Skip to main content
European Commission logo
English English
CORDIS - EU research results
Content archived on 2024-05-27

Climate Induced Changes on the Hydrology of Mediterranean Basins: Reducing Uncertainty and Quantifying Risk through an Integrated Monitoring and Modeling System

Final Report Summary - CLIMB (Climate Induced Changes on the Hydrology of Mediterranean Basins: Reducing Uncertainty and Quantifying Risk through an Integrated Monitoring and Modeling System)

Executive Summary:
1 Executive Summary

The Mediterranean region is experiencing a broad range of threats to water security. According to latest climate projections, as published in the 5th Assessment Report of the Intergovernmental Panel on Climate Change (IPCC 2013) or the European Environmental Agency (EEA 2012), the region is at risk due to its pronounced susceptibility to changes in the hydrological budget and extremes, which is expected to have strong negative impact on the management of water resources and on key strategic sectors of regional economies. However, the current potential to develop appropriate regional adaptation measures towards climate change impacts suffers heavily from large uncertainties.

CLIMB put a major focus on the assessment and quantification of uncertainties in climate change impacts, related vulnerabilities and risk, by means of a comprehensive multi-model ensemble approach. In its 4-year design, a network of 20 partners from 9 countries, analyzed ongoing and future climate induced changes in hydrological budgets and extremes across the Mediterranean and neighboring regions. The work plan was focused on seven circum-Mediterranean river or aquifer systems, where the consortium employed a novel combination of field monitoring and remote sensing, data assimilation, integrated hydrologic modeling and socioeconomic factor analyses targeted to reduce existing uncertainties in climate change impact analysis.

Advanced climate scenario analysis was utilized and available ensembles of regional climate model simulations were audited and downscaled. This process provided the drivers for an ensemble of hydrological models with different degrees of complexity in terms of process description and level of integration. The results of hydrological modeling and socio-economic factor analysis were applied to develop a risk model via a spatial Vulnerability and Risk Assessment Tool, serving as a platform for dissemination of project results, including communication and planning for local and regional stakeholders, namely the CLIMBPortal. An important output of the research in the individual study sites is the development of recommendations for an improved monitoring and modeling strategy for climate change impact assessment.

The results from CLIMB largely confirm, but specify on the catchment scale the current state of the art on climate change research for the Mediterranean region, demonstrating that climate change is impacting water resources in a manifold and distinct fashion. Triggered mostly by a strong increase in temperature and a moderate to strong reduction and seasonal re-distribution of precipitation, impacts will mostly be felt in water resources management, agriculture, tourism and its consequent implications on income, value-at-risk and water security under increasing water scarcity. Selected results indicate that i) climate change contributes, yet in strong regional variation, to water scarcity in the Mediterranean; other factors, e.g. pollution or poor management practices are regionally still dominant, ii) rain-fed agriculture needs to adapt to seasonal changes; stable or increasing productivity likely depends on additional irrigation, iii) tourism could benefit in shoulder seasons, but may expect income losses in the summer peak season due to increasing heat stress, iv) local & regional water managers and water users, lack, as yet, awareness of climate change induced risks, with emerging focus areas are supplies of domestic drinking water, irrigation, hydro-power and livestock, and, v) data and knowledge gaps in climate change impact and risk assessment are wide-spread and ask for coordinated monitoring programs. Site-specific recommendations for adaptation to climate change have been developed in a dialogue with local and regional stakeholder networks.

However, while substantial progress has been made, uncertainties in climate projections and subsequent (hydrological) impact assessment remain and still impose strong limitations on water-related decision-making under conditions of climate change. This is particularly true due to the general lack of regional data and the yet unresolved mismatch of spatial and temporal scales of operation from different scientific perspectives.

Project Context and Objectives:
2 Project Context and Objectives

2.1 Project Context

According to current climate projections, Mediterranean countries are at high risk for an even pro-nounced susceptibility to changes in the hydrological budget and extremes. These changes are expected to have severe direct impacts on the management of water resources, agricultural productivity and drinking water supply. The different regions of the Mediterra¬nean landscape are already experiencing and expecting a broad range of natural and man-made threats to water security, such ass include severe droughts and extreme flooding, salinization of coastal aquifers, degradation of fertile soils and desertification due to poor and unsustainable management practices. It can be foreseen that, unless appropriate adaptation measures are undertaken, the changes in the hydrologic cycle will give rise to an increasing potential for tensions and conflict among the political and economic actors in this vulnerable region.

A number of major obstacles exist to implementation of adaptation measures designed to achieve sustainable management of water resources in Southern Europe, North Africa and the Middle East. Effective adaptation measures need to be prepared in a multi-disciplinary approach. While there is scientific consensus that climate induced changes on the hydrology of Mediterranean regions are presently occurring and are projected to amplify in the future, very little knowledge is available about the quantification of these changes, suffering from a lack of suitable and cost effective hydrological monitoring and modeling systems. Current projections of future hydrological change, based on regional climate model results and subsequent hydrological modeling schemes, are highly uncertain and poorly validated. The conditions required to develop and implement appropriate adaptation strategies are lacking. To the extent that adaptation initiatives are being proposed and adopted, they are primarily by perceptions of individual stakeholders and are rarely based on a multi-disciplinary assessment covering both natural and associated social and economic changes.

In its four-year-design, CLIMB aims at the improvement of modeling capabilities and the development of appropriate tools to advance the capability to assess climate effects on water resources and uses. The project consortium employed a combination of novel field monitoring concepts, remote sensing techniques, integrated hydrologic (and biophysical) modeling and socioeconomic factor analyses to reduce existing uncertainties in climate change impact analysis and to create an integrated quantitative risk and vulnerability assessment tool. Together, these provided the necessary information to design appropriate adaptive water resources management instruments and select suitable agricultural practices under climate change conditions. The integrated risk and vulnerability analysis tool enabled the assessment of risks for conflict-inducing actions, e.g. migration. Results were communicated to stakeholders and decision makers in a transparent, easy-to-understand form, enabling them to utilize the new findings in regional water resource and agricultural management initiatives as well as in the design of mechanisms to reduce potential for conflict.

To better assess the manifold consequences and uncertainties in climate impact on man-environment systems, CLIMB (ENV) was embedded in a coordinated topic in EU-FP7, bringing it together with the projects CLICO (SSH) and WASSERMed (ENV) in the research cluster CLIWASEC (Climate Change, Water and Security; for an improvement of scientific synergy and policy outreach.

2.2 Project Objectives

The CLIMB consortium specified the following major objectives: An analysis of climate change impacts on available water resources is undertaken in study sites located in Sardinia, Northern Italy, Southern France, Tunisia, Turkey, Egypt and the Palestinian-administered area Gaza. The work plan is targeted to selected mesoscale river or aquifer catchments, with areas in the order of up to a few thousand square kilometers in the above mentioned partner and SICA countries, representing water management units for regional water authorities (see Figure 1).

Fig. 1: The CLIMB case studies

Small-scale experimental sites within the watersheds are evaluated in depth for the extension and adaptation of existing and novel hydrological modeling tools. The site specific analyses enables improved assessment and quantification of region-specific vulnerability and risk factors for agricultural, drinking, residential and industrial water. Advanced climate scenario analysis techniques are employed and dynamical and statistical downscaling of available ensembles of regional climate model simulations is performed. This process provides the drivers for an ensemble of hydro(-geo)logical models with different degrees of complexity in terms of process description and level of integration. The outputs of the climate-hydrological modeling chain are focused to deliver estimates of changes in hydrological components such as water balance, total available water, drought indicators, frequencies of extreme values in stream discharge, soil moisture or groundwater balance.
Field monitoring and measurement strategies for surface and subsurface hydrological processes are tested and adjusted to the specific requirements in the study sites. Synergistic radar and optical remote sensing techniques are employed to provide steady state parameters (e.g. land use, land cover, soil hydraulic properties), to retrieve dynamic model parameters (e.g. soil moisture and roughness, vegetation structures), to monitor process variables (e.g. infiltration, water stress) and to validate model results. Data assimilation procedures are developed in order to incorporate relevant data and process understanding into existing modeling concepts, contributing to the reduction of uncertainty in predicted hydrological quantities.

An important output of the research in the individual study sites is the development of a set of recommendations for an improved monitoring and modeling strategy for climate change impact assessment. Once the model concepts are adjusted to represent the current-state hydrology in the study sites, they are tested over a range of selected climate change scenarios to project future hydrological budgets and extremes.
The integration of hydrological model results and socio-economic factor analysis enables the development of a GIS-based, modular Vulnerability and Risk Assessment Tool. This tool serves as a platform for the dissemination of project results, including communication with and planning for local and regional stakeholders as well as for the discussion and comparison of results with the teams working in the before mentioned CLIWASEC cluster.
All activities are conducted and evaluated in close co-operation with regional agriculture and water resources experts. This co-operation serves both to ensure a focus on adaptation measures appropriate for the region and ensure an optimized dissemination of project results. Valid findings are made available for improved site-specific monitoring and modeling systems for water resources and use assessments under changing climate and land use conditions.

