significance since snow and ice melt represent a large stream-flow component and a vital local resource for freshwater supply, hydropower generation, irrigation, tourism activities, and other industrial uses. Glaciers and snow pack act as vital natural storage systems, storing water as snow and ice through the wetter winter periods and releasing these provisions as flows during the drier summer months. Changes in precipitation and temperature will therefore impact both the quantity and timing of water available across these different sectors. Changes in the frequency or intensity of hazards (e.g. floods, glacier lake outburst floods, rock falls and landslides) are also likely to have a range of socio-economic implications as temperature and precipitation changes affect slope stability, glacier melt, snow melt and the zero isotherm.
Hydropower
• Variability in glacier retreat patterns (size, aspect, shape, debris cover, etc.) has consequences for the management of hydropower plants and dams, which depend primarily on snow- and ice-melt.
• Reduction in surface water flows and seasonal shifts in water availability (more availability of water in the earlier months of the year and a longer summer period with lower run-off) will impact hydropower. Climate change also indirectly affects electricity load because energy consumption varies with air temperature.
• Technological, economic and behavioural changes in the electricity system are, however, expected to exert a stronger impact on hydropower.
• In the Po region, greater variability in river flows and decrease in snow fall will affect the filling of hydropower reservoirs (e.g. decreasing ability to use all the storage capacity) and increase the inter-annual variability of electricity production.
• Storage-hydropower plants are a more flexible technology with modifiable production periods, whose revenues are less vulnerable to shifts in seasonality than run-of-river.
• While more even contribution from runoff might advantage reservoir management, a decrease in total annual runoff expected for reservoirs fed by ice melt is likely to negatively affect production.
Agriculture
• Until 2050 major agro-climatic risks will likely be caused by high temperatures rather than by increased drought (negative effects on both crop and livestock production).
• With increasing temperatures, water consumption through crop evapotranspiration increases is likely to lead to additional irrigation demands to maintain optimal yields (e.g. +10% in July at Visp across a range of climate scenarios up to 2049).
• High demand for water for irrigation will put additional pressure on small rivers in catchments with little or no water supply from glaciers, while larger water sources in valley may not be subject to the same extent of variability.
• In drier areas with low summer precipitation (e.g. valley floor and the south-facing slopes), potential water shortages for crop growth would be likely, requiring more irrigation to maintain optimal crop yields (max. +35%).
• In extremely dry years irrigation requirement could potentially exceed surface water availability in smaller catchments with a nival runoff regime (e.g. Sionne) where water is drawn through small irrigation channels for grassland irrigation.
• In the upper Rhone valley, improved water management should include both regulations regarding the allocation of water to different users of the same source, installation and management of reservoirs, and technical measures to improve the efficiency of irrigation by avoiding losses of distribution systems, evaporative losses, and excessive runoff due to over-application of water.
• Shifts in intensification of grassland use may have negative consequences for other ecosystem services such as biodiversity, soil carbon storage, and nitrogen retention.
Tourism
• A more local approach to winter tourism exposure to climate change in the Rhone catchment (comparing across average winters, a snow-poor winter with average economic conditions and a snow-rich winter at the beginning of the current economic crisis) reveals high disparities between mountain resorts and a higher vulnerability than regional approaches have suggested.
• Although Valais witnessed a smaller reduction in skier days than other regions during snow-poor winters, some resorts were strongly affected, with a large disparity in the vulnerability of resorts (e.g. small and medium resorts).
• The hotel sector is less exposed to a snow-poor winter than the cable-car companies, but faces difficulties in some areas when longer trends and socio-economic factors are taken into account.
• There is a real need to increase cooperation among and within resorts and touristic regions, enhance co-ordination of water uses (e.g. between hydro and cable-car companies), to improve the promotion of the Canton in general, to improve the ability to consider alternatives to skiing during warmer winters (notably in more vulnerable regions), and to improve current regulation on artificial snow-making.
Policy and Governance Implications
These impacts therefore need to be taken into account in the rules, rights and policies that structure the way in which water resources are managed across its multiples uses as well as the methods for protecting society from hydro-climatic related hazards. However, traditional governance and management approaches have often unsuccessfully coped with current internal or stochastic climate variability. Practitioners and scholars therefore not only need to address current rules and practices for managing historical challenges within the current envelope of uncertainty, but also to assess the adaptability of current frameworks for managing water resources and hazards to future climate change impacts.
In the Po basin, despite the farmers’ observed investments to adapt to mean changes in climate at local level, unanticipated variability in climate continues to impact crop yields. Policy interventions (integration or water resources, water storage) must therefore deal with residual uncertainty remaining after local adaptation to climate change (Bozzola and Swanson, 2013). In the Upper Rhone basin, the increasing heterogeneity of precipitation and late summer reductions in run-off from reduced glacier melt are likely to further exacerbate current local critical situations, bottleneck periods for local water supply, which are themselves related to: high levels of autonomy at municipal level that block longer term catchment scale planning and smoothing of bottleneck periods; lack of formal mechanisms to manage competition across catchment areas; lack of rules on emerging challenges and uses(Hill Clarvis et al., 2013b).
Earlier snow melt and shifting glacier melt patterns are likely to impact the inflexible and long term user rights that govern uses such as hydropower by introducing an extra layer of uncertainty and shifting the hydrological baselines upon which fixed and un-integrated rules and policies are based on at different governance scales and across different sectors (Hill Clarvis et al., 2013a). In response to climate change impacts as well as the continuing challenges concerning uncertainty, adaptation strategies are recommended that are no-regret, flexible and iterative, that allow for safety margins and redundancy in new investments, take a long term and social and green infrastructural approach (to complement grey infrastructure), and that integrate both adaptation and mitigation requirements.
In addition to the challenges of communicating and adapting to different scales of climate related challenge (i.e. internal or stochastic variability versus increased uncertainty and variability from climate change impacts), water resources governance and management adaptation must also deal with a number of other scale based challenges. These include: trade-offs across risk response when short term adaptation actions potentially undermining long term social-ecological resilience (Adger et al., 2011; Hill and Engle, 2013); balancing out proactive and reactive responses, as well as responses to multiple forms of stress at different magnitudes of physical change and scales of governance (Hill, 2013a); trade-offs between narrowly defined adaptation policies and other policy frameworks and economic sectors;
Cross-scale and sector trade-offs need to be better understood in the process of developing adaptation and broader environmental policy, plans and projects that address the impacts of climate change (Adger et al., 2011; Hill and Engle, 2013). Furthermore, downscaling climate and socio-economic impacts to finer temporal and spatial scales can give decision makers a clearer view of how these tensions might play out in a future climate, and therefore allow them to better prepare for and respond to them.
At the European level, while European climate change and adaptation policy is still in its infancy, reflection of how to account for climate change impacts is happening in relation to the Floods Directive and the next round of River Basin Management Planning for 2015 according to the Water Framework Directive. In Switzerland, a comprehensive climate change adaptation policy is in early development at the federal level. However, water resources use and management transcends multiple policy frameworks, and underlying trade-offs across different policy frameworks and sectors must still be better accounted for and remediated as we move into an era of potentially less manageable pressures from the hydro climatic environment.
$
Outreach and societal implications
Societal Implications: Understanding and managing uncertainty
Not only are the driving forces of climate highly uncertain, but fundamental scientific knowledge gaps limit the reliability of model projections (Hallegatte, 2009) with uncertainties in how climatic and non-climatic pressures will interact on different aspects of hydrology and ecology (Wilby et al., 2010). Traditional decision making tools, water management and infrastructure have however tended not to be developed to take account of the broader levels of uncertainty produced by climate change projections (Hallegatte, 2009). Despite scientific efforts focussed on the reduction of uncertainty, through enhanced data collection and modelling (Schneider and Kuntz-Duriseti, 2002), the uncertainty surrounding the specificity of climate change impacts remains a major challenge to planning and managing for future hydro-climatic conditions. However, policy makers, decision makers and water managers are increasingly recognising the need to develop better tools to manage and cope with both existing and increasing levels of uncertainty from climate variability and climate change impacts (Hallegatte, 2009).
While in many areas better managing for current and future levels of uncertainty is key, effective monitoring and data provision remains a critical component to remediating challenges relating to the current lack of knowledge and data on hydro-climatic, other environmental criteria and usage particularly in nations that are most at risk with the least data infrastructure. This is and should remain a core component of the on-going work on climate services, WMO standards and data sharing at international as well as national levels. While this is vital in developing countries, wealthier nations should not cease to continue to maintain and where possible improve their own monitoring and data infrastructure.
The growing body of work on climate change and climate change impact projections, adaptation and adaptive capacity have developed a broad set of principles, determinants and indicators for adaptive governance and management of water resources. However, there remains a need for clearer guidance on practicable mechanisms and actions at different levels of policy making, institutions and water management (Folke, 2011). This could help better guide the concrete changes that need to be made in governance and management frameworks to improve their sustainability and adaptability. Often there has been too much focus on the technological aspects of water resources management and adaptation and not enough on the governance and social infrastructural aspects of water systems (Adger et al., 2009; Hill, 2013b).
Outreach
The ACQWA project has produced over 280 scientific publications, a number of special issues in top tier journals such as Science of the Total Environment and Environmental Science and Policy, and over 130 instances of dissemination and outreach activities at academic conferences, public presentations, policy maker briefing sessions, newspaper and magazine articles and television news stories.
Conclusion
There is a clear need for a more integrated and comprehensive approach to water use and management. In particular, beyond the conventional water basin management perspective, there is a need to consider other socio-economic factors and the manner in which water policies interact with, or are affected by, other policies at the local, national, and supra-national levels. As an example, it is unclear whether current EU water policies are consistent with energy, agriculture, and other industrial policies.
The problems highlighted during the Riederalp meeting and summarised in the Environmental Science and Policy paper are also related to the inconsistencies between physical and socio-economic data and models. For example, figures related to water use may not be available at the temporal and spatial detail required by hydrologic models. Hydrological information is often based on basins whereas economic (and social) data is aggregated into administration regions. Thus, economic and physical data are often incompatible, because they are collected by different entities for different purposes. Future research should thus address the development of compatible data sets and the conversion between different data formats, as well as the development of toolboxes for up-scaling, downscaling and bias correcting data. Furthermore, the use of water in production processes is often not mediated by the market. The use of economic flexibility mechanisms in the allocation of water resources is quite rare, despite their potential in improving the efficiency of water resources allocation. More research and policy initiatives in this direction are thus necessary.
Finally, many scientists working in large integrated projects highlight a large gap between Science and Policy. This is certainly at least partly due to problems of communicating in an appropriate manner the key research results that would be of use to policy-relevant strategies. Awareness of this problem is increasing within the EC and other policy institutions, and hopefully this new momentum will be sustained over time so that conclusions from EU and other water-relevant projects will be widely incorporated into future policies at the local, national, and supra-national levels. Ultimately, the implementation of guidelines, maybe even an EU Directive, on the good governance of data (sharing) could be envisaged as a possible framework, providing advice and general rules on data formats and standards, data storage after project completion or the general terms of access.
List of Websites:
www.acqwa.ch
Project co-ordinator name: Prof. Martin Beniston
Project co-ordinator organisation: UNIVERSITE DE GENEVE
Phone: +41 22 379 07 69
Fax: +41 22 379 07 44
E-mail: Martin.Beniston@unige.ch
Project website address: www.acqwa.ch
The overarching goal of the ACQWA project has been to quantify and assess the consequences of shifts in temperature and precipitation patterns, as well as the behaviour of snow and ice in many mountain regions in a changing climate. These transformations of mountain climates and environments are expected to change the quantity, seasonality, and possibly also the quality of water originating in mountains and uplands. As a result, shifts in water availability will affect both upland and populated lowland areas. Economic sectors such as agriculture, tourism or hydropower may enter into rivalries if water is no longer available in sufficient quantities or at the right time of the year. The challenge of the ACQWA project has thus been to estimate as accurately as possible future changes in order to prepare the way for appropriate adaptation strategies and improved water governance.
The project identified the vulnerability of water resources in mountain regions such as the European Alps, the Central Chilean Andes, and the mountains of Central Asia (Kyrgyzstan) where declining snow and ice are likely to strongly affect hydrological regimes in a warmer climate. Model results were then used to quantify the environmental, economic and social impacts of changing water resources in order to assess how robust current water governance strategies are and what adaptations may be needed in order to alleviate the most negative impacts of climate change on water resources and water use.
