Periodic Reporting for period 4 - DRY-2-DRY (Do droughts self-propagate and self-intensify?)
Periodo di rendicontazione: 2021-08-01 al 2022-07-31
Sintesi del contesto e degli obiettivi generali del progetto
Climate change and population growth are great threats with joint impacts on hydrology and ecosystems. Droughts, unlike other natural hazards, develop slowly and can affect extensive areas for prolonged periods of time. This leads to devastating environmental impacts and losses of human life due to famine and dehydration. Globally, droughts and tropical cyclones tie as number one cause of death by natural disaster. In Europe, however, no natural hazard has caused in recent years as many deaths as heatwaves. For instance, the European mega-heatwaves of 2003 in Western Europe and 2010 in Russia amounted together to 120,000 reported lives, i.e. 85% of all deaths caused by climate disasters in the continent between 1970–2012. Plus, recent heatwaves have frequently coincided with soil droughts, leading to compound impacts.
The overarching goal of DRY–2–DRY was to help mitigate the devastating impacts of droughts and heatwaves by furthering our understanding of the physical processes triggering their intensification and propagation. The main hypothesis was that, as a soil drought strikes, the reduced evaporation contributes to the drying and warming of the air. This leads to further expansion and concatenation of drought and heatwave conditions in the downwind direction. This implies that droughts and heatwaves can 'self-propagate' and 'self-intensify'. Just like a wildfire. Along the journey to test this hypothesis, we unravelled the effect of vegetation and soil moisture on the circulation of heat and moisture in the planet, and ultimately on temperature and precipitation. The potential of land cover change and management as a climate adaptation and mitigation strategy was also explored.
Altogether, DRY–2–DRY activities helped uncover a physical process that is behind the occurrence of many heatwave and drought events in subtropical regions and mid-latitudes. By doing so, the project contributed to a community effort to improve future predictions and early forecasts of these extreme events in order to attenuate their devastating impacts they have on ecosystems and societies.
The overarching goal of DRY–2–DRY was to help mitigate the devastating impacts of droughts and heatwaves by furthering our understanding of the physical processes triggering their intensification and propagation. The main hypothesis was that, as a soil drought strikes, the reduced evaporation contributes to the drying and warming of the air. This leads to further expansion and concatenation of drought and heatwave conditions in the downwind direction. This implies that droughts and heatwaves can 'self-propagate' and 'self-intensify'. Just like a wildfire. Along the journey to test this hypothesis, we unravelled the effect of vegetation and soil moisture on the circulation of heat and moisture in the planet, and ultimately on temperature and precipitation. The potential of land cover change and management as a climate adaptation and mitigation strategy was also explored.
Altogether, DRY–2–DRY activities helped uncover a physical process that is behind the occurrence of many heatwave and drought events in subtropical regions and mid-latitudes. By doing so, the project contributed to a community effort to improve future predictions and early forecasts of these extreme events in order to attenuate their devastating impacts they have on ecosystems and societies.
Lavoro eseguito dall’inizio del progetto fino alla fine del periodo coperto dalla relazione e principali risultati finora ottenuti
a – Improvement of agricultural and water management using satellite data
Early activities within DRY–2–DRY demonstrated that state-of-the-art satellite datasets of evaporation where insufficient to accomplish the first objective of providing evidence of the impacts of drought and heatwaves on terrestrial evaporation. That led to a large time investment to improve the GLEAM model (www.gleam.eu) aiming to better capture plant transpiration during conditions of stress. That required the application of the framework at increasingly higher resolutions by utilising data from the Sentinel constellation at 1 km resolution over continental scales, and at 100 m over regional scales. A longstanding collaboration with VanderSat was established to foster the development of GLEAM at hyper-resolution, in order to turn it into a very valuable asset for water management and agricultural irrigation planning. Spin-off activities included the collaboration with the European Space Agency (ESA) in several projects that feed into the EU Green Deal, the EU Data Strategy and the EU Destination Earth (DestinE), directed towards the modelling of a high-resolution digital replica of the Earth's hydrological cycle based on satellite observations.