2.3 Project Structure and Strategy

To cover many of the most evident and expected climate change and water security related problems in the region of interest, the Consortium has identified seven study sites in Southern Europe, North Africa and the Middle East.. The selected sites allow for the scientific analyses of the projected future course and impact of droughts and extreme flooding, salinization of coastal aquifers, degradation of fertile soils and water scarcity due to unsustainable management.
The project duration was first set to 48 months, then extended to 50 months. This duration was necessary and justified due to the difficult data situation in the region, the complexity of methods, the interdependency of work packages (WP) and the necessity to coherently disseminate the results and implications to local and regional stakeholders throughout and especially at the end of the project. All efforts in the scientific WPs are dedicated towards a gradual reduction of uncertainty in assessing climate change impacts on the hydrology of the sites under investigation and provide a better basis for achieving water security in southern Europe and the targeted SICA regions.

The work plan is divided in eight work packages. While WP 0 identifies and fosters the scientific synergies between CLIMB, WASSERMed and CLICO for the sake of a more focused and efficient policy outreach, WP 1 manages and co-ordinates the CLIMB Consortium internally. WP 2 provides and develops the common data infrastructure for and throughout the project. The WPs 3-6 focus on scientific research, development and innovation of technologies. None of these WPs stands alone, but are interconnected by means of interfaces, dependencies and feedback loops to ensure an iterative reduction of uncertainty and a more accurate assessment of water related ecological and economic risk. WP 7 is devoted to the interaction with stakeholders (see Figure 2).

Fig. 2: CLIMB Structure and Workflow

The workflow builds upon the scientific progress and accounts for a coherent dissemination of pro¬ject findings. The characterization of the various study sites (WP 3) is the scientific starting point of the project. All activities in WP 3 are accompanied by intense field campaigns for data collection, conducted by concerted partner actions in each site. WP 3 is sub-divided into three focal areas: i) screening and collecting existing ecological, meteorological, hydrological and socio-economical data (WP 3.1) ii) conducting hydro-geophysical field measurements to gather relevant information for in-depth hydrological process understanding (WP 3.2) and the parameterization of hillslope-scale hydrological models and iii) to up-scale and regionalize process descriptions using multi-sensorial and multi-scale remote sensing imagery (WP 3.3) and determine spatiotemporal distributions of crucial parameters (e.g. land use, soil moisture, roughness, infiltration, plant growth) for model applications at the catchment scale (WP 5).

The principle idea behind the structuring of WP 3 and WP 5 is the conviction that the gradual improvement of data availability and accuracy, consequently leads to improved model parameterizations and thus to a reduction of uncertainties in hydrological modeling. Most importantly, the partners complement each other in providing and interfacing surface water and groundwater models of different complexity and level of integration. This ensures that for each identified climate change related water security issue and site, there is an ensemble of hydrological models in changing configurations. This provides the opportunity to combine, compare, and cross-validate a wide methodological range in this field and supports the reduction of data- and model-prone uncertainties and the identification of the most efficient model complexity given specific boundary conditions and data availability. After calibration and validation of the model ensembles to current climate conditions, they are run over a range of selected climate change scenarios (elaborated in WP 4) to project hydrological budgets and extremes.

In parallel, WP 4 aims at designing robust procedures allowing auditing (intercomparison and verifica-tion) of products coming from different global and regional climate models. These procedures account for the statistics of average and extreme fields, for water balance conservation in the atmosphere and for prob¬lems related to spatial and temporal grid discretizations adopted in the different models. Further, the WP 4 bridges between the typical scale of climate models and the smaller scale required for hydrological modeling at the catchment scale (WP 5).

WP 6 establishes a comprehensive risk modeling approach for water resource problems under expected climate change in the selected study sites, integrating and quantifying the existing uncertainties (WP 6.1) stemming from insufficient data, inaccurate model descriptions, future climate projections and socio-economic vulnerability factors (WP 6.2). The results of the risk model are used to elaborate recommendations for future water management. The integration of hydrological model results and socio-economic factor analysis determines the development of a GIS-based, modular Vulnerability and Risk Assessment Model (WP 6.3) which serves as a platform for the dissemination of project results and as a communication and planning tool for local and regional stakeholders (WP 7).

Project Results:

3 Main Scientific and Technical Results

Findings from the CLIMB project are presented in the order of successive work packages and interacting working tasks. It includes a short introduction to the work of the CLIWASEC cluster and focuses on the main scientific and technical results.

3.1 CLIWASEC – Scientific Synergy and Policy Outreach (WP 0)

CLIMB, WASSERMed and CLICO joined forces to identify and foster scientific synergies and to establish a more focused and efficient policy outreach strategy. Major building blocks of this collaboration contained scientific exchange and review, identify and utilize complementary monitoring and modeling methods, harmonize and share data and discuss dissemination strategies or elaborate and propose adaptation alternatives. The projects agreed on joint annual general assemblies, a joint dissemination plan for presenting the results of the three projects in the scientific literature and the setting up a cluster project web-portal, which hosts and advertises further related projects. Policy briefs of the projects findings were prepared and posted on the cluster website on an event basis. At any time, regional, national and international stakeholders and policy bodies were and are still invited to express their research needs and recommendations.

To optimize benefits from the variety of cluster partners’ competences, joint research was devoted towards a better understanding and description of interfaces in such complex systems. Two main challenges were addressed: i) bridging scales and ii) quantifying and reducing uncertainty. Collaboration was build on the mutual integration of different methods from natural and social sciences. It contributed to better conceptualize each project’s research findings and propose solutions for water resource management under climate change, especially when a variety of different situations were covered in complementary case studies. The main findings of the CLIWASEC cluster were presented in a ‘Summary for Policymakers’ on climate change impacts on water an security in southern Europe and neighbouring regions (Figure 3).

Fig. 3: The CLIMB conceptual framework

3.2 Geodatamanagement – the CLIMBPortal (WP 2)

Managing Geodata was at the core of CLIMB. The WP 2 provided and developed the common data infrastructure for and throughout the project. In order to discover, visualize and provide access to selected main results of the project, a CLIMBPortal has been established under focusing on the utilization by scientists, regional planners and stakeholders (Figure 4). The CLIMBPortal serves as the central platform for interchanging spatial data and information. It stores and provides spatially distributed data about the current state and the future changes of the hydrological conditions within the seven CLIMB test sites around the Mediterranean. Maps of the outcome of the hydrological models - validated by the CLIMB partners - have been constructed and are offered to the public in forms of web mapping services (WMS). A selection of common indicators (e.g. runoff, drought index) including their changes over time are provided in different spatial resolution. Besides map information, the portal enables the graphical display of time series of selected variables calculated by the individual models applied within the CLIMB-project. Its implementation is based on GeoNode in v2.0.

Fig. 4: The CLIMB-Portal (

Among the major benefits of the CLIMB-Portal are solutions to i) the problem of heterogeneous data and file formats, ii) the missing of formerly unavailable ISO-compliant metadata, or iii) the lack of uniform presentations of model results. In addition, it offers i) compliance to required geospatial standards (e.g. INSPIRE), ii) maps and figures of easy-to-interpret hydrological indicators, iii) access to underlying data for registered users, iv) a WebGIS-client that integrates external Web Mapping Services and v) an open source solution for long-term availability of CLIMB results.

3.3 Environmental Monitoring and Data Collection (WP 3)

It is important to recognize that initially the case study characterization was limited to the available data, in several ways much to poor to perform a sophisticated modeling exercise. CLIMB activities included specifically designed field campaigns and data analysis, which greatly increased the level of understanding of conditions at each study site. These field activities were conducted at the required scale, either on the basin scale or even local fields, to perform a better understanding of the hydrological processes and to estimate the most sensitive parameters for the modeling phase. Addi-tional data collection may be necessary to fully understand specific processes; however, the executed field monitoring was selected to the extent practicable to fill data gaps and move forward with the implementation of hydrological models.

A guidance report was elaborated, early in the project period, to overview the main study site characterization requirements from the different CLIMB basins, leading to a sufficient overall site description. It is organized according to major site descriptors that require identification, for each case study, from existing monitoring programs and geographic information data bases. The site attributes were selected in order to provide an understanding of the basin management processes, highlight the physical basin characteristics, compare observations among study sites and establish a basis for hydrological modeling activities. A standardized description of each site attribute was also provided by the guidance report. The general attributes for each basin included location and ownership of information, land use/land cover basic site classification, basic hydrologic basin descriptors extracted from the basin topography, main surface and groundwater field monitoring activities, existing hydraulic structures (dams, hill dams, artificial recharge sites, etc), sources of pollution and general socio-economic information about the region.

Interaction between study site leaders:
Interaction between study site leaders provided a way to reconcile standards and requirements for basin descriptors among different protocols used by the different national databases. An effort was made by all partners to be as consistent as possible in the classification and the means of the site attributes. Transformation of geographic attributes from national coordinate systems to UTM system was needed and conducted

Data collection and analysis:
Field data collection efforts included topographic surveys, vegetation surveys, collection of water and soil samples and hydro-geophysics surveys. In addition to these efforts, an assessment of hydrologic features for each basin, analysis of the soil sampling data, and analysis of vegetation data and hydro-geophysics interpretation were also performed. Recent Land use/land cover (LULC) maps were also elaborated for all study sites. Multiple classification techniques are used to extract the spatial-temporary evolution of LULC information from multispectral remotely sensed images.