Current generations of state-of-the-art models were applied to various interacting elements of the climate system, that include regional atmospheric processes in complex terrain, snow and ice, vegetation, and hydrology. Results from the suite of models have enabled to project shifts in water regimes in a warmer climate in the case-study regions mentioned above. Observations, targeted models, and methodologies from the social sciences were applied to the impacts analyses on sectors such as tourism, agriculture and hydropower which will certainly be strongly impacted upon by changing water regimes. The results from these different approaches then served to develop a portfolio of recommendations for adaptation and updated water governance strategies.
In summary, the ACQWA project explored specific science and policy themes that included in particular the following elements:
• An integrated modelling approach to enable accurate projections of changes in the seasonality and quantity of runoff in the river basins under scrutiny in the ACQWA project;
• Identification of key economic impacts on a number of important water-relevant sectors, including the possible compounded effect of changing frequencies and intensities of extreme events;
• An assessment of the possible rivalries between economic actors that will be faced with changing water resources;
• A portfolio of possible water governance strategies to alleviate future problems of water allocation and use, of relevance to upcoming revisions of the EU Water Framework Directive.
Project Context and Objectives:
Beniston and Stoffel (2014) have emphasized the fact that mountains are recognized as “sentinels for environmental change”, in the sense that they exhibit dynamics in physical and biological systems that are often more readily identifiable than in other geographical entities of the globe (e.g. Loarie et al., 2009; Engler et al., 2011; Gobiet et al., 2014). The study of cryospheric, hydrologic, geomorphic and socio-economic change in sensitive mountain regions enables to further our understanding of how an important part of the terrestrial environment responds to, is affected by, and may adapt to rapid and sustained changes in temperature and precipitation regimes (Beniston, 2009; IPCC, 2013).
The importance of mountain regions as a provider of numerous ecosystem services was already recognized at the United Nations Conference on Environment and Development (UNCED, Rio de Janeiro, Brazil, 1992); mountain regions were included under Agenda 21 of the UNCED conference, which mentions under its Article 13 (UN, 1992) that “Mountains are important sources of water, energy, minerals, forest and agricultural products and areas of recreation. They are storehouses of biological diversity, home to endangered species and an essential part of the global ecosystem. From the Andes to the Himalayas, and from Southeast Asia to East and Central Africa, there is serious ecological deterioration. Most mountain areas are experiencing environmental degradation.”
Among these visible impacts, changes in glacier length and volume are perhaps the most spectacular manifestations of climate impacts in mountains; currently and with few exceptions, mountain glaciers from the equatorial to the high latitudes are experience glacier shrinkage (Paul, 2011; Bolch et al., 2012), highlighting the fact that this is a global phenomenon. Shifts in mountain snow-pack behavior in the past decades has also been observed, with collateral impacts on the timing of snow-pack melting and thus of surface runoff (e.g. Beniston, 2010), and also an influence on the start of the vegetation period for certain mountain plant species (e.g. Gottfried et al., 2012).
More subtle changes are reported for mountain ecosystems, in part because of the longer timescales involved in biological systems compared to mountain cryospheric systems, for example, and also because certain species are more adaptable than others (Dubuis et al., 2013), thus resulting in greater difficulties in attributing cause-to-effect relationships of climate change. In mountains, the transitions between biological entities and vegetation (ecotones) can occur over short distances, contrarily to what takes place in lowland plains (e.g. Gosz, 1993). Many changes in vegetation patterns related to sustained shifts in environmental conditions can be identified in these ecotone transition zones, as shown for example by Gottfried et al. (2012) at the boundary between high alpine vegetation and the snowline.
Finally, mountains are also the locale for numerous socio-economic activities, in particular tourism, agriculture, industry, mining, and energy (hydropower). These sectors are all sensitive to climate change, since climate exerts the essential controls on the availability of snow for ski tourism (e.g. Uhlmann et al., 2009; Morrison and Pickering, 2013), or of water for mountain agriculture, hydropower, and for mineral exploitation (e.g. Finger et al., 2012; Fuhrer et al., 2014; Gaudard et al., 2014), for example. However, it should be emphasized here that low priorities to sustainable land-use and natural resource management in many mountain regions in the world implies that changes in forest resources, mountain agriculture and water resources are driven not only environmental change but also by economic and demographic factors (Beniston, 2003).
It is in this context of rapid dynamics of environmental and social change in mountain regions that the ACQWA project (Assessing Climate impacts on the Quantity and quality of Water) was developed, in response to the first call for climate-relevant projects under the 7th European Union Framework R&D Programme (EU-FP7). The large integrating project, coordinated by the University of Geneva from 01.10.2008 – 31.03.2014 under EC contract nr 212250, brought together a consortium of 30 partners in 10 countries and on three continents for a total of 37 different research, public, or private research entities representing over 100 scientists.
The overarching goal of the project was to assess the vulnerability of water resources in mountain regions where snow and ice represent a major input of water for rivers originating in mountains, and where declining snow amounts and receding glaciers in a warmer climate are likely to have profound impacts on hydrological regimes. Future shifts in temperature and precipitation patterns, and changes in the behavior of snow and ice in many mountain regions will change the quantity, seasonality, and possibly also the quality of water originating in mountains and uplands (Sorg et al., 2012; Immerzeel et al., 2013). As a result, changing water availability will affect both upland and populated lowland areas (Hill et al., in press; Sorg et al., in press). The challenge of the ACQWA project has thus been to estimate as accurately as possible future changes in water availability, and the impacts these changes may impose on a range of water-dependent economic systems (Fig. 1).
Figure 1: Structure of the ACQWA project and its main components
The flow diagram in Fig. 1 illustrates the broad structure of the ACQWA project. Current generation of state-of-the-art models (e.g. Themessl et al., 2010; Gobiet et al., 2014; Heinrich et al., 2013) were applied to various interacting elements of the climate system, that include regional atmospheric processes in complex terrain, snow and ice, vegetation, and hydrology in order to project shifts in water regimes in a warmer climate in mountain regions as diverse as the European Alps, the Central Andes of Chile and Argentina, and the mountains of Central Asia (Kyrgyzstan). Observations, targeted models, and methodologies from both the social and the natural sciences were then applied to conduct analyses of climate impacts on sectors such as tourism, agriculture and hydropower which could be strongly influenced by changing water regimes. Because these economic sectors and other water-dependent industries may well enter experience conflicts of interests and rivalries if water is no longer available in sufficient quantities or at the right time of the year, a further goal of the ACQWA project was thus to define a portfolio of proposals to pave the way for appropriate adaptation strategies and improved water governance (Hill and Allan, in press; Hill and Engle, 2013). These are designed to help alleviate the more negative impacts of climatic change on water resources and to reduce the risks of conflict between the economic actors most affected by these changes.
The focus on water has been the key element of the ACQWA proposal, because it is an essential resource for human populations, animal and plant communities. Water is relevant in every aspect of mountain systems, in the physical, biological and socio-economic systems. It directly influences the energy supply (hydropower), tourism (snow, water usage, glaciers), forestry and agriculture (productivity changes with changes in water supply, need for irrigation) and services from natural and semi-natural ecosystems. Changes in any of these compartments resulting from shifts in temperature and precipitation regimes are expected to result in feedbacks on water availability.
Achieving sustainable water use poses particular challenges for policy making because of its nature as a public good and because it often has both upstream/downstream and trans-boundary/transnational characteristics. Any changes in climate affecting precipitation and the behavior of snow and, where relevant, glaciers, will have a major influence on the seasonality, amount, and quality of surface runoff. The main changes are thus expected in the surface water systems, which have been comprehensively analyzed in the ACQWA project, in order to quantify the changes affecting the streamflow regimes, which may lead to uneven temporal distribution of the resource throughout the year. In this respect, the project has mainly concentrated on surface waters, both because mountain regions are in general depending on them rather than on groundwater, and given that the configuration of topography and landscape that do not allow the presence of sizeable alluvial deposits necessary for groundwater systems to develop.
A changing climatic regime could alter the frequency and the magnitude of a wide range of geomorphic processes related to extreme precipitation events that could in particular increase the severity of floods and debris flows (e.g. Stoffel and Beniston, 2006). Extreme precipitation events would in addition contribute to larger rates of erosion, discharge and sedimentation. A further factor responsible for decreased slope stability in a warmer climate is the reduced cohesion of the soil through permafrost degradation, particularly in the higher elevations in the Alps. Deglaciation can in some instances lead to problems of water accumulation behind unstable moraines that, if they fail, result in intense flooding and debris flows referred to as glacier-lake outburst floods (GLOFs) that put communities and infrastructure at risk. Similarly, hanging glaciers also pose a threat that is taken seriously in zones where glaciers are located above villages and major communication routes (Funk, 2006). In their retreat, these glaciers reveal a large quantity of unstable rubble and, sometimes ice that could result in severe down-slope flow of material. Given changes in the distribution of population with more extensive and intensive land use, particularly from tourism, any increase in the number and intensity of natural disasters could have proportionally higher human costs. By taking into account the impacts of extreme events within the hydrological cycle, the ACQWA project has acknowledged the fact that resulting natural hazards could contribute to the disruption of access to, and use of, water for many economic purposes.
Shifting precipitation patterns by season and sharply curtailed glacier mass in the mountains will lead to modifications in hydrological regimes and will also mean glaciers will no longer feed water into river catchments at a time of the year when precipitation amounts are low and the snow-pack has completely melted. These changes will have significant impacts on several key socio-economic sectors in mountain regions, particularly since these are also subjected to various other forces that influence their viability. There will in addition be cascading effects on downstream areas. Climatic changes will affect overall land use patterns, which in turn feed back into effects on water and carbon fluxes. Mountain agriculture has been under pressure from lower-cost production in lowland areas. Potential increases in drought conditions will only serve to increase its vulnerability. Forests’ crucial role in protecting against erosion and protecting biodiversity and water storage are potentially threatened. As a result of shifting seasonality of precipitation and glacier melt, the reduction in capacity to store water could also diminish the potential for hydroelectric production just as European-wide efforts are being initiated to reduce dependence on fossil-fuel based energy sources. Not only mountain regions, but the entire European electrical grid could thus be affected. Some rivers, such as those that flow from the northern part of the Alps, may dry up partially or completely towards the end of the summer; this is already the case in the Mediterranean mountains, where the drought situation may well worsen. Although the energy potential of the Alps has by now largely been exploited, this is not the case for other regions of the world where this usage has barely been tapped (Romerio, 2002). By investigating the challenges of managing crucial but often limited water resources in many of the mountain regions that have been studied, the ACQWA project has aimed to better understand social and economic issues such as changes in social arrangements among mountain populations and their downstream neighbours, energy production in regions where water is underused as a means of helping abate greenhouse-gas emissions, and environmental issues such as the future evolution of water supply for use in domestic, tourist, or agricultural sectors.
Climatic change leading to shifts in hydrological regimes has the potential to increase competition over water that will be available at different times and in different quantities. Water is difficult to allocate because of its public good features, which are aggravated by upstream potential to capture the resource and by the fact that flowing water may cross internal and international borders. Changing land cover and land use will generate significant shifts in the amount and seasonality of water resources. For example, deforestation causes an increase in the average annual discharge, and an acceleration of runoff during rainstorm events, also enhancing erosion and downstream sediment supply. Changing social patterns and economic incentives have resulted in major land-use changes in many mountains of the world and, in some instances, have exacerbated the risks associated with excessive runoff and erosion potential.
The AQCWA project has sought to address such issues, particularly in regions with significant social change, for example resulting in land abandonment as farming loses attractiveness and spontaneous vegetation colonizes previously-managed terrain. Conflicting water use, for example between agriculture and hydropower, or between hydropower and tourism, as the resource diminishes through reduced precipitation in some areas and glacier retreat in others, has also been investigated within the ACQWA context (Hill et al., 2014). This has been achieved to acknowledge the fact that new water resource management is not just a matter of adjusting to shifts in the physical environment but is also associated with social changes generated by changing types and levels of water use and new market conditions affecting the distribution of the resource. These different issues have been tackled in the ACQWA project through a number of work packages (WPs), namely:
• WP1: Coordination and administration
• WP2: Climatic and socio-economic drivers of change
This WP provided a quantitative description of the primary (or direct) drivers such as climate change, and of the indirect socio-economic drivers, with relevance to mountain water resources.
• WP3: Modelling climate change impacts on water resources, including extremes.
The focus of this WP was to develop the climate scenarios at the regional and local scale and to model the effects of climatic change on water resources at the basin scale, using state-of-the-art cryosphere, biosphere, and hydrological models.
• WP4: Impacts on natural and socio-economic systems, adaptation strategies, and policy issues
The impacts, adaptation and policy WP investigated the manner in which changing water resources and water use may influence a range of sectors. It was also designed to consider a portfolio of adaptation and other response strategies, and to revisit current water governance issues with a view of improving water resource management.