b – Enabling a better representation of land processes in climate models
A central activity within DRY–2–DRY was to improve global evaporation and soil moisture data to benchmark climate model representation of land surface processes. The accurate model representation of evaporation is critical to predict droughts and heatwaves. Data produced during the project was used in the IPCC AR6 for the purpose of climate model benchmarking. The commitment to refining global satellite products for their use as benchmark in climate models led the DRY–2–DRY team to explore hybrid modelling, i.e. the combined use of process-based and machine learning algorithms. This demonstrated great potential to better understand land surface processes at global scales, including evaporation, soil moisture and plant water stress. Therefore, the project heralded an alternative way to study these variables and represent them in climate models. Likewise, earlier research within the project also demonstrated that novel satellite observations (e.g. of sun-induced fluorescence) can be used to estimate plant use of water. These lines of work leave a legacy beyond DRY–2–DRY. Updated datasets of these variables will continue being made available at www.gleam.eu which counted on more than 8,000 unique data users during the lifespan of DRY–2–DRY.
c – Uncovering whether the global water cycle is accelerating
A clear expectation derived from climate model simulations and basic thermodynamic principles is that precipitation will increase globally as the planet warms. The term 'water cycle acceleration' has been coined to refer to this process. This is not a homogenous response at regional scales, as it is conditioned on the ability of the land surface to evaporate more water as temperature rises. In fact, this is not always the case, and most dry regions are expected to become drier and receive less precipitation in the future. Yet globally, our research unequivocally shows that land evaporation will monotonically increase following global warming trends. Since 2015, our contributions to the annual State of the Climate Report from the Bulletin of the American Meteorological Society (BAMS) have confirmed these earlier findings. These recursive publications using data produced during DRY–2–DRY have consistently demonstrated that El Niño-driven droughts cause negative anomalies in global evaporation because they trigger droughts in the Southern Hemisphere, but decadal trends still follow global warming. This is an important finding that implies that the future water availability over the continents will be highly sensitive to greenhouse emissions, and the trends in the intensity of El Niño–La Niña cycles. The P.I.'s involvement within the evaporation scientific community led to a role as board member in the Global Climate Observing System (GCOS), where he is responsible for the 'Land Evaporation' Essential Climate Variable (ECV).
d – Better understanding the lethality of drought–heatwave compound events
Recent heatwaves frequently coincided with soil droughts, largely due to the feedbacks illustrated in Fig. 1. While this synchronisation with soil drought acts to intensify air temperature, it also lowers air humidity. Using the novel HAMSTER software tool developed during DRY–2–DRY, it was demonstrated that these compound dry–hot conditions are highly detrimental for ecosystem primary productivity as they lead to severe vegetation stress. However, for humans, the situation is different: the lowering of air humidity due to soil drought reduces heat stress, instead of increasing it as it occurs for plants. To understand the net impact of soil desiccation on the human morbidity and mortality of heatwaves, another software tool, called CLASS4GL, was developed within DRY–2–DRY. CLASS4GL combines weather balloon data, satellite observations and atmospheric modelling. The tool was used to disentangle the effects of land feedbacks on the local likelihood of rainfall during droughts and heatwaves. Then, combined with meta-analyses of morbidity and mortality, it was demonstrated that soil drought actually leads to a mild reduction in human heatwave lethality. This finding questioned the efficiency of traditionally proposed mitigation strategies involving surface moistening to increase evaporative cooling, and it gave rise to an interdisciplinary research line to explore the impact of drought on human physiology during heat strokes.
e – Predicting whether droughts will become more severe in the future
The DRY–2–DRY work also focused on future dry extremes, not just historical events. Work during the last stages of the project demonstrated that future drought duration is underestimated by the current generation of climate models. The relationship between the Coupled Model Intercomparison Project (CMIP5 and CMIP6) past estimates and future projections of meteorological drought duration was contrasted with past satellite observations, using what is often referred to as an 'emergent constraint'. Results showed that the 21st century global increase in dry-spell duration will be significantly larger than predicted by the CMIP5 and CMIP6 model ensembles. These findings revealed extensive world regions where global warming will cause meteorological drought aggravation, and they emphasised the importance of reducing climate model errors in precipitation, cloud formation and land–atmosphere feedbacks.