Elaboration of a Basin Information Repository:
A CLIMB basin Information Repository was constructed for all CLIMB basins and implemented as a database developed in Microsoft Access. This repository provides a screening of all available in-formation related to basic characteristics of topography, soils and land uses within the basins, available resources for surface water and groundwater, existing climatic data as well as general socio-economic conditions and sources of pollutions. It provides key site characteristics, the state of monitoring within each basin and a tool for comparison between the CLIMB case studies. It is thus undoubtedly a previously unavailable and very comprehensive compilation of data in each of the CLIMB case studies.

Fig. 5a and 5b: Examples of soil moisture maps over bare agricultural areas in the Thau basin, France

Remote Sensing Products

CLIMB employed radar and optical remote sensing data to provide crucial spatial information for the parameterization, calibration and spatial validation of the hydrological models. Focus was given to the following parameters and state variables (1) landuse/landcover (LULC), (2) soil moisture, (3) vegetation parameters (albedo and leaf area index (LAI)), (4) infiltration information and (5) snow cover. Figure 5a and 5b shows two examples of remote sensing soil moisture products, derived from using an inversion technique based on neural networks, which were used to improve model parameterizations in the Thau catchment, France (Baghdadi et al. 2012).

Policy Implications and Research Needs

The Basin Information Repository reveals serious informational, temporal and spatial data gaps in the investigated region.
Surface water is poorly monitored in most case studies. Climatic parameters, such as temperature, humidity, rainfall, wind velocity etc., key parameters for defining climate conditions, are not available in sufficient resolution and density.
Remote sensing and specifically designed geophysical field campaigns proved useful to fill some of the existing gaps, but certainly not all. Improvement can be expected from the upcoming ESA Sentinel satellite missions.
Thus, it will be important to establish specific monitoring plans for climate change follow up studies in the Mediterranean basin. The specific needs are often site specific and thus need a local and regional approach.
Research is still needed to a) better understand hydrological processes, specific to semi-arid climates, and b) to develop modeling schemes which are robust even under scarce data conditions.

3.4 Climate Model Auditing and Downscaling (WP 4)

Accurate assessment of future hydrological tendencies, as well as their impact on the agricultural and socio-economical sectors, depend critically on several sources of uncertainty. Climate model signals constitute the main forcing in a multi-model cascade (i.e. climatological, hydrological, socio-economical, crop evolution models etc.), implemented for the evaluation of climate change impacts. That said, uncertainty reduction of the climatological representation at hydrologically relevant scales has been the main objective of the CLIMB workpackage WP4 “Climate Models auditing and downscaling”, as a means of providing reliable inputs for subsequent hydrological modeling.

Climate Models selection
An intercomparison of the performances of several climate models was conducted, taking into account their effectiveness in reproducing the precipitation and temperature fields as well as the climatologies in all Mediterranean catchments considered in the CLIMB project. The set of climate model outputs was extracted from the FP6 ENSEMBLES project database, generated by coupling different GCM (Global Climate Model) and RCM (Regional Climate Models) outputs. To evaluate climate model performances, E-OBS precipitation and temperature datasets were used as reference fields: being distributed on the same grids as ENSEMBLES outputs, E-OBS allowed for standardization of the intercomparison for all considered catchments and climate models. Model selection was further constrained in order to maintain at least two different RCMs nested in the same GCM, and two different GCMs forcing the same RCM. The four best-performing GCM-RCM combinations were selected to drive hydrological model simulations for the CLIMB activities: a) 'HCH-RCA'= HadCM3 - High Sensitivity (UK) driving RCA (Sweden); b) 'ECH-RMO'= ECHAM5/MPI (Germany) driving RACMO2 (Netherlands); c) 'ECH-REM'= ECHAM5/MPI (Germany) driving REMO (Germany); d) 'ECH-RCA'= ECHAM5/MPI (Germany) driving RCA (Sweden).

Fig. 6: Climate Models selection.
Scatterplot of weighted precipitation and temperature errors in the mean and standard deviation of monthly climatologies over the 60-yr verification period (1951- 2100), for each of the 14 ENSEMBLES models with respect to E-OBS observational reference. Large circles indicate the four best performing selected models. Catchment: Rio Mannu, Sardinia (Italy).

Large-scale bias corrections
Most Climate Models, including the four best-performing ones used in the CLIMB project, display more or less deficiencies in reproducing the hydrological balance, climatological features and sea-sonality in several of the major river-catchments around the Globe. These problems are further amplified when considering smaller-sized catchments, as happens in CLIMB. Thus, we proceeded with calibration and bias correction of the climatic signals in each catchment area considered. More precisely, we used E-OBS precipitation and temperature data and a procedure based on probability distribution matching to correct biases and distribution errors in past- and future-climate signals.

Local-scale bias corrections
E-OBS products, while being a useful reference for a standard intercomparison evaluation among different areas, may fail to reproduce small scale features of precipitation and temperature fields in the targeted catchments. One major uncertainty is certainly related to the low density of meteorological stations pooled together in order to analyze and construct the E-OBS fields. Indeed, such a network may not be sufficient to accurately represent the local climatologies. This problem is particularly relevant for the precipitation fields, which are typically characterized by a small correlation range. In order to tackle the issue, the mean areal monthly precipitations and temperatures were calculated for each catchment area using local data (i.e. made available from CLIMB partners), and used to evaluate and properly correct the residual biases. Figure 7 shows an example of residual bias correction for a catchment in Italy.

Small scale interpolation and downscaling
An additional source of uncertainty is related to the smoothing effect induced by the coarse grid resolution of climate models (about 25 km resolution). Their outputs are incapable of reproducing the small-scale variability introduced by orographic effects and the intrinsic intermittency of meteorological fields. This is especially the case for precipitation, where interest is in the representation of hydrological balances and, more in general, the soil-water-atmosphere transfer processes at catchment scales. In the CLIMB project we applied different downscaling schemes to interpolate, to smaller scales (1-5 km resolution), the main variables controlling the hydrologically relevant exchanges. A multifractal model for space-time rainfall downscaling was applied to statistically reproduce the scale-invariance and small scale fluctuations found in the observed fields. Orographic effects were further introduced using a modulating function. Temperature fields were in-terpolated using a distance-depended weighting function, and a dynamic lapse rate to account for elevation variations. The later was calculated locally allowing for time-variations. Relative humidity was interpolated using a similar approach as that used for the temperature field, by considering the corresponding dew-point temperature. The wind speed and direction as well as the solar radiation were interpolated using a morphometric approach that accounts for the small scale terrain elevation, azimuth and slope. Figure 8 shows an example of interpolated precipitation fields.

Fig. 7: Local scale bias corrections.
Comparison of mean monthly observed MAP (mean areal precipitation; grey thick line) and mean monthly RCMs MAP, before cor-rection (dashed lines) and after correction (continuous line). Catchment: Rio Mannu, Sardinia (Italy).

Fig. 8: Small scale interpolation and downscaling.
Annual mean of high resolution precipitation field (about 1 km) after bias correction and application of a space-time rainfall downscaling model.

Overall uncertainty of climate forcing
All steps described above aimed at reducing the uncertainties related to the climatic component, while introducing the natural small scale variability unresolved at climate model grid scales. Hence, the main achievement of WP4 is the evaluation of climate model uncertainties at local scales, and the quantification of their contribution to the overall uncertainty of climate change impact assessments. As an example, Figure 9 compares the initial spread in the climatology of all climate models extracted from ENSEMBLES, with the spread of the four best performing ones selected for subsequent hydrological simulations.

Fig. 9: Overall climate forcing uncertainty.
Boxplot of climatological means of precipitation (mm/y) over five 30-year periods between 1951 and 2100 for the 14 ENSEMBLES models: lines of boxplots correspond to median and quartiles, whiskers extend to the minimum and maximum values from the RCMs. Symbols are superimposed for the 4 best performing RCMs and E-OBS climatologies. For E-OBS, climatologies are computed only for the first two 30-year periods.

3.5 Integrated Hydrological Modeling (WP 5)

The objective of this WP was to apply hydrological modeling in order to predict the response of the watershed to changing driving forces such as climate, land use, urbanization, and human activities. In order to better quantify the hydrological model uncertainty, an ensemble of different models with varying complexity in parameter space was applied in each study site. The activities can be classified into two categories: i) small- to medium-scale modeling to study and identify key hydrological processes and improve the predictive capability of hydrological models and ii) large-scale modeling to assess the impact of changes in climate, land use and water demand on water resources evaluated at the watershed scale. These activities were conducted in tight collaboration with the other WPs, and in particular with WP 3 and WP 4, which provided the input to the modeling activities, and with WP 6 which used the data from the modeling activity to conduct risk assessment. Modeling was conducted in a unique multimodel ensemble framework in order to properly address structural and parametric uncertainty.