• WP5: Dissemination, outreach and training
This WP was dedicated to project-specific workshop activities, publications, exchange of young scientists within the network, and public and stakeholder outreach activities.
Project Results:
The ACQWA S&T description will focus on the principal results emerging from the project, pertaining to the following elements:
• The ACQWA case-study regions
• Regional climate change in the ACQWA-case study areas
• Climate impacts by the middle of the 21st century on natural environmental systems that determine water availability: snow and ice; hydrology; extreme events
• Impacts on socio-economic systems (hydropower, agriculture, tourism) and semi-managed systems (forests, aquatic ecosystems)
• Lessons to be learned from the non-European case-study areas
A more complete description of the scientific results, including policy-relevance and possible adaptation strategies, is available in a 100-page ACQWA Science and Policy Brief (Beniston et al., 2013), and a shorter focus is provided in the introduction to the Special Issue of Science of the Total Environment dedicated to the project by Beniston and Stoffel (2014).
Case study regions used in the ACQWA project
Fig. 2 shows the main case-study areas investigated in the context of the ACQWA project. The Rhone and Po river basins in the European Alps have been used as a common “test ground” for model investigations, where the different methodological approaches have converged to the basin scale through appropriate up- or down-scaling techniques. Both basins represent ideal case-study areas, as they comprise all the elements of the natural environment that have been modeled (snow, ice, vegetation, hydrology) and have a wealth of data to enable models to be validated. At the same time, these are highly regulated watersheds, where economic activities related to hydropower generation, irrigated agriculture, and tourism take place in the context of a climate that is at the borderline between Mediterranean and Continental, and are therefore particularly vulnerable to climatic change (Beniston, 2003). The boundaries of the Rhone catchment study-area include the alpine segment, running from the Rhone Glacier in Central Switzerland to Lake Geneva. The boundaries of the Po case-study area used in the ACQWA project do not extend as far as the Adriatic Sea, for reasons of data access and hydrological model constraints. The investigations have thus focused more on the flows from the Alps of Piemonte and Val d’Aosta, with the “ACQWA Po” boundary that is limited to Cremona, on the Po River south of Milan and the western segment of the basin (Coppola et al., 2014). Regional climate model results, however, cover the entire basin as illustrated in the map on Fig. 1.
Figure 2: The principal ACQWA case-study regions: the Swiss Rhone and Italian Po basins, the Chilean Aconcagua basin, and the Kyrgyz catchments of Central Asia
Some of the methodologies developed in the intensive investigations of the European alpine catchments have been applied to the Aconcagua Basin in Chile, where receding glaciers already today pose a genuine threat to water availability (Pellicciotti et al., 2014). Investigating the coping strategies of Chilean economic sectors affected by changes in the quantity and seasonality of water resources can help highlight the types of problems that could arise in the Alps in coming decades (Hill et al, 2014). In Central Asia (Kyrgyzstan), on the other hand, the same processes of ice-mass wasting in the headwaters of the Syr Daria or Amu Daria rivers involve much larger glaciers (Sorg et al., 2012). During the 21st century, the meltwaters from the Tien Shan could potentially represent a source of economic opportunity, for example through the development of hydropower as a source of foreign revenue, but also a risk in view of the political instability and rivalries between different independent states of former USSR (Sorg et al., 2014).
Other research-specific case-study areas, not shown in Fig. 2 comprise the Aragón Basin in Spain for interdisciplinary investigations pertaining to agriculture and energy in a context of changing land-use and climate (Lopez-Moreno et al., 2014); and French Pyrenean watersheds for aquatic ecosystem studies in a hydrology, habitat and biota framework (Khamis et al., 2014a). These are located in the Cauterets region in the vicinity of the French Pyrenees National Park. By analogy with the other non-Alpine case study regions, some of the issues addressed in the Pyrenees in today’s world are likely to be those that will arise in the European Alps in tomorrow’s world.
Regional climate change in the ACQWA-case study areas
The high-resolution simulations carried out within the EU-FP6 ENSEMBLES project (www.ensembles-eu.org) formed the basis for the focused modelling work and climate impacts assessments within ACQWA. Two principal simulations were chosen from the ENSEMBLES multi-model dataset (namely ICTP_RegCM and MPI_REMO, both driven by ECHAM5-r3) and used by all partners. In addition, several impacts studies used the entire ENSEMBLES dataset in order to identify more completely climate-induced uncertainties. The IPCC (2001) A1B greenhouse-gas scenario through to the mid-21st century was applied across all the individual case studies to have a common scenario referenced period for all projections and impacts studies. The 2050s was set as the principal time horizon for the project, as this is a period in the future which is not too far removed from the time-scales typical of forward economic planning and decision-making. Using the ENSEMBLES simulations from a set of 22 high resolution regional climate models (RCMs), climate data in the regions of interest was compiled up to the year 2050. The ENSEMBLES RCMs were used with a horizontal grid spacing of 25 km, and the lateral boundary conditions were provided by eight different global climate models (GCMs). The restriction of using only the A1B emission scenario, rather than other scenarios or a range of emission futures, is of minor importance since the uncertainty due to the choice of emission scenario remains fairly small in the first half of the 21st century (Gobiet et al., 2014).
Results by 2050 using the multi-model mean climate change signals exhibit stronger warming along the Alpine ridge, especially in summer. The high sensitivity of the Alps becomes even more evident in the rather small Rhone case-study region located in the Valais region of south-central Switzerland, with projected median warming over 1.5°C in winter (DJF) and close to 2°C in summer (JJA). Warming is projected by all models of the dataset and for all seasons; the uncertainty of the projected changes is larger in summer and autumn than in winter and spring (Gobiet et al., 2014). In-depth analyses of this data suggests that the choice of the GCM that drives the RCM initial and boundary conditions has by far the largest effect on the total uncertainty, contributing more than 75% to the overall variance in most cases (Im et al., 2010).
Figure 3: Alpine-scale precipitation change in winter and summer by 2050.
(Source: A. Gobiet, University of Graz, Austria).
The massive presence of the Alpine ridge as a dividing feature between Mediterranean and Atlantic or Continental climates clearly influences the spatial distribution of precipitation and the projections of change, as seen in Fig. 3. In the Rhone catchment, the summer decrease and winter increase are small through to 2050, although these changes may well amplify into the second half of the 21st century as shown by a number of earlier studies (e.g. Beniston, 2006). In the Po catchment, temperature and precipitation changes are somewhat more marked than in the Rhone catchment to the north. Between the decades 2001-2010 and 2041-2050, increases of temperature according to different RCM simulations range from 2 - 3°C, and the variation of mean annual precipitation ranges from 1-10%, mainly in the winter and early spring period. Accelerated melting periods, earlier in the year, and likely increases in summertime evapotranspiration will inevitably counter the influence of the larger amounts of summer precipitation on river discharge that are projected for the region.
In the Andean zone of Central Chile, the Aconcagua catchment is projected to experience warmer winters and decreasing precipitation which, as in other mountain regions, will affect the behavior of the mountain snowpack and lead to changes in the timing of snow and glacier melt (Pelliciotti et al., 2014).
In the Central Asian republic of Kyrgyzstan, available climate simulations project by 2050 decreases in summer precipitation by around 5% and increases in winter precipitation around 8%. Temperature increases of between 2.5 and 4.5°C are projected for all seasons in the region. Overall, extreme events will tend to increase, in particular at both ends of the moisture spectrum with more summer droughts and winter or spring flood events (Sorg et al., 2012).
Snow and glacier response to climatic change
• Remote-sensing of changes in snow cover
Remote sensing provides a unique opportunity to address the question of snow cover regime changes at regional scales. Since the availability of daily optical satellite data at the end of the 1980s (NOAA-AVHRR), methods have been developed to compute changes in the surface area of snow cover (SCA) and snow cover duration (SCD). The main parameters analyzed are the timing and duration of the melting season under current and future climate conditions. In this context, a remote sensing database of snow cover dynamics over a time period of 10 hydrological years (2000-2010) was compiled (Dedieu et al., 2014). It focused on the four watersheds of the ACQWA project, namely the upper Rhone (5’300 km2) in Switzerland, the upper Po (37’800 km2) in Italy, the Aconcagua in Chile (5’800 km2) and the Syr Darya source region in Kyrgyzstan (110’000 km2). The satellite data were provided by the MODIS Terra MOD-09 images and the MOD-10 snow products (NSIDC).
Figure 4: Annual standard deviation of snow cover in the Po Basin
The results rely on previous studies already conducted in different regions to reconstruct time series of snow cover at the regional scale and to analyze snow regime trends under the current climate change context. The specific added values of work within ACQWA include a novel and original method for climate impact detection, as well as scientific results for regions with crucial lack of information (Kyrgyzstan, Chile). Significant progress has been made in using statistical tools to assess the interannual variability of snow cover. Maps of standard deviation highlight sensitive regions where strong temporal and spatial variability becomes a significant proxy of climate change related to recent changes in temperature, as shown for example in Fig. 4.
• Changes in snow cover in the alpine part of the Rhone catchment
Numerical simulations to assess the future course of seasonal snow cover under the influence of climate change were performed during the ACWQA project using the detailed snowpack model CROCUS coupled to the land surface model ISBA. In contrast to snowpack components of most land surface schemes within GCMs/RCMs and hydrological models, CROCUS explicitly accounts for internal processes occurring within the snowpack such as compaction, phase change/refreezing and snow metamorphism. During the ACQWA project, the model was driven by atmospheric fields from the RCM REMO at a horizontal resolution of 10x10 km² and run for a century, from 1950 to 2050. Regardless of the metric used to characterize the time evolution of snow conditions, the main outcome from the ACQWA-based study is that the seasonal snow cover in Alpine regions is most likely to decline within the next decades (Fig. 5), essentially as a result of increases in temperature (e.g. Beniston, 2012). This leads to a shift in the snow/rain partitioning towards relatively more rain and less snow precipitation. Such changes are seen to occur particularly at mid-altitude locations, between 1000 and 2000 m elevation above sea-level, which are most sensitive to air temperature fluctuations around the freezing point. The results obtained using the detailed snowpack model Crocus are consistent with the conclusions drawn from other impact models that use simpler snow schemes.
Figure 5: Upper left map mean snow water equivalent (SWEmax; mm); right: altitudinal profile of SWEmax for 1950-1969 period. Lower left: map of changes in SWEmax changes (mm) between 1950-1969 and 2030-2049;right: changes in altitudinal profiles. Dots in the right-hand figures refer to individual simulation grid points, whose average value over a given altitude band is provided as bars.
• Glacier changes in the Rhone and Po catchments
Understanding the glacier response to climate change in mountain regions is extremely relevant for water resources management. Future glacier response is also important to identify the boundary conditions for the evolution of the glacier hazard potential. Climate modifications can indeed influence stream flow regimes, increase downstream landslide and flood risk, and have an impact on hydropower production and other water uses, which strongly depend on melt water. The modelling of glacier response to climate change was carried out at the glacier scale by means of state-of-the-art continuous mass balance models for six different glaciers characterized by different morphological and surface characteristics and representative of the greater Alpine region. The models developed and implemented within the ACQWA project accounted for, in a spatially distributed fashion and at sub-daily temporal scales, accumulation and ablation processes as well as glacier evolution (Finger et al., 2012). The results thus allowed to quantify how glaciers tend to modify in response to changes in climate in order to reach the equilibrium with the current climate. Changes of the mass budget of a glacier and of glacier geometry (surface area, length and volume), were modeled at the highest level of detail in relation to knowledge of glacier behaviour and to the need of long-term modelling under stochastic climate scenario forcing.
The predictions of ice volume evolution suggest a progressive glacier retreat for the period 2001-2050 and a related ice volume reduction (see Fig. 6 for the Rhone Glacier, source of the Rhone River, in south-central Switzerland). Regardless of glacier features and future meteorological conditions, at no time in the coming decades are stationary conditions for ice volume attained, for any of the glaciers investigated. Ablation is generally sufficient to melt the entire amount of snow accumulated during the winter season over the remaining glacierized area and does not lead to changes in the mass balance trend.
Figure 6: 3-D representation of the extent of the Rhone Glacier between 2010 and 2050
(Source: P. Burlando, ETH-Zurich, Switzerland).
An interesting result concerns the behaviour of glaciers and/or glacier areas, which are covered by a thick layer of debris. Their response to climate change shows a slower negative mass balance in comparison to debris-free glaciers. The debris cover tends to insulate the ice beneath, thus preserving the ice. The thin ice thickness located at high elevation and directly in contact with the atmosphere tends to disappear, leading to a scenario in which most of the remaining glacier surface is located at low elevations and is not exposed, but covered by a mantle of rock debris. This result points to the importance of including debris-covered areas in realistic simulations of future evolution of glaciers.