f – Revealing whether droughts and heatwaves can self-propagate
A large part of DRY–2–DRY was dedicated to unravelling the effects of land feedbacks on the expansion and concatenation of droughts and heatwaves. Results demonstrated that the paradigm proposed in DRY–2–DRY was indeed valid: droughts and heatwaves do self-propagate. As a soil drought strikes, evaporation is reduced, which contributes to the drying but also the warming of the air. Moreover, the atmosphere reacts in such a manner that it allows the accumulated heat and dry air to persist from day to day. As that warm and dry air flows downwind, it creates a propagation front, leading to the expansion and concatenation of drought conditions. This propagation is fuelled by the soil dryness caused by the drought itself; similar to the propagation of a wildfire. This process of drought 'self-propagation' is especially important in drylands, where it can account for ~20–30% of the precipitation deficits causing droughts – see Fig. 2 (Schumacher et al., 2022). We also demonstrated that this self-propagation applies to heatwaves as well: as heatwaves dry the surface, the enhanced warming over the land is transported downwind leading to their expansion and even to a concatenation of multiple heatwave events – see Fig. 3 (Schumacher et al. 2019). During recent European mega-heatwaves, over 30% of the heat anomaly was caused through this process.
g – Challenging the notion of 'watershed' as management unit in hydrology
Watersheds (i.e. hydrological catchments) are tightly interconnected via atmospheric water transfers as shown during DRY–2–DRY. This contradicts the traditional views of water resources and water scarcity as local aspects for which watershed‐based management appears adequate. The truth is that precipitation, as a main source of freshwater, depends on moisture supplied through land evaporation from outside the watershed, sometimes thousands of kilometres away. This notion of evaporation as a remote supply to local precipitation is typically not considered in hydrological water assessments. The work within DRY–2–DRY showed, for instance, that up to ¾ of summer precipitation over European watersheds depends on moisture supplied from other watersheds. In other words: watersheds are not autarkic systems. This emphasized the need for global water governance to secure freshwater availability. It also highlighted the role of upwind watershed evaporation for drought occurrence in downwind watersheds. Overall, this has far-reaching implications: the measures that are currently taken in the management of local land and water resources have quantifiable repercussions on water availability in remote regions, which requires consensual decisions, not only between different basins, but also between countries.
Early activities within DRY–2–DRY demonstrated that state-of-the-art satellite datasets of evaporation where insufficient to accomplish the first objective of providing evidence of the impacts of drought and heatwaves on terrestrial evaporation. That led to a large time investment to improve the GLEAM model (www.gleam.eu) aiming to better capture plant transpiration during conditions of stress. That required the application of the framework at increasingly higher resolutions by utilising data from the Sentinel constellation at 1 km resolution over continental scales, and at 100 m over regional scales. A longstanding collaboration with VanderSat was established to foster the development of GLEAM at hyper-resolution, in order to turn it into a very valuable asset for water management and agricultural irrigation planning. Spin-off activities included the collaboration with the European Space Agency (ESA) in several projects that feed into the EU Green Deal, the EU Data Strategy and the EU Destination Earth (DestinE), directed towards the modelling of a high-resolution digital replica of the Earth's hydrological cycle based on satellite observations.
b – Enabling a better representation of land processes in climate models
A central activity within DRY–2–DRY was to improve global evaporation and soil moisture data to benchmark climate model representation of land surface processes. The accurate model representation of evaporation is critical to predict droughts and heatwaves. Data produced during the project was used in the IPCC AR6 for the purpose of climate model benchmarking. The commitment to refining global satellite products for their use as benchmark in climate models led the DRY–2–DRY team to explore hybrid modelling, i.e. the combined use of process-based and machine learning algorithms. This demonstrated great potential to better understand land surface processes at global scales, including evaporation, soil moisture and plant water stress. Therefore, the project heralded an alternative way to study these variables and represent them in climate models. Likewise, earlier research within the project also demonstrated that novel satellite observations (e.g. of sun-induced fluorescence) can be used to estimate plant use of water. These lines of work leave a legacy beyond DRY–2–DRY. Updated datasets of these variables will continue being made available at www.gleam.eu which counted on more than 8,000 unique data users during the lifespan of DRY–2–DRY.