The hydrological model ensemble has been applied to the study catchments, using as input the results of the climatic models for the reference (REF, 1971-2000) and the future periods (FUT, 2041-2070). Overall, most catchments show in the 2041-2070 scenario a reduction of precipitation and an increase of temperature, which lead to an increase of potential evapotranspiration. However, the actual evapotranspiration reduces with respect to the control period as an effect of a drier soil (i.e. a lower soil water content). An exception of this general behavior is the Noce catchment in the Southern Alps, where the increase of both temperature and precipitation lead to an increase of actual evapotranspiration on the future scenario. The most relevant changes observed in the study sites are summarized below, supported by a selected representative figure.

3.5.1 CHIBA, Tunisia

In the Chiba catchment, located in the Cap Bon region of Tunisia, hydrological simulations have been conducted with the models SWAT (CERTE, UT) and WASIM (DLR, LMU). The mean annual precipitation is projected to reduce of 18% in the future scenario accompanied by an increase in the range of 7-12% of potential evapotranspiration as an effect of the increase of temperature. However, due to a reduction of the soil water content (in the range 14-22%) the actual evapotranspiration reduces of about 11%, with a slight difference between the two models (about 1%). As expected, the difference between precipitation and actual evapotranspiration, which is the water potentially available, reduces of 43% and 52%, according to the simulations conducted with PROMET and SWAT, respectively. Most importantly, the runoff reduces significantly, ranging between 39% and 51%, again according to PROMET and SWAT, respectively. Despite the differences, due to the different epistemic error, both models agree in projecting a dramatic reduction of the available water resources, both in term of surface water and recharge to the underlying aquifers (percolation reduces of 52% in WASIM and 22% in SWAT. The reduction of the runoff seems distributed over the entire hydrological year with the larger reduction in the winter months, when runoff reaches its maximum. Figure 10 shows the evolvement of a Drought Index, indicating the number of months in a 30-year period (REF, lower row vs. FUT, upper row) with a significant mean monthly soil water deficit, modeled with the hydrological model WASIM and driven by the four selected climate forcings in WP 4.

Fig. 10: Drought Index value in Chiba for the reference and future period as predicted by the WaSIM using various climate forcing.

3.5.2 KOCAELI, Turkey

Three hydrological models (PROMET (VISTA), mGROWA (FZJ), MIKE-3 (YTU)) were applied in this site, driven by the same set, but regionally adjusted set of climate forcing. Due to the global warming predicted in the climate models and the resulting shift of rainfall periods with increasing precipitation during the winter months and decreasing precipitation during the summer months, total runoff is simulated to increase in winter. The level of groundwater recharge within the hydrological winter half year seems to remain stable. However, the increasing temperature in future spring and summer months combined with low precipitation will probably lead to a more extensive depletion of the soil water storage and thereby to an increasing irrigation demand on agricultural land in order to keep agricultural yields constant. HCH-RCA, in comparison to the other climate models, predicts the driest conditions for the future with a decrease of 106 mm (-13.1 %) in mean annual precipitation and an increase of mean annual temperatures of 3 °C (+23.3 %). This leads to a decrease in percolation of water from the lowest modelled soil layer, which indicates a decrease of ground water recharge and an increase in drought risk (Figure 11). MIKE 3 HD simulation runs were performed for the Izmit Bay using river discharge data provided by the PROMET (VISTA) and mGROWA (FZJ) models. The simulations were carried out to assess possible variations in sea level in Izmit Bay. The results obtained have shown that the maximum sea level change was found to be 0.6 m. The most important result obtained out of these model runs was that the sea level change can be observed clearly in the inner sections of the bay where the maximum sea depth is 35 m. It was also found out that these changes were more prominent for the future periods, especially for the year 2068.

Fig. 11: PROMET monthly mean actual evapotranspiration divided by potential evapotranspiration as an indicator for drought condi-tions. REF period 1971-2000 (left), FUT period 2041-2070 (right). 300m spatial resolution. Climate model forcing is HadCM3 RCA.

3.5.3 NOCE, Italy

In the Noce catchment located in the Southern Alps hydrological simulations have been conducted with PROMET (VISTA) and GeoTransf (UNITN). Despite the diversity in the modeling approach these two models project a similar catchment functioning in the future scenario. Due to the combined effect of the rise of temperature and of the increase of precipitation (6.4-8.7%) the actual evapotranspiration increases (10.51-12.29%), but to a less extent with respect to the precipitation, leading to a slightly larger amount of water resources available in the future period (Figure 12).
The total runoff increases by 16.5% and 15.6% according to PROMET and GeoTransf, reepctively. This is a significant amount, which correspond to 125 mm, with PROMET and 107 mm, with GeoTransf, of additional runoff. However, this volume is not uniformly distributed through the year. Due to the larger temperature in winter the freeing line is expected to rise and consequently more winter precipitations than in the past will be liquid. This leads to an increase of the winter runoff and a parallel significant reduction of the runoff in late spring and beginning of the summer, from May to August). Overall, the distribution of the runoff trough the year is projected to be less variable than in the reference period. The difference between the two models increases when considering percolation, which with some approximations can be considered as the recharge of the aquifers, PROMET projects a slight reduction (-2.5%), with GeoTransf projects a moderate increase (+15%). Both models agree in projecting a quite large reduction of Snow Water Equivalent (-65% and -63% with PROMET and GeoTransf, respectively) during the winter, and this is in agreement with the considerations above.

Fig. 12: PROMET modelled run-off for the REF (blue) and the FUT period (red) and for all 4 climate models (range). Displayed are mean monthly sums for the whole Noce catchment.

3.5.4 RIO MANNU, Italy (Sardinia)

In the Rio Mannu, located in the southern part of Sardinia, hydrological simulations have been conducted by using SWAT (CRS4), WaSIM (LMU) and tRIBS (CINFAI). Similarly to the other catchments in the Mediterranean area the annual precipitation reduces of about 12%, while the potential evapotranspiration increases (+0.75% with SWAT and 14% with PROMET) due to the increase in temperature. However, both PROMET and SWAT project a similar reduction in the actual evapotranspiration due to the smaller soil water content (-10.2% with SWAT and -8.9% with PROMET). The total water potentially available (the difference between precipitation and actual evapotranspiration) reduces between 14.8% and 22.6% as projected by SWAT and PROMET, respectively. This results in a significant reduction of the runoff, which is projected to reduce of 15.5%, 19.7% by SWAT and PROMET, respectively. tRIBS provide results in line with those of the other two models but with larger reductions, with the reduction of runoff of 31%.
The models seem to represent the current hydrological situation in a plausible manner. All water balance components like precipitation, discharge, potential evaporation, actual evapotranspiration, and percolation show significant changes in the modeled future time series. Temporal changes within a hydrological year and over the complete simulation time series are more distinct than spatial changes. Over the whole hydrological year higher temperatures and higher values of potential evaporation are projected. Precipitation rates are expected to slightly increase in the winter months (January), but will decline for most of the other month with strongest impacts in spring. This might cause situations of water shortage at earlier times of the year. Strongest changes within the hydrological year have been determined for actual evaporation, soil water content and soil water deficit in spring and autumn. The strong decline of actual evapotranspiration and soil moisture in the spring time is a consequence of the decreased precipitation rates. The major area of the catchment is used for intensive agricultural production. Especially this agricultural production in the catchment is endangered by those climate change impacts in the future spring season. It can be assumed that the irrigation water demands will rise while water availability is dropping in the spring time and/or that the growing season is shifting to an earlier onset of the year.

Fig. 13: SWAT modelled change in the number of low flow days between FUT (red) and REF (blue)

3.5.5 THAU, France

Two hydrological models were applied to the Thau region in southern France, SWAT (CERTE, UT) and mGROWA (FZJ). For precipitation, projections of the climate models suggest a general decrease of rainfall. In winter a marked decrease in precipitation is likely to occur ranging between -15% to -25% in the Thau catchment. The decrease in precipitation is likely to be more pronounced in summer than in the other seasons where it can reach -30%. According to the SWAT simulations, the Thau catchment will experience a decrease in monthly flow discharge from April to December, with May being the most altered month (-70%), while insignificant increase in monthly discharge of January and March is projected. Although 95% of the projected magnitudes of change confirm that summer monthly discharges are likely to decrease at the horizon 2040, the corresponding projected amplitude of change is uncertain. This uncertainty tends to increase in the direction and amplitude of change as moving from the dry season (May to September) to the wet season (October to April) which confirms the results of the projected change in precipitation and low and high flow frequency.
According to mGROWA, total runoff is projected to decrease as well. Additionally, drought severity will increase in any possible future development path under consideration. The increasing temperature in future spring and summer months combined with low precipitation will probably lead to a more extensive depletion of the soil water storage and thereby to an increasing irrigation demand on agricultural land in order to keep agricultural yields constant. Groundwater recharge levels within the hydrological winter half year seem to remain stable only in one GCM-RCM-combination, whereas the others tend to project significant lower levels (Figure 14). It can be concluded from the bandwidth represented by the ensemble that groundwater withdrawals have to be restricted in order to maintain the groundwater resources of the region, which are in general very little. As a consequence of climate change, the region will be even more dependent from water import to satisfy a present-day demand.