No significant variations in the mean spatial snow melt rates over the simulated decades were observed. However, the variability introduced by spatial stochastic climate forcing allows for spatial changes to occur. Some simulations showed that in some decades high snow melt rates at low elevations could compensate for low rates at higher elevations; in addition, higher snow melt rates were obtained at high elevation, due to increased temperatures in the future, whereas no snow melt occurred at low elevations as a result of the earlier disappearance of snow and the higher fraction of liquid precipitation events. Snow melt contribution to the runoff hydrograph, as well as its variability, thus reflect the characteristics of the meteorological forcing, indicating the importance of accounting for climate scenarios that are localized, spatially variable and highly resolved in time (Fatichi et al., 2014).
A progressive shift and change in shape of the ice melt rate frequency distribution was identified across all melting seasons. This is due to new melt events during times of the day that in the current and past climate did not generally lead to melting. Thus, anticipated ice exposure in spring generates low melt rates, as temperature is not high enough to cause higher melt rates. In summer, more frequent melt occurrence at night is related to higher night temperatures or when the solar radiation contribution is still low, such as during early mornings and late afternoons. Low melt increase is also observed in autumn in some decades and can likely be associated to melt at night. Higher frequency of high ice melt rates was also observed in the last two decades investigated (2031 to 2050) both in spring and summer, due to significantly higher temperature and reduced elevation range of the reduced area of glacier cover. In general, the overall ice melt rate is more variable than snow melt due to the transition of most of the investigated glaciers to small scale glaciers. In this respect, the timing of the transition and, more generally the impact of climate forcing, is significantly influenced by the glacier morphology, the initial ice thickness distributed over a large elevation range and by the intensity of the local climate change signal, which can be different from the large scale patterns simulated by GCMs and RCMs. The ice melt contribution to runoff in glacierized catchments tends accordingly to disappear gradually. An exception is represented by the Aletsch Glacier, the largest alpine glacier situated in south-central Switzerland, where no large changes for the simulated decades (2001 to 2050) are observed in the ice melt hydrograph.
The gradual disappearance of ice volume induced by climate change may confine glaciers to higher elevations causing a possible reduction of the speed of glacier retreat. A key role in this respect is played by the winter accumulated precipitation, which can mitigate (high accumulation) or enhance (low accumulation) the glacier retreat and the backward shift of the accumulation area.
• Stability of glaciers
Three different types of instabilities can be identified according to the thermal properties of the interface between the ice and the bedrock (Failletaz et al., 2012). If cold, the maturation of the rupture is associated with changes in surface velocities and the seismic activity generated by the glacier which, if known, can have predictive value. For the other types of instabilities, water plays a key role in the initiation and the development of an instability leading to rupture. If the ice/bed interface is partly temperate, the presence of melt-water at the interface reduces its basal resistance, which enhances the instability but also renders its prediction difficult. The third type of instability concerns steep temperate glacier tongues which experience enhanced basal motion during the summer melt period. Although instabilities of this nature are still difficult to forecast, a novel numerical model that includes water flow in a sub-glacial drainage network has been developed that has predictive capability. In the context of climate change, the stability of some alpine glaciers may be affected in the near future due to changes in the thermal regime at the ice/bedrock interface. Although some presently-hazardous glaciers may present a lower risk in the near future because of their retreat, some others may evolve towards a critical situation and present a genuine hazard for communities and infrastructure lower down in the valleys. A timely identification of such transitions towards potentially critical hazards is today a challenge that the ACQWA project has contributed to improving.
The modelling techniques developed and used in the ACQWA project represent a step forward with respect to existing literature as they allowed to highlight some important aspects of the impact of climate change on future glacier evolution (Failletaz et al., 2012). The key elements of the added value include: (i) the identification of the important role played by distributed in space and highly resolved in time localized scenarios, which explicitly model the internal variability of future climate, in modulating the glacier evolution trajectory jointly with the influence of glacier morphology, altitudinal range and ice thickness; (ii) a quantification of the evolution trajectories of different representative glaciers by means of statistical descriptors of ice- and snow-melt changes following ensemble based simulations by means of advanced mass balance glacier models; (iii) a decomposition of the mechanisms by which glacierized catchments will change their runoff regime, also showing relative changes of the runoff components and how these depend on the local climate and the glacier configuration (size, aspect, shape, etc.); (iv) the quantification of the role of debris cover in modifying the glacier retreat pattern in space and time, thus highlighting the importance of including this component in glacier evolution assessments for more realistic predictions.
Hydrological response to climatic change
• Distributed basin-scale responses to climate scenarios (Rhone Catchment)
The physically based distributed modelling methodology, as one approach to assessing changes in hydrology, allowed to model the response of the (upper) Rhone river to a high number of stochastically downscaled ensemble members at hourly temporal resolution and with a spatial discretization of 250x250 m, also keeping the highest detail of representation of anthropogenic disturbances, such as hydropower, water abstraction and irrigation. This level of detail is unprecedented, particularly for studies in Alpine regions, and represents a new basis, as compared to existing studies, for investigating options for adaptation policies. The results show how climate change effects on stream flow propagate from high elevation headwater catchments to the river in the major valley, highlighting the damping effect of the river network on the mean and on the extremes, even when the latter show large increases in the upper river reaches. The simulations also indicate that changes in the natural hydrological regime imposed by the existing hydraulic infrastructure are likely larger than climate change signals expected by the middle of the 21th century in most of the river network (Fatichi et al., 2014).
Results suggest that internal (stochastic) climate variability is a fundamental source of uncertainty, typically larger than the projected climate change signal. Therefore, climate change effects in stream flow mean, frequency and seasonality are masked by possible natural climatic fluctuations in large part of the analyzed regions (Fatichi et al., 2014). Simulations also identify regions where strong precipitation increase in the February to April period leads to flow larger than natural climate variability during the melting season. Despite the strong uncertainties induced by stochastic climate variability, an elevation dependence of climate change impacts on stream flow could be identified, with a severe reduction due to the missing contribution of water from ice melt at high-elevation and a damped effect downstream. The presence of reservoirs and river diversions tends to decrease the uncertainty in future stream-flow predictions that are conversely very large for highly glacierized catchments. Despite uncertainty, reduced ice cover and ice melt are likely to have significant implication for aquatic biodiversity and hydropower production. A decrease of August-September discharge and an increase of hourly-daily maximum flows appear as the most robust projected changes for the different parts of the Rhone catchment. This will have a significant impact on hydropower production and management. Water abstraction and irrigation needs, as described by the available data, will likely be marginally affected, as they represent a very small fraction of the use of the available water resources. However, while local changes may be of some relevance, it is unlikely that major changes in total runoff for the entire upper Rhone basin will occur in the decades up to 2050.
• Distributed basin-scale responses to climate scenarios (Po Basin)
The most common approach to assess the hydrologic impacts of global climate change involves the use of climate models to simulate climatic effects of increasing atmospheric concentrations of greenhouse gases, and hydrological models to simulate water-related impacts of climate change. River discharges and their temporal distributions are strongly affected by high mountain areas that are particularly sensitive to global warming. Therefore the quality of hydrological impact investigations, even for larger catchments, depends on the capability to model those specific processes in mountains.
Comparison of parsimonious and fully physically based models showed that simpler models, despite their approach for computing evapotranspiration based only on temperature, are sufficiently robust and accurate to perform hydrological impact investigations of climate change for alpine river basins investigated in the ACQWA project (Ravazzani, 2013). The bias resulting from the approximation of the method implemented to compute evapotranspiration is lower than uncertainty associated with different climate models, however.
Impacts of climate change on hydrological processes were assessed by comparing hydrological model results for the upper Po river basin driven by two different regional climate models (REMO and RegCM3) for the decade 2041-2050 with respect to the decade 2001-2010. Increase of temperature ranges between 15.2 and 17.5%, while variation of mean annual precipitation ranges from 1.1 to 9.6%. Precipitation increase is mainly concentrated in the period from January through March, and causes an increase of snow water equivalent during this period, followed by an accelerated melting period in the mountains. The rise in temperature causes an increase of actual evapotranspiration, mainly in the summer period that counteracts the influence of the larger amounts of summer precipitation on river discharge. Impacts of climate change on flow duration curves of mountain tributaries of the Po River, as provided by different regional models, exhibit a general decrease of discharge for high durations (low flows) and an increase in discharge for low durations (high flows). The two climate models yield different results for the impact on flow duration curves in Po the Po river basin. In particular, the REMO model data results in an increase in discharge for both low and high durations, while use of the RegCM model data yields a decrease in discharge for both low and high durations except for the period ranging from 12 to 58 days.
• Modeling changes in the Po, Rhone and Kyrgyz catchments with the CHyM model
In order not to depend on one single model, the effects of predicted climate changes on Hydrological cycle of Po and Rhone catchments have been studied by forcing the CHyM hydrological distributed model with 8 different climate scenarios simulated by RegCM and REMO regional climate models. A further simulation with the RegCM model has also been used to simulate the future hydrological scenario on the Kyrgyzstan catchments. The future discharge trend has been analyzed for the whole drainage network, obtaining a map of such effects for different seasons (Coppola et al., 2014).
Downscaling of climatic scenario at hydrological scale can be considered an important challenge after the downscaling from GCMs to RCMs models. The results show that the simulated effects of climate change on the hydrological cycle appear to be considerably different for the Po and Rhone basins and also for different regions within the same basin. The same conclusions can be reached using the 8 different simulated climatic scenarios. The decrease of water resources is shown to be more critical for the entire flood plain during the fall season, leading to a loss of 200 m3/sec at the outlet of the Po River. The decrease of flow discharge is estimated to be more than 50% of the seasonal average for a large portion of the drainage network (Im et al. 2010). The effects of climate change on the hydrological cycle appear less evident in the high part of the Rhone valley and, more generally, in the higher part of the Alpine region (Fig. 7). The situation is quite different for Kyrgyzstan, where, for a large portion of simulated domain, increases in discharge during winter months and decreases of water availability during summer are observed.
Figure 7. 30-year mean discharge difference (%) in the Po basin (left) and Rhone basin (right).
Extreme events
Extremes of heat and moisture can have a strong bearing on the rate of snow melt, glacier retreat, and thus on water regimes in mountain watersheds (floods and droughts and associated impacts on a number of water-dependent economic sectors). Geomorphic hazards can add an extra burden on local economies by damaging buildings, disrupting communicating routes and potentially rivers themselves (silting, damming, etc.). A focus on extreme events and natural hazards was thus an important focus of the ACQWA project.
• Novel statistical methods to deal with extreme events
Weather and climate extremes affect societies and ecosystems and sometimes induce fatalities and large financial losses. To reduce the impact of such events and to improve risk assessment studies, it is essential to obtain accurate statistical features of extremes. In the framework of the Extreme Value Theory, a new method has been established, based on Generalised Probability Weighted Moments and Kernel regression (Naveau et al., 2011).
Novel statistical methods were proposed to analyze extreme events, to develop new algorithms to implement such methods and to apply those approaches to hydrological and climate data. Among many byproducts of this project, three specific topics were highlighted: spatial clustering of weekly maxima of hourly French rainfall, non-stationary analysis of heavy precipitation in Switzerland, and the study of global changes in seasonal extreme precipitation (Kallache et al., 2011). For the first topic, one of the main objectives of statistical climatology is to extract relevant information hidden in complex spatial-temporal climatological datasets. To identify spatial patterns, the most well-known statistical techniques are based on the concept of intra and inter clusters variances (like the k-means algorithm or EOF’s). As analyzing quantitatively extremes like heavy rainfall has become more and more prevalent for climatologists and hydrologists during the last decades, finding spatial patterns, simple and fast clustering tools tailored for extremes have been lacking. In this context, a novel algorithm based on multivariate extreme value theory was proposed. Comparing with classical clustering on weekly maxima of hourly precipitation recorded in France (Fall season, 92 stations, 1993-2011) shows that other patterns, specific to extremes events, were missing when employing traditional approaches.
The second example dealt with the inference of high rainfall return levels in a situation of non-stationarity in space and time. A new estimation technique of such high return levels based on semi-parametric approaches within the mathematical framework of Extreme Value Analysis was implemented (Smith et al., 2013).