c – Uncovering whether the global water cycle is accelerating
A clear expectation derived from climate model simulations and basic thermodynamic principles is that precipitation will increase globally as the planet warms. The term 'water cycle acceleration' has been coined to refer to this process. This is not a homogenous response at regional scales, as it is conditioned on the ability of the land surface to evaporate more water as temperature rises. In fact, this is not always the case, and most dry regions are expected to become drier and receive less precipitation in the future. Yet globally, our research unequivocally shows that land evaporation will monotonically increase following global warming trends. Since 2015, our contributions to the annual State of the Climate Report from the Bulletin of the American Meteorological Society (BAMS) have confirmed these earlier findings. These recursive publications using data produced during DRY–2–DRY have consistently demonstrated that El Niño-driven droughts cause negative anomalies in global evaporation because they trigger droughts in the Southern Hemisphere, but decadal trends still follow global warming. This is an important finding that implies that the future water availability over the continents will be highly sensitive to greenhouse emissions, and the trends in the intensity of El Niño–La Niña cycles. The P.I.'s involvement within the evaporation scientific community led to a role as board member in the Global Climate Observing System (GCOS), where he is responsible for the 'Land Evaporation' Essential Climate Variable (ECV).
d – Better understanding the lethality of drought–heatwave compound events
Recent heatwaves frequently coincided with soil droughts, largely due to the feedbacks illustrated in Fig. 1. While this synchronisation with soil drought acts to intensify air temperature, it also lowers air humidity. Using the novel HAMSTER software tool developed during DRY–2–DRY, it was demonstrated that these compound dry–hot conditions are highly detrimental for ecosystem primary productivity as they lead to severe vegetation stress. However, for humans, the situation is different: the lowering of air humidity due to soil drought reduces heat stress, instead of increasing it as it occurs for plants. To understand the net impact of soil desiccation on the human morbidity and mortality of heatwaves, another software tool, called CLASS4GL, was developed within DRY–2–DRY. CLASS4GL combines weather balloon data, satellite observations and atmospheric modelling. The tool was used to disentangle the effects of land feedbacks on the local likelihood of rainfall during droughts and heatwaves. Then, combined with meta-analyses of morbidity and mortality, it was demonstrated that soil drought actually leads to a mild reduction in human heatwave lethality. This finding questioned the efficiency of traditionally proposed mitigation strategies involving surface moistening to increase evaporative cooling, and it gave rise to an interdisciplinary research line to explore the impact of drought on human physiology during heat strokes.
e – Predicting whether droughts will become more severe in the future
The DRY–2–DRY work also focused on future dry extremes, not just historical events. Work during the last stages of the project demonstrated that future drought duration is underestimated by the current generation of climate models. The relationship between the Coupled Model Intercomparison Project (CMIP5 and CMIP6) past estimates and future projections of meteorological drought duration was contrasted with past satellite observations, using what is often referred to as an 'emergent constraint'. Results showed that the 21st century global increase in dry-spell duration will be significantly larger than predicted by the CMIP5 and CMIP6 model ensembles. These findings revealed extensive world regions where global warming will cause meteorological drought aggravation, and they emphasised the importance of reducing climate model errors in precipitation, cloud formation and land–atmosphere feedbacks.
f – Revealing whether droughts and heatwaves can self-propagate
A large part of DRY–2–DRY was dedicated to unravelling the effects of land feedbacks on the expansion and concatenation of droughts and heatwaves. Results demonstrated that the paradigm proposed in DRY–2–DRY was indeed valid: droughts and heatwaves do self-propagate. As a soil drought strikes, evaporation is reduced, which contributes to the drying but also the warming of the air. Moreover, the atmosphere reacts in such a manner that it allows the accumulated heat and dry air to persist from day to day. As that warm and dry air flows downwind, it creates a propagation front, leading to the expansion and concatenation of drought conditions. This propagation is fuelled by the soil dryness caused by the drought itself; similar to the propagation of a wildfire. This process of drought 'self-propagation' is especially important in drylands, where it can account for ~20–30% of the precipitation deficits causing droughts – see Fig. 2 (Schumacher et al., 2022). We also demonstrated that this self-propagation applies to heatwaves as well: as heatwaves dry the surface, the enhanced warming over the land is transported downwind leading to their expansion and even to a concatenation of multiple heatwave events – see Fig. 3 (Schumacher et al. 2019). During recent European mega-heatwaves, over 30% of the heat anomaly was caused through this process.