Fig. 14: Mean annual groundwater recharge simulated using mGROWA based on the climate model ensemble and change in the future period against the past period, Thau case study area.

3.5.6 GAZA, Palestinian Administered Areas

The hydrological model WaSIM (LMU) and the hydrogeological model CODESA-3D (CRS4, IUG) were applied in combination to this aquifer site. Three out of the four applied RCMs project a severe decrease in precipitation for the future period 2041 – 2070. However, the RCM driven by HadCM3 indicates a slight increase in precipitation. As the limiting factor for evapotranspiration is water availability, the real evapotranspiration is hence slightly decreasing by up to 5%. Due to raising temperatures, however, potential evapotranspiration is increasing, causing a smaller share of actual to potential evapotranspiration, shown here as evapotranspiration index ETI. The ETI is especially reduced in spring months, causing further irrigation needs for agricultural purposes during this sensitive period. The deterioration in the water balance of ~25% leads to further exploitation of the groundwater aquifer, if irrigation methods will not be improved in the future, or stronger regulations for the water consumptions are met in the Gaza Strip. The almost dramatic increase in drought months (Figure 15) emphasizes the need for adaptation strategies to maintain agricultural activities in the area and preserve the groundwater aquifer. Considering the ongoing salinization, caused by sea water intrusion, a certain water level in the aquifer is essential to preserve the areal freshwater reserves.
Since the water demand is estimated to increase, according to future trends in the Gaza Strip until 2035, three different scenarios have been considered:
- the ‘worst scenario’, which assumes the estimated increasing of pumping;
- the ‘best scenario’, according to a management scenario which considers a decrease of pumping;
- the ‘intermediate scenario’, representing an ‘averaged ‘management scenario between 1) and 2)
The more striking result obtained within the Gaza coastal aquifer study is that the only way to prevent and control sea water intrusion consists in assessing and properly adapting a groundwater management strategy; in fact, different climate scenarios lead to differences in the groundwater system that can hardly be appreciated when compared with pumping effects.

Fig. 15: Total number of months in a thirty year period, reference (lower row) and future period (upper row) for all four RCMs, with mean soil water deficits <-0.3 according to the CLIMB Drought Index.

3.5.7 NILE DELTA, Egypt

The hydrological model WaSIM (LMU) and the hydrogeological model MODFLOW (UZ) were applied for reaches of the Nile Delta between the Gharbia Governorate and Alexandria. The water balance simulation model WaSiM was forced with an ensemble of four GCM-RCM combinations, resulting in 4 scenarios for future conditions. Simulations where carried out for the reference period 1971-2000 and the future period 2041-2070. All four of the applied RCMs project a decrease in precipitation for the future period 2041 – 2070, however, magnitude varies strongly from 1.7 to 23%, but for absolute annual precipitation rates of 30 to 50 mm. As the limiting factor for evapotranspiration is water availability, the real evapotranspiration is hence slightly decreasing by up to 2.5% for all RCM simulations (Figure 16). Due to raising temperatures, however, potential evapotranspiration is increasing by about 5%, causing a smaller share of actual to potential evapotranspiration, shown here as evapotranspiration index ETI. The ETI is especially reduced in early spring, causing further irrigation needs for agricultural purposes during this sensitive period. Due to the small share of pre-cipitation to evapotranspiration the water balance is already far negative, due to irrigation from the groundwater and from surface water. As the changes in precipitation are comparatively small, in absolute numbers, to evapotranspiration, the effect on the water balance is almost negligible. However, the precipitation decrease causes less recharge for the aquifer, and also less water in the Nile upstream available for irrigation, which leads to further exploitation of the groundwater aquifer, if irrigation methods will not be improved in the future, or stronger regulations for the water consumptions are met in the Nile Delta.

Fig. 16: Real evapotranspiration for the Nile Delta resulting from WaSiM simulations. Red band shows the range of the four future runs, forced with different RCM scenarios, blue band the range for corresponding reference period

3.6 Uncertainty Analysis, Socioeconomic Factor Assessment and Risk Modeling (WP 6)

The main objective of the WP was to establish a comprehensive risk modeling approach for water resource problems under anticipated climate change in two selected Mediterranean Basins. Taking advantage of data and information provided by WP3, WP4 and WP5, the risk assessment of income losses (and out-migration) expected to arise directly from climate change and/or indirectly from climate induced changes in the hydrology of the two super sites Rio Mannu di San Sperate (Italy) and Chiba basin (Tunisia) was conducted. The related analyses focused on two highly water intensive sectors of substantial economic importance for the two studied river basins or their close surroundings, namely agriculture and tourism. Key results of the analyses were incorporated into the comprehensive risk assessment that was established. This holistic risk assessment for the two super sites comprised the analysis of the factors that will influence the supply and the demand of water under future scenarios, taking into account direct climate change stressors and interacting stressors that will exacerbate the impact of climate change on water security under future scenarios. The holistic risk assessment was summarized in terms of water supply/demand matrices, including quantitative information from the hydrological modelling (water supply) and from socio-economic modelling (water demand) as well as further important qualitative information. Results from uncertainty analysis were considered as well. Overall, the water supply/demand matrices compiled within WP6 are supposed (i) to help getting a more comprehensive picture of the trends to be expected with respect to water use conflicts between stakeholders, (ii) to serve as an impetus for a fruitful discussion between water managers, water users, government officials as well as other stakeholders and experts, and (iii) to support the identification of priority fields for adaptation. Basic suggestions on potential adaptation strategies and future water resource management options were elaborated. The main achievements summarize as follows:

A new comprehensive uncertainty framework for water related risk assessment has been de-veloped for all test sites within WP6

The concept focuses on investigating two kinds of uncertainties within the hydrological impact mod-elling of WP5 as well as the socio-economic vulnerability factor/risk assessment (related to income loss) for the agricultural sector and tourism. One kind of uncertainty arises due to the effect of using several climate data sets on a final and robust model configuration of a respective hydrological impact model or vulnerability model, named Climate Signal Uncertainty Study (CUS). The second kind of uncertainty stems from the effect of performing CUS with several hydrological impact models or different settings of a vulnerability model named Model Structure Uncertainty Study (MUS). CUS and MUS will be investigated in the super sites Chiba basin, Tunisia, as well as Rio Mannu di San Sperate, Sardinia, for the two hydrological impact models WaSiM (Water balance Simulation Model) and SWAT (Soil Water Assessment Tool) because those hydrological impact models are applied in both super sites. The vulnerability models are used within the socio-economic vulnerability factor/risk assessment modelling only. Those efforts are focused on the two super sites as well. Figure 17 illustrates the structure of the uncertainty assessment within CLIMB.

Fig. 17: Structure of the uncertainty assessment within CLIMB.

Quantification of climate change impacts on water components and associated uncertainties

In Chiba basin WaSiM and SWAT project a decline in PRC (precipitation) of about -18% between 1971-2000 and 2041-2070. SWC (soil water content) dwindles in the range of -15% (WaSiM) and -25% (SWAT) respectively. TAW (total available water) shows a reduction of roughly -30% in both hydrological impact models. Model structure uncertainty (MUS) is small for projections related to PRC and TAW. MUS is medium when looking at SWC. Uncertainty errors of the analyses stemming from using four different climate models (Climate Signal Uncertainty Study CUS) are below 10% for all three indicators (PRC: ~3%, SWC: ~3%, TAW: ~8%). In Rio Mannu WaSiM and SWAT project a decline in PRC of -13%, SWC dwindles about -11% and TAW is reduced by -14% (SWAT) to -21% (WaSiM). MUS is small for PRC as well as SWC and medium for TAW. Uncertainty errors of the analyses due to CUS are below 10% except for TAW derived from the model WaSiM (18%). Besides smaller differences, both hydrological impact models show comparable trends and magnitudes of change for the two super sites Chiba basin, Tunisia, as well as Rio Mannu di San Sperate, Sardinia. Those findings are useful for the subsequent assessment of associated risks for crop yield reduction as well as income loss in the agricultural sector.

A comprehensive risk model has been designed for and implemented in the two super sites Chiba basin, Tunisia, as well as Rio Mannu di San Sperate, Sardinia

Findings about climate induced changes derived within the hydrological impact modelling efforts in WP5 and associated uncertainties are combined with the socio-economic vulnerability factor/risk as-sessment for the agricultural sector and tourism in order to create a comprehensive risk model of income loss for each sector. For the agricultural sector, an empirical approach as well as the AQUACROP model are used to simulate crop yield responses associated to the simulated climate change impacts stemming from WP5 for two crops, namely tomatoes (Chiba basin only) and winter wheat (both super sites), and for two kinds of management practices, namely under rain-fed and irrigation conditions. For the tourism sector, temperature and precipitation data from climate models are driving a special tourism model which assesses the impact of changing climate conditions on revenues in tourism. For both sectors, four different combinations of global and regional climate models are used for the reference (REF) and scenario period (FUT), each covering a time span of 30 years. This results in considerable uncertainty due to a span of simulated annual values. Figure 18 shows sources of uncertainty and their propagation in CLIMB. The hydrological impact modelling was performed with four climate models (CUS) and two hydrological impact models WaSiM and SWAT (MUS) in the super sites Chiba basin, Tunisia, as well as Rio Mannu di San Sperate, Sardinia. The climate signal uncertainty and model structure uncertainty resulting from a span of the results from those efforts are used as input for the comprehensive subsequent risk assessment in WP6. A series of quantitative (derived within preceding modelling efforts) and qualitative aspects (directly or indirectly derived from modelling work and local knowledge) are brought together to create a holistic picture for the risk assessment (Figure 19).