The third example focused on exploring how anthropogenically-warmed climate is expected to sustain a larger increase of precipitation extremes. Here, new evidence was provided on global seasonal precipitation extremes for the 21st century, using 8 new high- resolution global climate model simulations. In the mid and high latitudes of both hemispheres, a significant intensification of extremes is evident in all seasons at the end of the century (Toreti et al., 2013).
The method provided an accurate estimation of heavy, extreme daily precipitation by statistically modeling exceedances above a high threshold and handling for spatio-temporal non-stationarities. The method is fast (no optimization is required) and flexible (non-parametric) and can be applied to large data sets from catchment area series to global climate models. Importantly, the method is computationally inexpensive (Naveau et al., 2011).
A similar approach was applied to high resolution CMIP5 Global Climate Models. Results show a remarkable intensification of extreme precipitation events at the end of the 21st century (especially under the RCP8.5 scenario) in all seasons. A parallelized R package of the method has been developed and is available for use by hydrologists, climatologists, engineers. This part of ACQWA research enabled the development of a state-of-the-art fast and flexible methodology that takes into account spatio-temporal non-stationarities in the statistical modeling of the daily precipitation exceedances distribution. The method can be also applied to large data sets. The new methodology allows in a computationally inexpensive approach to accurately characterize the statistical features of extreme precipitations, such as the return values illustrated in Fig. 8 for Switzerland (Stucki et al., 2012).
Figure 8. 50-year return levels of precipitation (mm) for Switzerland, 1961-2010. Spatial-temporal differences are evident with the north-south gradient and the tendency towards higher return levels in the 1980s as well as the recent decade over the Upper Rhone valley.
• Hot spells in the Rhone valley
In order to achieve a better understanding of the potential impacts of climate and climate change over the target areas, temperature extremes carefully and accurately studied. Here and for the first time, warm spells both in winter and summer were identified and analyzed for the last six decades (1951-2009) over the Upper and Southern Rhone valley by applying a recently developed approach (Toreti et al., 2012).
Figure 9:Intensity of winter and summer warm spells from 1951-2009 in the Upper Rhone Valley. Grey dots are associated with values of intensity calculated for each grid point. The black solid line represents the median of the values.
The approach is based on the identification of consecutive sequences of days (with a certain tolerance) with daily maximum temperature above a station-specified threshold. An increase in the intensity and frequency seems to have affected the two areas in the last 30 years in both seasons, especially after 1990 (Christidis et al., 2012). Furthermore, the higher intensity of the warm summer spell in the Southern Rhone valley, especially at the end of the period, can be highlighted. Finally, the 1983, 2003 and 2006 summer events are clearly visible in both valleys (Fig. 9).
The application of the state-of-the-art methods to recently released data sets gave the possibility to identify, estimate and analyze for the first time warm spells over the target areas both for winter and summer. The approach can be easily adapted for other regions of the world and it is available to the climate community (e.g. Hegerl et al., 2011).
• Hydrologically-relevant geomorphologic hazards
Changes in temperature and precipitation are likely to have a range of secondary effects on the occurrence of natural hazards, in particular also in mountain environments. However, while theoretical understanding exists for increased mass-movement activity as a consequence of predicted climate change, impacts can hardly be detected currently in observational records. One of the most obvious consequences of climate change at higher elevations is the glacier downwasting and related formation of ice-marginal lakes, ice avalanches and gravitational processes originating from the debuttressing of previously glacierized walls and hillslopes (Jomelli, 2012; Stoffel and Huggel, 2012). Glacier downwasting is likely to promote many rock slope failures at rather short future time scales, probably in the order of decades. Important effects of climate change on slope stability are also related to the warming and thawing of permafrost. Slopes currently underlain by degrading permafrost will probably become less stable at progressively higher altitudes with ongoing climate change. The probability of rock instability and the incidence of large (>106 m3) rockfalls will likely increase in a warming climate. A large number of recent slope failures have been documented in permafrost areas, related to increasing temperatures.
The impact of climate change on hydrologically-relevant geomorphologic hazards was further investigated by the development of (i) the slope stability component of the distributed hydrological model used for climate change impact analysis and (ii) an advanced slope stability model, which can simulate the effect of climate change on the occurrences of shallow landslides at high resolution in time and space.
The soil slip simulation tool that was implemented in the distributed hydrological model during this study is based on the concept of the infinite slope model (Tiranti et al., 2013). This model is aimed at exploring the ability to reproduce the slope response using a coarse terrain resolution at large catchment scales (river basin mesoscale, of the order of hundreds of km2) and thus for large scale simulations that are aimed at predicting hazard changes at the regional scale.
The second model was developed to investigate more in detail and at finer spatial scales the stability of slopes and their dependence on the detailed description of the involved hydrological and geotechnical processes. The HYDROlisthisis model is aimed at fine spatial scale resolution (of the order of the meter) and thus suitable for local scale (i.e. hill-slope and catchment scales of the order of 10 km2 at most) investigations of slope stability response to climate change forcings. The model consists of an hydrological and a geotechnical component, which are coupled together. The hydrological component takes into account the 3D variably saturated flow through soil, surface runoff, and hysteresis of the Soil Water Retention Curve (SWRC), topography-dependent solar radiation, potential evapotranspiration and root water uptake. The geotechnical component, which is based on a multidimensional limit equilibrium analysis, considers simple earth pressure conditions acting on the lateral sides of the soil column. These forces are computed using the coefficients of active, passive and at rest pressure, which are computed taking into account unsaturated conditions (Stoffel et al., 2014).
Both models showed, when applied to case studies, to be able to capture the observed slope instabilities at the scale for which they were developed. In particular, they showed significantly-better performance than statistical models, especially because they explicitly account for the spatial and temporal variability of the soil water dynamics, which is a key variable in the triggering of shallow landslides. This result is particularly valuable in the context of analyzing the shallow landslide hazard potential due to climate change, as they can explicitly account not only for the changes of the climatic forcing, but also for the consequences that this can have on the soil water dynamics and the related soil-vegetation medium response.
Climate and land use impacts on water availability and management: The case of the Aragón river basin, Pyrenees
Water availability is probably the main constraint for the development of modern agriculture, industry and tourism in the Ebro valley. The work developed within the ACQWA project provides robust information to water managers and policy makers concerning the response of river flows under projected environmental change. The study provides advance warning concerning limitations to maintain current water demand even if the regulation capacity is increased in headwater areas. Moreover, the study suggests that controlling land cover may be a mitigation strategy to minimize the probable reduction of available water resources (Lopez-Moreno et al., 2014).
A warming between 1-2°C and a decrease of precipitation between 5-20% is projected for the Spanish Pyrenees according to climate model simulations using the IPCC A1B scenario for the 2021-2050 period. A noticeable increase of warm events during winter months is also expected in the region. In the Pyrenees, an increase of 1ºC implies a 20% decrease in snow accumulation at 2000 m above sea level. This sensitivity increases at lower altitudes and decreases at upper elevations.
According to observed land cover change and the analysis of current aerial photography and remote sensing data, a scenario of land cover for the middle of the 21st century in the Pyrenees has been developed. Shrub areas are projected to evolve toward pine forests, with a 100 meters rise in the tree line and an increase of the shrub areas in the subalpine areas.
Increased vegetation in the basin could decrease annual stream flow in the Upper Aragón river basin by 16%, mainly in early spring, and autumn. Projected climate change could decrease annual stream flow by 13.8%, mainly in late spring and summer. Combined effects of forest regeneration and climate change may thus to reduce annual stream flows by 29.6%. Simulating the management of the main reservoir of the region using the modeled hydrological data, it is likely that serious difficulties to meet the current water demand, based on its current storage capacity (476 hm3) will emerge. If the current project to enlarge the reservoir to a capacity of 1059 hm3 is completed, the potential exists to apply multi-annual stream flow management, which will enhance the capacity to maintain the current water supply. However, under future climate and land cover scenarios, reservoir storage will rarely exceed half of the expected capacity, and the river flows downstream of the reservoir may be dramatically reduced (Lopez-Moreno et al, 2014).
Impacts on the hydropower sector
• The Swiss Rhone catchment
This research provided an analysis of the hydropower future in the context of climate change, opening of the electricity market to competition, decarbonisation of the energy system and, in some countries such as Switzerland and Germany, even phasing out of nuclear energy. The case study of the hydropower installation of Mattmark, located in the Rhone catchment represents the focal point of the analysis conducted in the context of ACQWA (Gaudard and Romerio, 2014). In addition, the study encompasses Swiss and European dynamics, in particular those related to energy policy and markets. The first part of the research, devoted to Mattmark, provides quantitative results, whereas the Swiss and European analyses are mainly qualitative.
The main interest of this study was to highlight the link between climate change, electricity markets and energy policy. In particular, it shows how electricity generation is affected by climate change, the opening of the electricity market to competition, as well as the development of micro and super-grids, new storage technologies and intermittent energy resources such as solar and wind energy. These factors also influence the added value created by hydropower, which represents an important source of revenue in mountain regions. The analysis takes into consideration the electricity demand, which is affected only modestly by climate change. The impact of its variation on the wholesale power market is estimated by means of econometric tools. Microeconomic models and techniques based on operational research are used to simulate the markets’ behaviour, hydropower reservoir management, as well as electricity generation.
Figure 9: Multiple constraints on electricity markets
In Fig. 10 (Gaudard et al., 2014), red boxes represent the drivers that are transforming the electricity system, which includes centralized and decentralized generation; storage services; consumption; supply and demand; flows of power through the electric lines (plain lines) and flows of information, notably price signals (dotted lines). Spot, future, balancing and ancillary services markets will determine the value of hydropower. Climate change will affect hydropower because of the possible reduction in surface water flows and seasonal shifts in water availability. The technological, economical and behavioural changes in the electricity system are, however, expected to exert a stronger impact on hydropower.
The outcome of the research conducted on the Mattmark site provides key information to decision-makers about the factors that will determine the future output and value of hydropower. Despite the margins of uncertainty, public bodies and private or public companies may use this information in the definition of their policies and strategies. The particular case of the renegotiation of the concessions for water rights, which represents a topical issue in several Alpine regions, is particularly relevant. In the case of Switzerland, most of the concessions will come to an end between 2030 and 2060. It should be mentioned here that, thanks to its flexibility, hydropower and reservoirs play a very important role in ensuring the security of electricity supply and the network’s stability in Europe.
• The Po Basin
Climate change can influence hydropower production in two ways: directly, through changes in precipitation and, as a consequence, in inflows; and indirectly through the electricity load because energy consumption varies with air temperature. This could be very important in determining the management of hydropower reservoirs and may cause conflicts with concurrent water uses.
In particular, reservoir management is aimed to provide the water resource when it is needed, transforming the natural regime, with its modulation across the year and its random fluctuations, in a regulated and more useful flow.
The main results from the ACQWA focus on the Po Basin have been summarized by Maran et al. (2014) as follows:
1. A large local variability in electricity production, which follows an analogous rainfall pattern: a reduction of 10% is estimated in the Val d’Aosta while a 20% increase is expected in the Toce valley;
2. The monthly modulation of power production throughout the year is expected to change: in both areas, the greater variability in summer inflows will affect the filling of hydropower reservoirs and, in some years, it will not be possible to use all the storage capacity.
3. the interannual variability of production is projected to increase.
All these effects are mostly the consequence of greater variability in river flows and the decrease in snow fall. The effects on hydropower production of variations in energy prices, i.e. the other major forcing factor, were also studied. The two Italian small case studies are very similar: in general, larger volumes of water are stored in winter and greater quantity of water are used in Spring, corresponding to a higher energy production in this period (Fig. 11).
Thanks to the collaborations of the ACQWA project, universities, research centers and private companies have had the possibility to deal with a high quantity of data produced by climatological and hydrological models at large spatiotemporal scales. These data allowed the implementation of detailed management models, simulating the management of hydropower plants. Large areas of north-western Italy were analyzed and an assessment of the impacts of climatic change on the water and electricity sectors in these zones has been made available.
Figure 11: Frequency distribution of yearly production for the Toce (left)
and the Valle d’Aosta (right) networks
However, new storage capacity implies the construction of new dams and reservoirs that could modify heavily the natural landscape of the Alpine region and are not easily accepted by local communities. This potential for conflict calls for the development and implementation of methodologies and processes to increase the adaptive capacity of water governance and management systems. ACQWA results could be used as an important aspect to be considered in decision making procedures regarding the development of new infrastructures for the water sector.
Impacts on the agricultural sector
• The Swiss Rhone catchment
In the Swiss Rhone catchment, irrigation has historically been important. Today, about 11’000 ha of land is irrigated with over half occupied by grasslands used for livestock production, followed by orchards and vineyards. This allocation of irrigation reflects the economic importance of the production of livestock and permanent crops, while arable crops play a minor role (Fuhrer and Jasper, 2012).