g – Challenging the notion of 'watershed' as management unit in hydrology
Watersheds (i.e. hydrological catchments) are tightly interconnected via atmospheric water transfers as shown during DRY–2–DRY. This contradicts the traditional views of water resources and water scarcity as local aspects for which watershed‐based management appears adequate. The truth is that precipitation, as a main source of freshwater, depends on moisture supplied through land evaporation from outside the watershed, sometimes thousands of kilometres away. This notion of evaporation as a remote supply to local precipitation is typically not considered in hydrological water assessments. The work within DRY–2–DRY showed, for instance, that up to ¾ of summer precipitation over European watersheds depends on moisture supplied from other watersheds. In other words: watersheds are not autarkic systems. This emphasized the need for global water governance to secure freshwater availability. It also highlighted the role of upwind watershed evaporation for drought occurrence in downwind watersheds. Overall, this has far-reaching implications: the measures that are currently taken in the management of local land and water resources have quantifiable repercussions on water availability in remote regions, which requires consensual decisions, not only between different basins, but also between countries.
Progressi oltre lo stato dell’arte e potenziale impatto previsto (incluso l’impatto socioeconomico e le implicazioni sociali più ampie del progetto fino ad ora)
WP1 – Impact of droughts and heatwaves on evaporation
Still today, the lack of accurate historical datasets of key variables, such as evaporation or soil moisture, prevents reliable hydrological and agricultural monitoring. This hinders decision-making in water resource management, and it is a hurdle when it comes to understanding the influence of droughts and heatwaves on water resources. Project efforts to explore novel means to retrieve land surface states and fluxes based on satellite observations led to (a) high-resolution datasets that are unprecedented in their breath and scope, (b) the use of satellite observations like SIF that were still untapped in hydrology, and (c) novel approaches making use of artificial intelligence that at the start of the project were still inexistent. Therefore, DRY–2–DRY continually pushed the frontiers of the field of satellite-based evaporation science. Improving agricultural and water management was a natural consequence of this work that was enabled thanks to the collaboration with VanderSat. Likewise, model benchmarking activities using data produced during the project allowed to identify biases in state-of-the-art IPCC climate models (AR6), and to better understand the influence of climate and climate change on the global water cycle. This line of work, aiming for higher resolution accurate evaporation data, will remain a priority in the research team in the future.
WP2 – Local feedbacks and event self-intensification
Long-standing questions on the importance of land feedbacks and drought self-intensification were answered during the project. The proposed methodology, which combined satellite observations, in situ data and mechanistic models of the lower atmosphere, was extended and culminated in CLASS4GL. The approach showed great promise to scrutinise land feedbacks during extreme events, and enabled the systematic study of the influence of drought–heatwave synchronisation on the morbidity and mortality associated to heat stress. Unlike for plants, the lowering of air humidity due to soil drought was shown to reduced heat stress, instead of aggravating it. This net beneficial impact of soil desiccation on human morbidity and mortality during heatwaves was unexpected. These findings suggested, for instance, that large-scale irrigated agriculture may intensify lethal heat stress instead of reducing it, and it questioned the efficiency of traditionally proposed mitigation strategies involving surface moistening to increase evaporative cooling.
WP3 – Teleconnected feedbacks and event self-propagation
Back in 2017, mesoscale and teleconnected feedbacks from soil moisture and vegetation on rainfall remained largely understudied. Most existing work had centred on local feedbacks or moisture transport within the continents, but outside the context of droughts and heatwaves. The drought and heatwave self-propagation (Fig. 3, 4) – caused by the expansion and concatenation of events as evaporation declines – was a fully novel concept that was postulated for the first time in the DRY–2–DRY proposal. Years later, several high-impact articles clearly demonstrated that, as the air is warmed and dried during a soil drought, it creates a propagation front in the downwind direction. This 'self-propagation' is especially important for drought events in dryland, but applies to heatwaves in Europe as well. An important implication is that upwind land conditions can be used to predict upcoming droughts and heatwaves in regions downwind, probably weeks ahead. This predictability will be explored and exploited in future activities within the host research team.
WP4 – Land cover and land management influence
The impact of vegetation and land-cover change on catchment-scale water resources during dry seasons has long been explored, but most land cover management strategies in the con-text of geo-engineering have targeted afforestation experiments and crop selection to increase terrestrial carbon sinks or surface albedo. The influence of land cover changes and management on the occurrence of local or remote droughts and heatwaves remains understudied. The work in DRY–2–DRY to study atmospheric connections among watersheds led to important findings and implications regarding the need for a more holistic water governance. Conclusions contradicted the traditional views of water resources and water scarcity as local aspects for which watershed‐based management appears adequate, and highlighted the fact that precipitation depends on moisture supplied through land evaporation from thousands of kilometres away. Upwind land and water management can influence drought occurrence in downwind watersheds. This novel perspective strives to provide a more integral management of global water resources, while advocating for the consideration of inter-watershed transfers through the atmosphere as a key water resource. This view was further expanded to demonstrate the importance of plant transpiration for precipitation anomalies in African watersheds, using the HAMSTER tool developed during DRY–2–DRY.