Fig. 18: Propagation of uncertainties in CLIMB.

Fig. 19: Causal chains of climate induced changes affecting the hydrology of the two super sites Chiba basin, Tunisia, as well as Rio Mannu di San Sperate, Sardinia, and interacting stressors leading to an increase in risk in the water sector. Risk is described here as a more qualitative likelihood of occurrence that the water supplies cannot meet the water demand in a specific case study site.

Risk assessment in Agriculture (example: Chiba basin)

Risk assessment for the agricultural sector in Chiba basin investigates climate change impacts on crop yields of tomato and wheat. The risk analysis indicates the following:

Tomato yields with unlimited water for irrigation are expected to increase by 20-23% in 2040-2070 as compared to the present. However to achieve these yields, the amount of water required for irri-gation is expected to increase by approximately 11%
The yields are sensitive to the amount of irrigation water required. There is an irrigation threshold below which the risk of crop failure increases and the yields decrease. The modelled yields are for irrigation at this threshold value.
- If irrigation requirements are kept constant and “business as usual management” then yields are expected to decrease by 2%
- If irrigation decreases by 10% as expected and “business as usual management” then yields are modelled to decrease by 45%
- If irrigation decreases by 10% as expected but only August 15th plantings are allowed then yields are modelled to decrease by 24%
The water efficiency of the tomato production can improve by using mulches as an adaptation mea-sure.
Wheat yields, without irrigation, are expected to increase by 7 – 16% in 2040-2070 as compared to 1971-2000
Introducing irrigation increases yields by 30% but may require about 2,000 m3/ha/year in 2040-2070
Using empirical modelling and assuming no increase in producing area or irrigation, climate change is expected to
- Increase the production of vegetables by 12%;
- increase the production of citrus fruit by 2%;
- increase the production of legumes fruit by 8%;

Risk assessment in Tourism

Using a simple climatic beach index about tourists’ perceptions on unacceptably temperature and rain conditions for “sea, sand and sun” (3S) tourism, potential impacts of climate change on tourism in the wider surroundings of the two super sites were analysed. According to this index, climatic conditions for the dominant tourism type in the surroundings of the two super sites – i.e. 3S tourism – are expected to further improve in shoulder seasons, but may deteriorate in the current summer peak season (particularly in July and August) due to increased heat stress. Hence, based on the currently observed relationship between tourism demand (indicated by overnight stays) and the applied climatic beach index, there is a chance of climate-induced income gains during the shoulder seasons in spring and autumn and a risk of climate-induced income losses during the summer months. Annual net impacts are however expected to be (slightly) positive (Figure 20).

Fig. 20 : Expected change in overnight stays due to a change from reference climatic conditions (1971-2000) to future climatic condi-tions (2041-2070). Shaded areas and error bars represent the uncertainty resulting from the consideration of four different climate forcings and two different versions of a simple climatic beach index, whereas circles and bars indicate the average of the eight different modelling results.

3.7 Dissemination and Stakeholder Interaction (WP 7)

In order to have an operational dissemination of hydrological modeling results, the WP7 of CLIMB addressed the question of security threats through an analysis of water uses and rivalries in each case study investigated. Stakeholders have been considered into two main categories: water users and water managers. Individual interviews, meetings and implementation of questionnaires have been used in each case study to provide the following results. As far case studies of CLIMB represent a diversity of geographic and political situations around the Mediterranean Basin, we present here the results of all the case studies assuming it represent an assessment of an average situation of water uses within the « Mediterranean Region ».
In order to describe threats on water uses conflicts, results of two main outputs are commented :
- the rating of level of priorities of water uses
- the main pressures on water resources for the past and the next 20 years

3.7.1 The ranking of priority of water uses

Within the different sites, stakeholders have a clear representation of what should be the first priority of water uses: domestic use to supply (Figure 21) inhabitant requirements. However, considering the second level of priority (Figure 22), it seems not so clear if irrigation should be the prior allocation instead of other uses as livestock supply or industrial needs. This representation differs from water managers and water users, water managers considering irrigation and hydropower when there is this activity as the second level of priority. Confusion begins at third level (Figure 23) of water allocation with an extension of the water uses list.
This demonstrates the need to communicate the water management plan and the diversity of water uses within the managed areas to prevent conflicts. In fact, some water users are not aware of the multiplicity of water uses in their water basin as for the origin of the water. This is less true where water governance is really efficient, mainly in the French and in the Gaza sites. However, these tensions between water uses would maybe not occur during severe droughts events where stakeholders would understand the need to supply domestic water use first but probably more during less intense water scarcity events when the existing water would have to be shared within different water uses and water users.

Fig. 21: First priority water uses in the case studies

Fig. 22: Second priority water uses in the case studies

Fig. 23: Third priority water uses in the case studies

3.7.2 Main pressures during the past 20 years and for the next 20 years

It is notable that for the Mediterranean Region represented by CLIMB case studies, the main pressure on water resource during the last 20 years has been linked to population growth and urbanisation even before irrigation (Figure 24). For the near future, stakeholders assume urbanisation will not increase as in the past but consider however an increase of domestic water use. Then, tourism is not considered as a major demand. This latest confirms the difficulty for stakeholders to represent the entire water catchment with, in fact, for all cases, an important tourism pressure due to the interconnection of water resource origin.

Stakeholders are devided for two main features. Firstly for the evolution of water needs for irrigation. Some consider it will increase and others trust it will decrease. In fact, this issue is really site-specific: in some cases agricultural areas can not be extended thus with an assumed stability of water demand, as in the North Italian case study with an entire conversion of apple trees to drop irrigation while agricultural areas can not be extended. For other, as for the Gaza strip, population growth will imply new water demands for irrigation to supply agricultural productions. Secondly, for the evolution of water availability. As for the situation within the past, stakeholders are devided in considering the availabilty of water but with, for the near future, a more common view from water managers that water availability will increase. Contrary to this, both managers and users assess new water resources will be required.

These results are important as they reflect the lack of knowledge from water users of water management solutions implemented. In fact, to answer to the increase of water demand, external water resources have been integrated : in the Sardinian case study, water resource has been politicaly regionalised to confirm one unique water authority and connect the different basins, up to technological constraints limits. In the French case study, to comply with population growth pressure and the salinisation of local water resources, connexion with two external hydrological basins, the Hérault and the Rhône rivers, have been managed. For the Tunisian case study, if there is a local dam collecting runoff waters, water resource is national with major transfers from the North to the South. It is more complex for the Gaza strip and the Nile delta case studies because of geopolitical issues in these transboundary rivers.

Water use for environmental (regulatory) purposes is supposed to increase. This can be mainly attributed to the European sites with the implementation of the WFD and the concept of minimum ecological flow.

The decrease of water quality, not really attributed to the salinisation of water resources, is assumed to be more important for both water managers and water users. This reflects the awareness of stakeholders of the impact of human activities on water quality.

The results have underlined that the terms «climate change» have almost not been cited by stakeholders during both interviews and open questions in the questionnaires. It is emphasised by the fact that the evolution of rainfalls quantity is not considered as an issue for the next 20 years. This confirms the need to continue efforts on disseminating, in relation with local water managers, the state of knowledge on climate change impacts in the Mediterranean Region.

To conclude on the analysis of water uses and water rivalries in the Mediterranean Region represented by 7 case studies with CLIMB, the main answer to the increase of water demand, without considering climate change as a driving force, has been a progressive externalisation of water resource. It seems there is no limit to this extension within national borders and clear limits with international ones.

Figure 24 : Results of the main causes of water use evolution in the case studies of CLIMB according to water managers and water users for the last 20 years (grey) and for the next years (colour)

It has also been spotlighted that all analysed water management plans are mentioning desalinization as an option, for the European case studies as for the other ones. Is seems this represent the next step of water supplying in the Mediterranean Region. This important output raises some questions to water policy:

Are the national and international legislations ready to answer potential issues linked to this new water resource ?

Should this new water resource be included in hydrological modelling for water management planification scenarios ?

The last observation, valid for all sites, is actors' predictions of increases in water needs over time, with desalination and reuse options presented to be intensified.

3.7.3 Conclusions around climate change impacts on water uses in CLIMB case studies

How do stakeholders prepare and plan to adapt themselves to the impacts of climate change on local water uses? From the stakeholders’ point of view, could these changes produce or increase the rivalries on water resources?

The first observation is that for all sites, the interviewees did not mention all water uses within the boundaries of the site. This means that users and in some cases also water managers, do not fully represent the uses of water in the studied regions. At the same time, this lack of awareness could have a substantial effect on decisions, especially in times of water shortage and probably impact competition for water access.