Over the past decades, a trend was identified towards a higher frequency of extreme droughts on timescales shorter than 2 months, whereas for longer timescales, no clear indications of a change over time could be found. For the future, the mean of several agro-climatic indices even suggest a shift towards warmer and rather wetter mean conditions during the growing seasons, but an increase in risks caused by high temperature (i.e. heat stress). The thermal growing season becomes longer, with potentially positive effects on pasture and livestock production, most pronounced at mountain sites, whereas at the valley bottom, a trend occurs towards increasing risks of frost in permanent crops due to an asynchronous change in the beginning of the growing season and late frost occurrence, and in heat stress for livestock (Fuhrer et al., 2014).
With increasing temperatures, water consumption through crop evapotranspiration increases, thus leading to additional irrigation needs to maintain optimal yields. This concerns much of the lower part of the valley, but also the pasture-dominated south-facing slopes, especially on soils with low water holding capacity. Simulations reveal a moderate average increase in water requirement for irrigation in 2021-2050 relative to 1981-2009. In the currently driest areas, additional potential water requirements from 1981-2009 to 2021-2049 would range between 0 and +200 mm per average growing season, depending on the climate scenario. During extremely dry years, such as 2003 or 2011, the increase would be much higher and could exceed the water availability in surface waters of smaller catchments with a nival runoff regime. An example would be the catchment of the Sionne (27.7 km2) where water is drawn through water channels (‘Suonen’) for grassland irrigation, and where discharge is controlled by snowmelt and rain. A much reduced area could be irrigated under these low-flow conditions, which may become more frequent in the future. At the catchment-scale, the estimated mean total water requirement is 32 × 106 m3 per year (1981–2009), and a 45% increase during the 2003 European heat wave in the driest area of the catchment. With an extreme scenario, the increase is about 44% by 2050, which is consistent with the values of 2003 as shown by Smith et al. (2012).
Coping with climate change in agriculture requires knowledge of possible trends in agro-climatic conditions, with a focus at the smaller scales such as catchments or sub-catchments where decisions are taken. At those scales, risks from water shortage during parts of the season are likely to become more important in the long-term future, particularly in view of a drastic decrease in glacier volume and snow cover. The methodology developed in this project can help to identify emerging water conflicts. Given the importance of grazed pastures in the mountain zones for their economic, ecological and cultural role, it can be expected that agricultural policy will continue to support the maintenance of this extensive form of agricultural production, which requires irrigation. In other parts of the valley, maintenance or even expansion of the production of high-value crops such as grapevines and fruit trees, or even vegetables, would continue to use water for irrigation. However, the amount is much smaller than for pastures, given the smaller surface area concerned, and the shorter length of the irrigation period. Moreover water resources at the valley bottom are much larger and may not be subject to the same extent to inter- and inner-annual variability on the timescale considered because the dependence on rain and snow is less than in sub-catchments with a nival regime. And, technical measures to optimize and control irrigation, or building reservoirs would be easier to implement.
Thus, in the shorter-term, the demand for water for irrigation will increase and put pressure on smaller rivers in catchments with little or no water supply from glaciers, particularly in years with limited precipitation and snowmelt during springtime. In the sub-catchments concerned, water management will play an even more essential role under future climate conditions than in the past. This improved water management should include both regulations regarding the allocation of water to different users of the same source, installation and management of reservoirs, and technical measures to improve the efficiency of irrigation by avoiding losses of distribution systems, evaporative losses, and excessive runoff due to over-application of water (Smith et al., 2012).
An important result of this assessment is that until 2050 major agro-climatic risks under climate change will likely be caused by high temperatures rather than by increased drought. Situations with high temperatures may have negative effects on both crop and livestock production. Measures to cope with this risk could include the use of additional water for cooling purposes, shifting crop cultivation windows, and shifts in cultivar selection in both arable crops and permanent crops (grapevines, fruit trees). In addition, extending the grazing period and shifting of grazing zones to higher and/or cooler parts of the catchment could help to avoid heat stress in animals. However, such shifts in intensification of grassland use may have negative consequences for other ecosystem services such as biodiversity, soil carbon storage, and nitrogen retention. Thus, coping with heat stress in livestock production will require careful consideration of the sensitivity of alpine grassland ecosystems to intensification (Fuhrer et al., 2014).
• The Po Basin
A basin-wide enquiry for the Po Basin has looked into the optimal policy that might facilitate adaptation to climate change in agriculture, in particular policies that can help to smooth water input fluctuations. The different roles of farmers and the public sector in adapting to these changes have been addressed. The first and most fundamental level of adaptation to climate change in agriculture occurs at the level of the local farmer. Farmers undertake strategies to adapt to the form of climate change that they are able to foresee, through observation of the recent trends in indicators such as average temperatures and average precipitation. However, they can do little to respond to the greater uncertainty inherent in climate change. Farmers’ adaptation to expected climate change will often take the form of investment in assets to shift water temporally, using locally appropriate water storage techniques (Bozzola and Swanson, 2014). It might also be possible for water management to be pursued through more efficient irrigation practices. Because of the low costs and relative abundance of the water resource in the Po basin, farmers have traditionally relied on inefficient irrigation methods, which are still one of the main causes of waste of fresh water resources. Local farmers adopt strategies to cope with expected climate change, but the important question for policy makers concerns the role of governance in supporting adaptation.
A key question here is the role that remains for policy in light of local adaptation. Evidence is given by this study that there are some impacts of climate change that farmers’ adaptations do not reach at present. Despite the farmers’ observed investments to adapt to mean changes in climate, variability in climate continues to impact crop yields. Table 1 shows the results of a simple correlation analysis of the relationship between variability in temperature, precipitation (during the months March to August) and agricultural yields in the Po basin.
Log
Yields Precipitation anomalies Temperature anomalies
Maize -0.16*** 0.02
Barley -0.28*** -0.15***
Soft Wheat -0.43*** -0.1*
Sugar Beet -0.21*** -0.2***
Soya -0.23*** 0.08
Table 1: Correlation between crops yields and standard deviation of precipitation and temperature. Significance Levels of the correlation coefficients at: ***1%; **5%; *10%.
The negative correlation between agricultural yields and observed anomalies indicates that agricultural production has suffered when unanticipated climatic variability occurs. In short, farmers’ adaptations appear to be based on individual (farmer-based) forecasts deriving from current trends in observed weather patterns, but are not responding to the increasing variabilities (Bozzola and Swanson, 2014).
Impacts on mountain tourism: the case of Valais, Switzerland
This part of the ACQWA project has investigated the climate-related and socio-economic drivers of winter tourism in the Swiss canton of Valais, as well as its expected water consumption in the future. Winter tourism is amongst the main economic sectors within the canton. The most significant finding is the industry’s exposure to reduced snow cover. A report by the OECD published in 2007 depicted Valais as showing little vulnerability to the reduction of snow cover. However, instead of approaching the tourism industry at the cantonal level, a more regional/local approach of the canton (48 resorts grouped into 17 regions) has been adopted. By comparing two average winters (2004-06) with a snow-poor winter with average economic conditions (2006-07) and a snow-rich winter at the beginning of the current economic crisis (2008-09), the study reveals high disparities between the resorts. This is an indication that vulnerability assessments and, more particularly, adaptation measures should be considered at a much lower level than has been done so far (Schaub and Andonova, 2012).
Valais seems to be more exposed or vulnerable to climate change than previously expected. Although the canton witnessed a smaller reduction in skier days than other regions during snow-poor winters, some resorts were strongly affected. And just in removing the biggest resort of the canton (Zermatt) from the statistics, the impact of the exceptionally warm winter of 2005-06 appears to be much bigger: Instead of losing only 4.9% in the number of skiers, the canton would have lost 7.8%. If the hotel sector proved less exposed to a snow-poor winter than the cable-car companies, the accommodation sector is facing serious difficulties in some regions when longer trends and socio-economic factors are taken into account. The canton could consider some support to the hotel sector so that they reach a “critical mass” of tourists, a strategy adopted by Austria, for instance.
The study can be seen partly as a “wake-up call” in many respects. The main political implications of these findings are that policy-makers need to take into account the big disparity existing between regions, and how vulnerable small and medium size resorts can be to the changes that are expected to impact tourism in years to come. The main challenges are to find ways for increasing cooperation amongst and within resorts and touristic regions, improve the promotion of the canton in general (the label “Valais Excellence” and the website of “Valais Tourism” constitute a good start in this respect), and more vitally, the ability to consider alternatives to skiing during warmer winters, especially for more vulnerable regions
.
For the regions themselves, it would be advantageous to build further winter-sport infrastructure exclusively in snow-reliable regions (thus increasing their center of gravity) and to adopt a more integrated structure. The target – the type and origin of population they would like to attract for winter holidays – should be chosen carefully. More generally, stakeholders should be aware of potential changes in the tourists’ preferences –a major unknown for the next decades. The future water consumption of the tourism industry can only be estimated. It can be said that the current coordination between the hydroelectric sector and the cable-car companies – in which the first sells water to the second for artificial snow-making purposes – should be maintained and perhaps even increased in some cases. More studies should be led to determine whether the current law on artificial snow in the canton should be improved in the future. In addition, the type of tourists (according to: age, country of origin, and the social class) they expect to attract should be considered. The preference of a high social class can change within a small time frame, and it is uncertain whether the middle class would continue its skiing activities at the same level as it does currently, or even if it can afford this time of recreation in the future.
Impacts on mountain forests
Climate change has the potential to substantially alter the provisioning of essential ecosystem services, and mountain regions are likely to be both particularly vulnerable and heterogeneous in their response to climate change. To date, few studies have attempted to quantify the impact on mountain ecosystems under a wide range of climate scenarios, including novel “2° scenarios” that are based on the assumption of an early stabilization of greenhouse gas concentrations in the atmosphere. Similarly, although mountain ecosystem services are crucial both locally (e.g. forest-mediated protection of human infrastructure from avalanches or rockfalls) and regionally (e.g. impact of ecosystems on runoff generation), there is a scarcity of studies focusing on shifts in ecosystem services.
A systematic assessment of climate change impacts on mountain forest properties and the ecosystem services they provide was performed in the context of the ACQWA project (Wolf et al., 2012). The focus was on five ecosystem services (carbon storage, runoff, timber production, diversity, and protection from natural hazards); four regionally downscaled climate scenarios that cover a wide range of possible future conditions were used; three complementary, state-of-the art models of forest dynamics that were used to simulate a wide range of forest properties were employed; and the simulations in two climatically contrasting case study areas (catchments) of the European Alps at several spatial scales, from the stand to the entire catchment, were performed.
Figure 12: Changes in forest-derived avalanche protection in the Saas valley (part of the Rhone catchment) under four climate scenarios as projected using the forest landscape model LandClim. Changes are shown as normalized difference compared to 2010 levels.
The results suggest that the sensitivity of mountain forest ecosystem services to a 2 °C warmer world depends heavily on the current climatic conditions of a region, the strong elevation gradients within a region, and the specific ecosystem services in question (Fig. 12). Model projections show that large negative impacts will occur at low and intermediate elevations in dryer/warmer regions. Here, relatively small climatic shifts result in negative drought-related impacts on forest ecosystem services (Manusch et al., 2013).
In contrast, at higher elevations, and in regions that are initially cool-wet, our simulations suggest that forest ecosystem services will be comparatively resistant to a 2 °C warmer world. It was furthermore found that considerable variation exists in the vulnerability of forest ecosystem services to climate change, with some services such as protection against rock-fall and avalanches being sensitive to 2 °C global climate change, but other services such as carbon storage being reasonably resistant. While these results indicate a heterogeneous response of mountain forest ecosystem services to climate change, the projected substantial reduction of some forest ecosystem services in dry regions suggests that even a 2 °C increase of global mean temperature cannot be seen as a universally “safe” boundary for the maintenance of mountain forest ecosystem services (Elkin et al., 2013)
The results achieved in the ACQWA context rely on research activities that were begun more than a decade ago and benefited considerably from the funding made available in ACQWA in the sense of a synthesis and further development of existing approaches. Beyond this, the added value of the project was the availability of state-of-the-art climate scenarios and the comprehensive focus of the study, as ACQWA has always emphasized cross-sectoral activities rather than isolated impact assessments of individual components of mountain systems.