Still today, the lack of accurate historical datasets of key variables, such as evaporation or soil moisture, prevents reliable hydrological and agricultural monitoring. This hinders decision-making in water resource management, and it is a hurdle when it comes to understanding the influence of droughts and heatwaves on water resources. Project efforts to explore novel means to retrieve land surface states and fluxes based on satellite observations led to (a) high-resolution datasets that are unprecedented in their breath and scope, (b) the use of satellite observations like SIF that were still untapped in hydrology, and (c) novel approaches making use of artificial intelligence that at the start of the project were still inexistent. Therefore, DRY–2–DRY continually pushed the frontiers of the field of satellite-based evaporation science. Improving agricultural and water management was a natural consequence of this work that was enabled thanks to the collaboration with VanderSat. Likewise, model benchmarking activities using data produced during the project allowed to identify biases in state-of-the-art IPCC climate models (AR6), and to better understand the influence of climate and climate change on the global water cycle. This line of work, aiming for higher resolution accurate evaporation data, will remain a priority in the research team in the future.
WP2 – Local feedbacks and event self-intensification
Long-standing questions on the importance of land feedbacks and drought self-intensification were answered during the project. The proposed methodology, which combined satellite observations, in situ data and mechanistic models of the lower atmosphere, was extended and culminated in CLASS4GL. The approach showed great promise to scrutinise land feedbacks during extreme events, and enabled the systematic study of the influence of drought–heatwave synchronisation on the morbidity and mortality associated to heat stress. Unlike for plants, the lowering of air humidity due to soil drought was shown to reduced heat stress, instead of aggravating it. This net beneficial impact of soil desiccation on human morbidity and mortality during heatwaves was unexpected. These findings suggested, for instance, that large-scale irrigated agriculture may intensify lethal heat stress instead of reducing it, and it questioned the efficiency of traditionally proposed mitigation strategies involving surface moistening to increase evaporative cooling.
WP3 – Teleconnected feedbacks and event self-propagation
Back in 2017, mesoscale and teleconnected feedbacks from soil moisture and vegetation on rainfall remained largely understudied. Most existing work had centred on local feedbacks or moisture transport within the continents, but outside the context of droughts and heatwaves. The drought and heatwave self-propagation (Fig. 3, 4) – caused by the expansion and concatenation of events as evaporation declines – was a fully novel concept that was postulated for the first time in the DRY–2–DRY proposal. Years later, several high-impact articles clearly demonstrated that, as the air is warmed and dried during a soil drought, it creates a propagation front in the downwind direction. This 'self-propagation' is especially important for drought events in dryland, but applies to heatwaves in Europe as well. An important implication is that upwind land conditions can be used to predict upcoming droughts and heatwaves in regions downwind, probably weeks ahead. This predictability will be explored and exploited in future activities within the host research team.
WP4 – Land cover and land management influence
The impact of vegetation and land-cover change on catchment-scale water resources during dry seasons has long been explored, but most land cover management strategies in the con-text of geo-engineering have targeted afforestation experiments and crop selection to increase terrestrial carbon sinks or surface albedo. The influence of land cover changes and management on the occurrence of local or remote droughts and heatwaves remains understudied. The work in DRY–2–DRY to study atmospheric connections among watersheds led to important findings and implications regarding the need for a more holistic water governance. Conclusions contradicted the traditional views of water resources and water scarcity as local aspects for which watershed‐based management appears adequate, and highlighted the fact that precipitation depends on moisture supplied through land evaporation from thousands of kilometres away. Upwind land and water management can influence drought occurrence in downwind watersheds. This novel perspective strives to provide a more integral management of global water resources, while advocating for the consideration of inter-watershed transfers through the atmosphere as a key water resource. This view was further expanded to demonstrate the importance of plant transpiration for precipitation anomalies in African watersheds, using the HAMSTER tool developed during DRY–2–DRY.