As an example, the multi-sector and comprehensive management practices applied in the Rio Mannu Basin in Sardinia, including interconnection between the various basins, are a pivotal feature and one that creates potential rivalries and conflicts at the regional scale. Whereas previously low water conditions could only be handled using resources available in the basins, the new interconnection system increases the basin's water supply through transfers from regions with more plentiful reserves. While basins with shortages are thus able to access additional water volumes, which serves to reduce internal rivalries surrounding water, the regional scale becomes the stage of other possible rivalries among actors who had not been in competition previously, since they hadn't been sharing resources existing on their territory. Consequently, an interconnection-based management strategy has indeed expanded the water supply, yet by the same occasion the "multi-sector" system composed of hydraulic facilities imposes a multiyear scheduling on authorities that's capable of preventing crisis situations by means of "rationalizing" (managers' parlance) uses, an option that was simply infeasible when these facilities were being supervised either by distinct managers (typical in the wide array of Consorzi di Bonifica entities) or by individual river basin (Tirso, Flumendosa). The crises experienced have un-derscored limitations in the former system, which was highly fragmented and organized along sector-specific lines. As part of this reorganization however, basin hydrological criteria no longer influence water management policy. Regional authorities have adopted a scope of water management that covers the entire regional territory. The exchanges between water industry actors are no longer taking place within the confines of a basin but can extend beyond, in transgressing the hydrographic boundaries. Such exchanges are no longer tied to a physical space and lose their territorial rationale to focus solely on the political arena and the ability of individual actors to impose their project. This situation is identical to that found in Tunisia, where the Chiba Basin also receives water flowing from the North, but for how long will residents and users in the Northern territories accept an outflow of their water? The same configuration applies to the Thau Lagoon, which just began importing Rhone River water via the Aquadomitia project.

The third observation pertains to handling uncertainties and the risk associated with water supply interruptions. During decision-making processes, especially when a shortage hits, authorities must choose between two competing risk-taking strategies and two perceptions of risk. The first approach stems from the manager, for whom risk-taking (as in whether or not to use resources presently stored to absorb a future drought risk) is associated with measuring tomorrow's potential damage (relative to urban uses) as regards urban uses. The second approach is offered by agricultural producers, for whom the potential risk of a future lack of water motivates guaranteeing the full season's production, which thus rejects the notion of rationing present stored volumes. These various perceptions of risk may cause situations of conflict, particularly in the context of climate change.

Given these considerations, the presence of artificial reservoirs in the region (Rio Mannu, Noce, Chiba) provides for a multi-year regulation of natural water supply through successfully managing this uncertainty over time. The mere presence of these reservoirs however is not sufficient to effectively manage the variability in natural water inflows. The periodic scheduling of water regulation in reservoirs is a key parameter for coping with climate-related uncertainties. As a case in point, performing such regulation results from a combination of several factors, involving the relative ratio between volumes allocated to various uses, reservoir reloading time, the temporal variability (i.e. from one reservoir to another) in natural inflows, and the geographic locations of all such inflows. These factors need to be cumulated alongside a set of social factors associated with water uses, including the procedure and techniques for generating agricultural water, the relationship between service users and the water resource, crop choices, etc.).

A fourth observation pertains to the effects of tourist activities on three study sites, namely Noce, Thau and Chiba. The real challenges are correlated with the needs of such use, i.e. winter tourism for Noce and beach stays for the other two sites. In the former case, the most minimal snowfall leads to greater reliance on artificial production by means of snow blowers, which implies an additional use of water at a time of rising demand for drinking water in the basin given the presence of tourists and holiday-makers. Moreover, a shorter season combined with the need to climb higher to practice winter sports might represent a revenue loss relative to this economic activity and thereby stimulate an intensification of other activities, notably agriculture (which dips into basin water and is at times carried out by the same set of actors in a solidarity move). Recreation needs in connection with summertime tourism activities and beach resorts (swimming pools, hotels, etc.) exert considerable pressure on urban water uses. Such pressure further complicates resource use since water is being supplied via the public network. As a final point, the economic activities related to these recreational uses offer substantial economic growth potential.

3.7.4 Conclusions on the Climb dissemination process

This process of using interactions with stakeholders to disseminate CLIMB outputs was interesting, considering that scientists are comfortable with uncertainties and assess it, but could be uncomfortable to disseminate it, while stakeholders at local level are pragmatic and capable to deal day-to-day with uncertainties.
The qualitative analyse of water uses in both current and through climates changes scenarios situations permitted to provide lots of information. At the meantime, the process was very time consuming, even for case study leaders. Nevertheless, the process answered its objectives by completing the representation of the case study by case study leaders and by increasing its local area network. This has been confirmed at the occasion of the final dissemination meetings when they took place in the case studies.
This situation of reciprocity contributed to a real interface of learning through a continuous process. It should permit, locally, to avoid maladaptation and decrease vulnerability. The CLIMB method permitted to transfer knowledge between entities, in fact, using completely different languages:

Italian, French, Arabic, Turk, English, German,
Cultural differences and different perceptions of interactions ways!

The transfer of theoretical and empirical knowledge has not been only top down but also bottom up:

Addressed to scientists and stakeholders (managers and users) and translate “knowledge” in “know-how” to both,
Combine time scales usually disconnected: creation of research outputs and its use by end-users.

This has been possible because of:

Interdisciplinary teams, as in particular the WP7 leader with geographers, hydrologists, social anthropologists, public policy scientists and political scientists,

Dissemination cells with the local pivote role of a case study leader whose interest is to keep a local dialogue with stakeholders,

Support from the European Commission in granting substantial importance to dissemination processes.

Potential Impact:
4 Potential Impact and Main Dissemination Activities

4.1 Potential Impact

The strategic impact of the CLIMB project is fully in line with the research and dissemination requirements described in the objectives of the Environment Theme under FP7. Being a fully integrated approach to understand, analyze and predict climate change and its impacts, the project was specifically dedicated to the focal areas described in sub-activity 6.1.1 (Pressures on Environment and Climate) for 2009. CLIMB’s work plan was targeted towards a substantial advancement of process knowledge and modeling capabilities for a better understanding of the interactions between the biosphere, ecosystems and human activities and thus to better assess climate effects on water resources and uses. The project combined genuine science activities with a strong link to practical application in the targeted regions of the Mediterranean area and thus provided a balance between the three building blocks of environmental research, namely understanding, assessing impact and responding to threats to security in man-environment systems.

An increase in general knowledge of water management issues in (semi-)arid climate, that can be applicable to some EU areas more prone to changes brought about by climate change and/or global warming, led to the development of innovative practical and/or theoretical approaches and technologies in environmental monitoring and environmental modeling.

Taking into account the latest advancements in the field of climate change impacts on the environment, the development of new technologies was focused on the provision of new monitoring systems and modeling tools to significantly reduce uncertainties of climate change impacts on the hydrology in the specified regions as outlined in Call ENV2009- The systematic approach and the resulting tools, in particular the CLIMBPortal, is internationally recognized and is expected to be widely adopted by following research activities and operational water resources managers for the development of sound and sustainable adaptation measures to counteract adverse effects of climate change.

Expertise was intensively shared and exchanged with numerous other FP7-projects, such as DEWFORA, DROUGHT-R&SPI and particularly WASSERMed and CLICO within the CLIWASEC research cluster, but also with larger initiatives, such as the Mediterranean Water Scarcity and Drought Working Group (MED-EUWI) or the Mediterranean branch of the Global Water Partnership (GWP-Med). This is considered very beneficial to provide supportive guidance for a more concise implementation process for current water-related directives, such as the EU Water Framework Directive (2000/60/EC) or the EU-Flood Risk Management Directive (2007/60/EC).

More specifically, the work in CLIMB provided a Summary for Policymakers comprising the final results of the CLIWASEC cluster and a Special Report on model comparison from two of the three projects. In its effort to grant easy-access to data and results from the project, CLIMB developed a WebGIS-Server and Client architecture open to the public. It disseminates the impacts of climate change on selected hydrological indicators, including a rigorous assessment of related uncertainties, as determined from the multi-model ensembles employed in the seven case studies. Further, it comprises a risk modeling tool, assessing the value-at-risk due to water shortages in agriculture and the tourism sector, based on the identification of key socio-economic indicators. Site-specific adaptive measures are proposed and recommendations for future water resources management are given, taking into account a thorough diagnosis of climate change impacts on water uses and rivalries. It is expected that CLIMB results can be regionalized in general for water-stressed areas, in which climate and socioeconomic conditions render water-related problems compelling and urgent. This can happen in various ways to:

- foster and intensify the dialogue between scientists, managers, water experts and stakeholders in addressing local impacts of climate changes and identifying means for their assessments

- awareness among stakeholders about climate change impacts on water resources and land uses, which will lead to adequate approaches and adaptation strategies for water resources management and for food security

- empower stakeholders and scientists by providing new tools of decisions making in assessing climate change impacts

These science-management-policy links are indispensable to provide visibility of the research findings beyond the borders of the scientific community and will allow for an uptake of research results into policy and management practice. The diversity of study sites in CLIMB supports additional benefits for the development and implementation of adaptation measures, as larger scale stakeholder networks can develop when commonalities in problems and problem prevention can be addressed and mutually discussed through the respective connection to the CLIMB Consortium, which provided a forum to discuss the implications of the scientific results. An important output of the research in the individual study sites was the development of a set of recommendations for an improved monitoring and modeling strategy for climate change impact assessment, addressing in particular the minimum requirements towards data collection and model complexity to achieve sufficient predictive power for climate change impact assessment in the targeted regions and beyond.