Impacts on mountain aquatic ecosystems
• Mountain lakes in the Gran Paradiso National Park, Valle d’Aosta, Italy
During the ACQWA project, measurements of the physical, chemical and biological parameters in 20 high-altitude, ultra-oligotrophic lakes in the Gran Paradiso National Park (PNGP) were undertaken. All these lakes are characterized by extreme seasonality; most are naturally fishless, while some were stocked with fish about forty years ago (Tiberti et al., 2012). Both fishless and stocked lakes were sampled in order to determine the effects of introduced fish and of the interplay between the introduction of alien species, strong seasonality and shifts in environmental conditions. Among the different measurements, zooplankton densities at different depths in order to gain insight on their vertical distribution, their migratory behavior along the water column and the response of zooplankton to vertical abiotic gradients (e.g. light intensity and UV gradients). The PNGP measurements were complemented by the development of deterministic one-layer models to simulate the dynamics of high-altitude lake ecosystems, in order to estimate the response of the lake ecosystems to climate and environmental change (temperature, duration of ice cover, etc). These observational and modeling studies have enabled some of the first quantitative determinations of the effects of introduced fish and of the interplay between stocking and environmental change (Tiberti et al., 2013).
• Impacts on Pyrenean aquatic ecosystems
In mountainous river basins, the water source balance (i.e. the contribution of rainfall, glacier-melt, snowmelt, and groundwater to flow) is extremely sensitive to climate change. Projected warming is likely to alter stream flow quantity/quality regimes and modify in-stream physico-chemical habitat. As a result, biodiversity patterns of alpine running water organisms are highly likely to be altered, as vulnerable species must either adapt physiologically and/or genetically or migrate to more suitable habitats to persist (Finn et al., 2013).
As part of the ACQWA project an inter-disciplinary approach was adopted to investigate alpine stream ecosystem responses to climate change. Both the drivers of change (i.e. climate) and responses to change (i.e. hydrology – habitat – biota) were quantified using a combination of scenario simulation, space for time substitution surveys and in-situ experimental work. Macro-invertebrate community structure and function, population genetics, habitat characteristics and water source contributions were recorded at 26 sites representing a gradient of glacial influence, across five river basins in the French Pyrenees. Contemporary river flow, habitat, taxonomic and genetic patterns were related to both glacier cover and melt-water contribution to bulk discharge. Downscaled climate data was then used to drive (i) a watershed simulation (TOPKAPI) of river flow and glacier/snow melt dynamics and (ii) a water temperature regression model. Temperature and hydrological projections were then linked to observed patterns to predict future hydro-ecological change (Khamis et al., 2013a).
Further work, to improve understanding of finer scale hydrological and biological process , investigated; (i) the response of an alpine spring community to the introduction of a large bodied invertebrate predator (predicted to expand its range) and (ii) the inter-annual variability of the heat budget and thermal dynamics of a glacier-fed river reach.
Until now application and testing of the water source contribution approach as a hydro-ecological management tool, has been limited to single river basins. Results from this study highlight the utility of this approach for both management and prediction of climate driven ecosystem change. Furthermore, knowledge transfer between project partners enabled the successful completion of watershed simulations for a glacierized river basin in the relatively data poor Pyrenees, where until now few simulations have been conducted. This provided a unique insight into future hydrological and physico-chemical habitat change and the potential implications for biodiversity. Quantification of melt-water contribution enabled the identification of climate sensitive macro-invertebrate taxa which are expected to exhibit considerable range contraction as glacier retreat (Khamis et al., 2013b). Trait profiles of glacial stream taxa were also recorded and were shown to have the potential to act as indicators of changing water source dynamics (i.e. reduced glacier contribution to discharge), particularly when comparing sites from multiple biogeographic regions. Improved process understanding from experimental work and high resolution hydro-meteorological observations could prove valuable for further development of deterministic temperature models and species distribution models.
The Pyrenees represent the southern limit of contemporary glaciation in Europe, and only small cirque glaciers remain of the extensive ice cover which existed during the Little Ice Age. Thus, findings here can be viewed as a future analogue for mountain ranges which currently have significant glacier cover but which will shrink in the future. The successful application of the water source approach in a changing cryosphere across multiple river basins highlights its potential for use in future alpine conservation planning. Furthermore, the coupling of hydrological simulation and space for time surveys has a potential to inform future baseline conditions and could prove valuable for determining appropriate climate adaption conservation strategies (Khamis et al., 2014b).
Lessons learned from non-European regions
• Central Andes, Aconcagua Catchment, Chile
Chile represents a highly contrasting case to the European basins, from both a physical and a governance perspective. Water rights are a marketable commodity with minimal environmental regulation and no sectorial prioritization. Government institutions are highly centralized, with limited agency and capacity of water managers at the regional level. Increasing drought periods (from a combination of reduced summer runoff and altering precipitation patterns) are likely to compound current issues relating to the overuse of surface and groundwater from both legal and illegal abstractions in the Aconcagua Basin (Hill and Allan, 2014).
Major challenges from the governance perspective relate to the accuracy and applicability of monitoring data; lack of available, accurate, systematized and accessible information on water rights, water judgments, water market and prices, health and availability of water resources; climate data and uncertainty calculations are not used in planning; constricted agency and capacity of technical experts at the regional level; lack of trust and agreement on scientific studies and hydrological data that blocks concensus building; lack of a formal flexible conflict resolution mechanism.
A common thread across the different case studies are the challenges or opportunities presented by regional networks, actors and the differing levels of trust that can be capitalized for more integrated and longer term planning of water resources management. In the Aconcagua, there are a number of institutions that could be fostered to improve adaptive capacity. The Mesa del Agua is a collaborative attempt to strengthen inter-sectoral cooperation for the Aconcagua Project that should be made more inclusive with a broader, climate-related, goal orientation. While there is potential flexibility through the water rights system, denoting flexibility and autonomy to react quickly to changing conditions in the river, improvements to trust between river sections and more expedient and flexible conflict resolution mechanisms could greatly enhance its role in adaptation to drought conditions (Hill and Allan, 2014).
• Water governance in the Mendoza watershed in Cuyo, Argentina
During the ACQWA project, an analysis of the institutional environment in Argentina, its attributes and functionalities was undertaken, against the requirements to adequately respond to expected impacts on water resources and governance (Girón et al. 2014). Further, research focused on how the inherent constraints resulting from a dysfunctional institutional system have a potential to undermine adaptive capacity and adaptation actions to climate change. During the last three decades a weak institutional context created problems of credibility, coordination and cooperation and has had severe impacts on the quality of public policies. A weak institutional context is defined by low enforcement of the rules and/or by an application that is broadly discretional and one in which institutional change is frequent and radical. Hence, Argentina’s public policies tend to be unstable, poorly coordinated, weakly enforced, and highly rigid, and the decision making process lacks credibility. This persistent environment of instability, policy volatility, political short-termism and precarious enforcement pervades the water governance regime across scales and in the case study area, even in a formally highly decentralized federal system (as in other regions and sectors affected by the impacts of climate change). The decision making context, including drivers of water use, in a complex governance landscape, reinforces informational and cognitive, financial, social and cultural barriers to implement adaptation efforts and enhance adaptive capacities to improve the resilience of the social-ecological system. The scientific evidence that under projected climate change conditions the water resources of the Mendoza region could be reduced in a significant way thus requires the adoption of adaptive governance principles whose application is obstructed by rigidity, lack of cooperation and coordination, serial replacement (as opposed to predictable law and institutional structures) and fragmented or truncated learning processes and knowledge creation and dissemination.
• Climate change impacts on Kyrgyzstan
Within the ACQWA project, the Syr Darya river basin in the Tien Shan ranges have been chosen as a pilot area for a specific case study. The goal is to illustrate the impact of climate change on a transboundary river, where complex responses result from asymmetric power relations and less robust institutions. Kyrgyzstan has the role of a water tower as many rivers – such as the Syr Darya – originate within the country’s territory. The difficulty for the Central Asian states is to apply the principles of equitable use of water and to agree on a balanced reservoir management, which would allow the generation of energy in winter – benefiting upstream countries such as Kyrgyzstan – and irrigation for large-scale agriculture in summer – benefiting downstream countries such as Uzbekistan. Current challenges in the water operating regime are likely to be exacerbated by climate change impacts as water shortages during summer become more frequent with expected decreases in summer precipitation and reduced glacial meltwater releases due to smaller glacier volume (Sorg et al., 2012).
The Tien Shan mountains are located in the center of the Eurasian continent; the distance to the nearest sea exceeds 1500 km. Arid and semi-arid climate conditions result from atmospheric flows coming from the Atlantic and the Arctic, which lose major parts of moisture over the European territories and Siberia. Research within the framework of the ACQWA project and based on previous studies under the UNFCCC have demonstrated that the most probable climate change scenario fits the emission scenario B2-MESSAGE. By 2050, Kyrgyzstan will thus experience an increase in average annual temperature by +2.5°С and an increase in annual precipitation sums by +2.5%, with increasing amounts in winter and decreasing amounts in summer.
Sorg et al. (2012) show that these climate changes may have various impacts, in particular water shortages exacerbated by a decrease in glacial meltwater releases in the long-term due to reduced glacier volume. Glacier shrinkage is most pronounced in peripheral, lower-elevation ranges near the densely populated forelands, where summers are dry and where snow and glacial meltwater is essential for water availability. Under the high greenhouse-gas emissions SRES A2 scenario, 28-35% of today’s glacier volume in the Syr Darya catchment may melt by 2050.
Furthermore, shifts in seasonal runoff maxima have already been observed in Kyrgyzstan, and summer runoff will probably further decrease if precipitation and discharge from thawing permafrost bodies do not compensate sufficiently for water shortfalls. Runoff may decrease by 15%. Currently, however, river runoff in the formation zone of the Syr Darya basin is increasing due to intensive glacier melting. This trend is likely to continue up to 2025. After 2025, however, contribution of glaciers to river runoff in the headwaters of the Syr Darya will decrease irreversibly; as a consequence, the total hydropower potential of the Syr Darya basin may decrease by up to 15%.
Potential Impact:
The ACQWA project was formulated in response to the first call for climate-relevant projects under the EU 7th R&D Framework Programme (FP7). The philosophy of the project was based on the need to accurately assess the vulnerability of water resources in high-elevation, mid-latitude populated mountain regions. In such regions, declining snow and ice in a warmer climate are likely to strongly affect hydrological regimes, in terms of quantity, seasonality, and also quality. As a consequence of changing water availability, both upland and populated lowland areas will be affected. Rivalries and conflicts of interest may emerge as economic sectors such as agriculture, tourism or hydropower compete for water that may no longer available in sufficient quantities or at the right time of the year for these sectors to function. The challenge for the ACQWA project was thus to estimate as accurately as possible future changes in order to prepare the way for appropriate adaptation strategies and improved water governance. The project has enabled a suite of state-of-the art models to be applied, adapted, or developed to address many of the issues related to a changing physical world and to the socio-economic impacts that these changes will inevitable generate. Model results have also been used to assess how robust current water governance strategies are and what adaptations may be needed to alleviate the most negative impacts of climate change on water resources and water use.
Introduction
As the evidence for human induced climate change becomes clearer, so too does the realisation that its effects will have impacts on socio-economic systems and terrestrial ecosystems with multiple implications for society. Mountains are recognised as very sensitive physical environments with local populations that are highly exposed to rapid changes in the resource base on which their economic livelihoods are dependant. Moreover, policy priorities for alpine regions are often set by downstream actors, leading to trade-offs across different policy contexts that have the potential to be further exacerbated by changes in the hydro-climatic environment. While governance is well recognised as a core issue in current water resource related challenges, to date there is still a paucity of information on how adaptable water governance regimes in mountain areas could be to hydro-climatic changes impacts and the socio-economic impacts that these changes imply.
Large integrating projects generally represent a step forward in furthering our understanding of various complex processes and interactions between environmental, economic, social, and technological systems. The ACQWA project is no exception to this rule, and the five years of research has indeed enabled a number of issues to be refined and clarified, but has also identified problem areas that would need to be addressed in future investigations of this nature.
In January 2011, the ACQWA project organised a workshop in Riederalp, Switzerland, where over 25 EU projects focusing on water resources and water management were represented. Institutional and financial obstacles to data access for use in modelling exercises were identified, and gaps in scientific knowledge that contribute to uncertainty were highlighted. A working paper was subsequently published in 2012 in Environmental Science and Policy to report on the main conclusions of this crucial meeting. The discussions summarised in the paper have identified a number of sectors where these gaps often represent barriers to successful research outcomes, and suggested ways and means of alleviating some of these difficulties. A major issue that has been raised is that of data for research purposes. Policies aimed at ensuring free and unrestricted access to data, especially those generated by the numerous research projects that focus on issues of water availability, quality and management have been recommended. Implementation of the recommendations formulated in the Environmental Science and Policy paper may help pave the way for a more rapid and efficient production of research results that are of importance for policy guidance at the local, national and supra-national (EU) levels.
Socio-Economic Implications
ACQWA utilised advanced modelling techniques to quantify the influence of climatic change on the major determinants of river discharge at various time and space scales, and analyse their impact on society and economy. The main focus was to develop continuous transient scenarios from the 1960s up to 2050. By focussing on developing scenarios of change up to 2050 for a set of river basins, the project aims to develop climate information downscaled to temporal and spatial scales that are more useful to the challenges decision makers face (Beniston et al., 2011). This shortened modelling horizon allows for a more realistic assessment of the potential impact on the governance and socio-economic system components.
Key findings for different socio-economic sectors and themes
The following sections highlight the key findings from the ACQWA project with regards to socio-economic implications in key sectors across the case studies, namely, hydropower, agriculture and tourism (Fatichi et al., 2013a; Fatichi et al., 2013b; Fuhrer and Jasper, 2012; Fuhrer et al., 2013; Gaudard et al., 2013; Gaudard and Romerio, 2013; Gaudard et al., 2014; Gobiet et al., 2014; Hill Clarvis et al., 2013b; Stoffel et al., 2013; Toreti et al., 2013). In mountain regions, climate change takes on particular significance since snow and ice melt represent a large stream-flow component and a vital local resource for freshwater supply, hydropower generation, irrigation, tourism activities, and other industrial uses. Glaciers and snow pack act as vital natural storage systems, storing water as snow and ice through the wetter winter periods and releasing these provisions as flows during the drier summer months. Changes in precipitation and temperature will therefore impact both the quantity and timing of water available across these different sectors. Changes in the frequency or intensity of hazards (e.g. floods, glacier lake outburst floods, rock falls and landslides) are also likely to have a range of socio-economic implications as temperature and precipitation changes affect slope stability, glacier melt, snow melt and the zero isotherm.
Hydropower
• Variability in glacier retreat patterns (size, aspect, shape, debris cover, etc.) has consequences for the management of hydropower plants and dams, which depend primarily on snow- and ice-melt.
• Reduction in surface water flows and seasonal shifts in water availability (more availability of water in the earlier months of the year and a longer summer period with lower run-off) will impact hydropower. Climate change also indirectly affects electricity load because energy consumption varies with air temperature.
• Technological, economic and behavioural changes in the electricity system are, however, expected to exert a stronger impact on hydropower.
• In the Po region, greater variability in river flows and decrease in snow fall will affect the filling of hydropower reservoirs (e.g. decreasing ability to use all the storage capacity) and increase the inter-annual variability of electricity production.
• Storage-hydropower plants are a more flexible technology with modifiable production periods, whose revenues are less vulnerable to shifts in seasonality than run-of-river.
• While more even contribution from runoff might advantage reservoir management, a decrease in total annual runoff expected for reservoirs fed by ice melt is likely to negatively affect production.
Agriculture
• Until 2050 major agro-climatic risks will likely be caused by high temperatures rather than by increased drought (negative effects on both crop and livestock production).
• With increasing temperatures, water consumption through crop evapotranspiration increases is likely to lead to additional irrigation demands to maintain optimal yields (e.g. +10% in July at Visp across a range of climate scenarios up to 2049).
• High demand for water for irrigation will put additional pressure on small rivers in catchments with little or no water supply from glaciers, while larger water sources in valley may not be subject to the same extent of variability.
• In drier areas with low summer precipitation (e.g. valley floor and the south-facing slopes), potential water shortages for crop growth would be likely, requiring more irrigation to maintain optimal crop yields (max. +35%).
• In extremely dry years irrigation requirement could potentially exceed surface water availability in smaller catchments with a nival runoff regime (e.g. Sionne) where water is drawn through small irrigation channels for grassland irrigation.
• In the upper Rhone valley, improved water management should include both regulations regarding the allocation of water to different users of the same source, installation and management of reservoirs, and technical measures to improve the efficiency of irrigation by avoiding losses of distribution systems, evaporative losses, and excessive runoff due to over-application of water.
• Shifts in intensification of grassland use may have negative consequences for other ecosystem services such as biodiversity, soil carbon storage, and nitrogen retention.
Tourism
• A more local approach to winter tourism exposure to climate change in the Rhone catchment (comparing across average winters, a snow-poor winter with average economic conditions and a snow-rich winter at the beginning of the current economic crisis) reveals high disparities between mountain resorts and a higher vulnerability than regional approaches have suggested.
• Although Valais witnessed a smaller reduction in skier days than other regions during snow-poor winters, some resorts were strongly affected, with a large disparity in the vulnerability of resorts (e.g. small and medium resorts).
• The hotel sector is less exposed to a snow-poor winter than the cable-car companies, but faces difficulties in some areas when longer trends and socio-economic factors are taken into account.
• There is a real need to increase cooperation among and within resorts and touristic regions, enhance co-ordination of water uses (e.g. between hydro and cable-car companies), to improve the promotion of the Canton in general, to improve the ability to consider alternatives to skiing during warmer winters (notably in more vulnerable regions), and to improve current regulation on artificial snow-making.
Policy and Governance Implications
These impacts therefore need to be taken into account in the rules, rights and policies that structure the way in which water resources are managed across its multiples uses as well as the methods for protecting society from hydro-climatic related hazards. However, traditional governance and management approaches have often unsuccessfully coped with current internal or stochastic climate variability. Practitioners and scholars therefore not only need to address current rules and practices for managing historical challenges within the current envelope of uncertainty, but also to assess the adaptability of current frameworks for managing water resources and hazards to future climate change impacts.
In the Po basin, despite the farmers’ observed investments to adapt to mean changes in climate at local level, unanticipated variability in climate continues to impact crop yields. Policy interventions (integration or water resources, water storage) must therefore deal with residual uncertainty remaining after local adaptation to climate change (Bozzola and Swanson, 2013). In the Upper Rhone basin, the increasing heterogeneity of precipitation and late summer reductions in run-off from reduced glacier melt are likely to further exacerbate current local critical situations, bottleneck periods for local water supply, which are themselves related to: high levels of autonomy at municipal level that block longer term catchment scale planning and smoothing of bottleneck periods; lack of formal mechanisms to manage competition across catchment areas; lack of rules on emerging challenges and uses(Hill Clarvis et al., 2013b).
Earlier snow melt and shifting glacier melt patterns are likely to impact the inflexible and long term user rights that govern uses such as hydropower by introducing an extra layer of uncertainty and shifting the hydrological baselines upon which fixed and un-integrated rules and policies are based on at different governance scales and across different sectors (Hill Clarvis et al., 2013a). In response to climate change impacts as well as the continuing challenges concerning uncertainty, adaptation strategies are recommended that are no-regret, flexible and iterative, that allow for safety margins and redundancy in new investments, take a long term and social and green infrastructural approach (to complement grey infrastructure), and that integrate both adaptation and mitigation requirements.
In addition to the challenges of communicating and adapting to different scales of climate related challenge (i.e. internal or stochastic variability versus increased uncertainty and variability from climate change impacts), water resources governance and management adaptation must also deal with a number of other scale based challenges. These include: trade-offs across risk response when short term adaptation actions potentially undermining long term social-ecological resilience (Adger et al., 2011; Hill and Engle, 2013); balancing out proactive and reactive responses, as well as responses to multiple forms of stress at different magnitudes of physical change and scales of governance (Hill, 2013a); trade-offs between narrowly defined adaptation policies and other policy frameworks and economic sectors;
Cross-scale and sector trade-offs need to be better understood in the process of developing adaptation and broader environmental policy, plans and projects that address the impacts of climate change (Adger et al., 2011; Hill and Engle, 2013). Furthermore, downscaling climate and socio-economic impacts to finer temporal and spatial scales can give decision makers a clearer view of how these tensions might play out in a future climate, and therefore allow them to better prepare for and respond to them.
At the European level, while European climate change and adaptation policy is still in its infancy, reflection of how to account for climate change impacts is happening in relation to the Floods Directive and the next round of River Basin Management Planning for 2015 according to the Water Framework Directive. In Switzerland, a comprehensive climate change adaptation policy is in early development at the federal level. However, water resources use and management transcends multiple policy frameworks, and underlying trade-offs across different policy frameworks and sectors must still be better accounted for and remediated as we move into an era of potentially less manageable pressures from the hydro climatic environment.
$
Outreach and societal implications
Societal Implications: Understanding and managing uncertainty
Not only are the driving forces of climate highly uncertain, but fundamental scientific knowledge gaps limit the reliability of model projections (Hallegatte, 2009) with uncertainties in how climatic and non-climatic pressures will interact on different aspects of hydrology and ecology (Wilby et al., 2010). Traditional decision making tools, water management and infrastructure have however tended not to be developed to take account of the broader levels of uncertainty produced by climate change projections (Hallegatte, 2009). Despite scientific efforts focussed on the reduction of uncertainty, through enhanced data collection and modelling (Schneider and Kuntz-Duriseti, 2002), the uncertainty surrounding the specificity of climate change impacts remains a major challenge to planning and managing for future hydro-climatic conditions. However, policy makers, decision makers and water managers are increasingly recognising the need to develop better tools to manage and cope with both existing and increasing levels of uncertainty from climate variability and climate change impacts (Hallegatte, 2009).
While in many areas better managing for current and future levels of uncertainty is key, effective monitoring and data provision remains a critical component to remediating challenges relating to the current lack of knowledge and data on hydro-climatic, other environmental criteria and usage particularly in nations that are most at risk with the least data infrastructure. This is and should remain a core component of the on-going work on climate services, WMO standards and data sharing at international as well as national levels. While this is vital in developing countries, wealthier nations should not cease to continue to maintain and where possible improve their own monitoring and data infrastructure.
The growing body of work on climate change and climate change impact projections, adaptation and adaptive capacity have developed a broad set of principles, determinants and indicators for adaptive governance and management of water resources. However, there remains a need for clearer guidance on practicable mechanisms and actions at different levels of policy making, institutions and water management (Folke, 2011). This could help better guide the concrete changes that need to be made in governance and management frameworks to improve their sustainability and adaptability. Often there has been too much focus on the technological aspects of water resources management and adaptation and not enough on the governance and social infrastructural aspects of water systems (Adger et al., 2009; Hill, 2013b).
Outreach
The ACQWA project has produced over 280 scientific publications, a number of special issues in top tier journals such as Science of the Total Environment and Environmental Science and Policy, and over 130 instances of dissemination and outreach activities at academic conferences, public presentations, policy maker briefing sessions, newspaper and magazine articles and television news stories.
Conclusion
There is a clear need for a more integrated and comprehensive approach to water use and management. In particular, beyond the conventional water basin management perspective, there is a need to consider other socio-economic factors and the manner in which water policies interact with, or are affected by, other policies at the local, national, and supra-national levels. As an example, it is unclear whether current EU water policies are consistent with energy, agriculture, and other industrial policies.
The problems highlighted during the Riederalp meeting and summarised in the Environmental Science and Policy paper are also related to the inconsistencies between physical and socio-economic data and models. For example, figures related to water use may not be available at the temporal and spatial detail required by hydrologic models. Hydrological information is often based on basins whereas economic (and social) data is aggregated into administration regions. Thus, economic and physical data are often incompatible, because they are collected by different entities for different purposes. Future research should thus address the development of compatible data sets and the conversion between different data formats, as well as the development of toolboxes for up-scaling, downscaling and bias correcting data. Furthermore, the use of water in production processes is often not mediated by the market. The use of economic flexibility mechanisms in the allocation of water resources is quite rare, despite their potential in improving the efficiency of water resources allocation. More research and policy initiatives in this direction are thus necessary.
Finally, many scientists working in large integrated projects highlight a large gap between Science and Policy. This is certainly at least partly due to problems of communicating in an appropriate manner the key research results that would be of use to policy-relevant strategies. Awareness of this problem is increasing within the EC and other policy institutions, and hopefully this new momentum will be sustained over time so that conclusions from EU and other water-relevant projects will be widely incorporated into future policies at the local, national, and supra-national levels. Ultimately, the implementation of guidelines, maybe even an EU Directive, on the good governance of data (sharing) could be envisaged as a possible framework, providing advice and general rules on data formats and standards, data storage after project completion or the general terms of access.
List of Websites:
www.acqwa.ch
Project co-ordinator name: Prof. Martin Beniston
Project co-ordinator organisation: UNIVERSITE DE GENEVE
Phone: +41 22 379 07 69
Fax: +41 22 379 07 44
E-mail: Martin.Beniston@unige.ch
Project website address: www.acqwa.ch