4.2 Main Dissemination Activities

The CLIMB project promoted practical application of new strategies to assess climate induced changes on the hydrology of Southern Europe and neighboring countries and provides the transfer of existing and emerging knowledge in climate change impact research to different stakeholders. The main goal was the maximum use of project results by addressing researchers, policy makers, decision makers etc. The main target groups are scientists (universities, research organizations), commercial organizations (SMEs, industry), stakeholders (public authorities/organizations, NGOs), and politicians. For a complete list of all dissemination activities, please refer to section 6 (Table 6.2) of this report.

The CLIMB results have been exploited and disseminated inside and outside the consortium, com-posed of representatives from the academic world (universities, research institutes), SMEs, and non-profit organizations:

- Partners with the relevant knowledge in: climate change impacts, hydrological modeling, risk assessment, economic assessment etc.

- Different types of organizations: universities, research institutions, SMEs, and non-profit or-ganizations

- Different participating countries in CLIMB (European member states, International Cooperation partner countries, Canada)

Within the scientific community the CLIMB results have been distributed via publications in peer-reviewed scientific journals, the Special Sessions at internationally renowned scientific conferences and conjoint publications of Special Issues in appropriate cross-cutting scientific journals with WASSERMed and CLICO as partner projects in the research cluster CLIWASEC. For the latter, it is planned to establish an own Special Issue on ‘Climate, Water and Security in the Mediterranean and neighboring regions’ in the highly-ranked scientific journal STOTEN (‘Science of the Total Environment’).

Interaction with stakeholders and potential end-users

The dissemination of project related awareness, knowledge in a transparent, easy-to-digest and user-friendly way and impact on the utmost important level of stakeholders and potential end-users is challenging and requires special attention. CLIMB maintained close connection to stakeholders in all case study sites by means of regular consultations, to catalyse dissemination and application of the results. The main objective was to convince stakeholders and end-users of the feasibility and usefulness of the CLIMB project results:

Creation of a strong network on water resources issues in the Mediterranean region (e.g. via frequent local workshops). Workshops, for end-users and developers, offered opportunities to the project to gather feedback from the target groups and address the needs of the stakeholders involved (09.07.2010 + 14.09.2010; 17.06.2011; 24.02.2013: Thau / France; 01.06.2010 + 09.12.2012: Cap Bon / Tunisia; 05.10.2010: Noce / Italy; 01.02.2011 + 04.02.2011 + 02.04.2012: Sardinia, Italy; 11.11.2012: Senorbì / Italy; 14.02.2010 + 14.03.2010 + 01.02.2012: Al Gharbiyah / Egypt)

Improvements of models, new assessment tools have been communicated to stakeholders and decision makers in a form that enabled an easy utilization of the new findings in regional water resource and agricultural management initiatives as well as in the design of mechanisms to reduce potential for conflict.

Arrangement of national meetings focused on the demonstration of concepts, tools and results for users and policy makers to discuss and consequently improve the potential of these tools to support decision making at larger spatial scales.

Special design of CLIMBs annual general assembly’s (Cairo, Cagliari, Munich, Istanbul, Brussels) that included meetings, workshops and excursions for stakeholders and end-users, for them to better famil-iarize with the concepts and techniques developed and applied and to strengthen the science-policy interfaces. CLIMB aimed to capture decision-makers interest by providing constructive and substantial recommendations, elaborated by means of direct communica¬tions with regional stakeholders or in response to the thematic priorities issued by the EC.

Fig. 25: CLIMB meetings with stakeholders and organization of General Assemblies

Dissemination materials and tools

For the promotion of the CLIMB-project and its activities several digital tools have been used (public accessible website: frequently updated electronic study site-newsletters, press releases etc.) to support the awareness rising on the project and the exploitation of project results. The CLIMB website is the primary dissemination route for the presentation of the project, in addition to the online provision of press releases. The website is structured in different areas with respective subfolders (News & Events, Project, Partners, Media, Links, Dissemination, and Publications). CLIMB has also set up and maintained the CLIWASEC cluster webportal on climate change, water and security ( providing links to websites of the CLIMB, WASSERMed and CLICO projects.

CLIMB leaflet and CLIWASEC leaflet

A CLIMB leaflet in English and the languages which have been represented in the project via the participating countries (Arabic, French, German, Italian, and Turkish) included the general information about the CLIMB project (motivation, objectives, outreach, the structure, the partners, the study sites and the contact information).

Fig. 26: The CLIMB leaflet

Also in the research cluster CLIWASEC where CLIMB has been cooperating with the other FP7-projects WASSERMed and CLICO a leaflet has been designed and distributed for dissemination purposes.

Fig. 27: The CLUSTER leaflet (version 1, 2010)


Posters with the general scope of CLIMB, the outcomes of the several work-packages as well as the challenges in the different study sites have been designed and regularly updated to deal with the specific scope of respective events.

Fig. 28: CLIMB posters with general scope (left) and specific scope (right)

During several occasions, those CLIMB posters have been presented at scientific conferences, trade fairs, researcher´s nights, stakeholder meetings, General Assemblies etc. Therefore different inter-sectoral target groups in all countries with CLIMB partners have been reached (science, industry, policy, decision makers, civil society)

26.04.2010: Integrated River Basin Management under the Water Framework Directive (Lille / France)
10.06.2010: 7th Intern. Recycling Environmental Technologies and Waste Management Trade Fair (Istanbul / Turkey)
24.09.2010: La nuit des chercheurs (Angers / France)
02.02.2011: CLIWASEC-stakeholder meeting (Cagliari / Italy)
05.04.2011 22.04.2012 11.04.2013: EGU General Assembly (Vienna / Austria)
30.08.2011: Pedometrics conference (Prag / Czech Republic)
13.02.2012: CLIMB General Assembly (Munich / Germany)
22.04.2012: EGU General Assembly (Vienna / Austria)
01.11.2012: INNOVATION TURKIYE EXPO (Istanbul / Turkey)
08.10.2013: GlobalSoilMap conference 2013 (Orléans, France)

Press releases and articles to attract and maintain media attention

CLIMB disseminated the project results and ongoing activities also via public media, newspapers and magazines.

Fig. 29: Examples of press releases during the running period of CLIMB (11.01.2010; 15.02.2012; 21.11.2013)

Fig. 30: Impact of CLIMB in science-oriented and popular press

CLIMB videos

CLIMB produced also project-videos which have been used during stakeholder Workshops in the different case study sites. Those videos can be achieved via CLIMBs public website (
In addition, the video “Water in Africa in a changing climate” has been produced in 2013 by Séverine Dieudonné with the representation of CLIMBs scientific co-ordinator Prof. Ralf Ludwig as well as CLIMBs Nile delta case study leader Prof. Badr Mabrouk.

Fig. 31: Video “Water in Africa in a changing climate” on YouTube

Fig. 32: CLIMB in public TV stations (Left: General Assembly in Cagliari/Italy 2011, centre: Conference “Environmental Protection is a must in Alexandria / Egypt 2010; right: International Water Technology Conference in Alexandria / Egypt 2011)

In total, the distribution of the more than 450 registered dissemination activities during the runtime of CLMB is displayed in Figures 33 and 34. Accordingly, scientific conferences (38%) rank first among dissemination activities, followed by meetings & workshops (29%), PR-material (23%) and scientific publications (10%). 36% of all registered activities are covered by the coordination unit (22% BayFOR, 14% LMU).

Fig. 33: Distribution of dissemination activities during the CLIMB runtime

Fig. 34: Distribution of dissemination by partner during the CLIMB runtime

List of Websites:
5 Website and Contacts

5.1 Website

The CLIMB project website has been initiated in early 2010. presents the basic ideas behind CLIMB, informs about the actual status and latest news of the project and provides information about the research team and plenty of Downloadables.

Fig. 35: The CLIMB homepage (

5.2 Logo

The creation of the project’s image started with the design of a distinctive logo. This logo has been included in the design and production of the CLIMB website, the leaflet, as well as in all the inter-nal and external communication material produced by the consortium.

Fig. 36: CLIMB-Logo

5.3 Contact Details

The Coordination Team

Prof. Dr. Ralf Ludwig (Project coordinator)
Department of Geography
Ludwig-Maximilians-Universitaet Muenchen
Luisenstr. 37, 80333 Munich, Germany
Tel. +49-89-2180-6677, Fax +49-89-2180-17858

Dr. Thomas Ammerl (Project manager)
Bavarian Research Alliance GmbH
Prinzregentenstrasse 52, 80538 Munich
Tel. +49-89-9901888120, Fax +49-89-9901888-29;