Skip to main content

Do droughts self-propagate and self-intensify?

Periodic Reporting for period 3 - DRY-2-DRY (Do droughts self-propagate and self-intensify?)

Reporting period: 2020-02-01 to 2021-07-31

Droughts and heatwaves cause agricultural loss, forest mortality and drinking water scarcity. Their predicted increase in recurrence and intensity poses a serious threat to global future food security. Several historically unprecedented events have already occurred over the last decade in Europe, Australia and the USA. For instance, the cost of the 2017 Californian drought is estimated to be about US$3 billion. Still today, the knowledge of how droughts and heatwaves start and evolve remains limited, and so does the understanding of how climate change may affect them.

Positive feedbacks from land have been suggested as critical for the occurrence of recent events: as rainfall deficits dry out soil and vegetation, the evaporation of land water is reduced, then the local air becomes warmer and too dry to yield rainfall, which further enhances heatwave and drought conditions. Importantly, this is not just a 'local' feedback, as remote regions may rely on evaporated water transported by winds from the drought-affected region. Following this rationale, droughts and heatwaves self-propagate and self-intensify.

However, a global capacity to observe these processes is lacking. Furthermore, climate and forecast models are immature when it comes to representing the influences of land on temperature and rainfall. In fact, these land feedbacks have been suggested as a key reason why seasonal forecasts fail to provide accurate information relevant for mitigation, and why climate model projections of future drought remain uncertain. Do climate models underestimate these land feedbacks? If so, future drought and heatwave aggravation will be greater than currently expected. At the moment, this remains largely speculative, given the limited number of studies of these processes.

Here, we propose to use novel in situ and satellite records of soil moisture, evaporation and precipitation, in combination with new mechanistic models that can map water vapour and heat trajectories in the atmosphere, and explore multi-dimensional feedbacks. DRY–2–DRY does not only advance our fundamental knowledge of the mechanisms triggering droughts and heatwaves, it will also provide independent evidence of the extent to which managing land cover can help 'dampen' these events, and enable progress towards more accurate short-term and long-term forecasts.
The goals of the project were soon expanded from the original exploration of drought self-propagation and self-intensification to embrace also the analogous processes during heatwaves. This was motivated by the recent occurrence of extreme mega-heatwaves in Europe, and the realisation that the project holds the skills and tools to further the scientific understanding of the drivers behind heatwaves. The objectives of the project were extended accordingly. Results are being broadly disseminated in international publications and conferences. The exploitation of these results is being explored in light of their importance in (a) support of the climate extreme forecasts, (b) understanding global carbon, energy and water budgets, (c) contributing to climate change prediction policies, and (d) guiding on how to improve water management.

The first objective of DRY–2–DRY is to provide evidence of the impacts of drought and heatwaves on terrestrial evaporation. This is a crucial first step to understand the role of land feedbacks. To this end, we compiled a drought and heatwave archive with information on satellite observations and drought and heatwave indices to be used in the remaining of the project. For those events, satellite solar-induced fluorescence (SIF) has been explored as a diagnostic of the evaporation response to drought and heat stress, as well as a stand-alone early warning drought diagnostic (Pagán et al., in review). A novel metric has been proposed which captures natural ecosystem heterogeneity in response to drought stress based on SIF data (Pagán et al., 2019; Maes et al., in review). Moreover, a satellite-based transpiration retrieval scheme has been developed that uses this SIF data. A beta version at high resolution has been published (Martens et al., 2019) and a new data assimilation scheme is being implemented to ingest multiple satellite data from the Sentinel constellation. Current datasets are available for download at

The second objective of DRY–2–DRY is to disentangle the effects of land feedbacks on the local likelihood of rainfall during droughts and heatwaves. To this end, a novel open source tool to study the role of local land feedbacks has been developed and is accessible via Wouters et al., 2019. This tool allows us to easily visualise and physically interpret global balloon sounding data together with satellite data during droughts and heatwaves worldwide. The tool has also been introduced as an educational software in the teaching programs at Ghent University. Novel observational diagnostics of land–atmospheric feedbacks have also been developed and further refined, in order to study the role of dry soils on rainfall occurrence in dry ecosystems (Petrova et al., 2018). Our latest analysis in this regard focuses on the role of surface drying for human health during heatwaves: Are dry soils beneficial because of reducing air humidity or detrimental because of increasing air temperature? The tools we have developed allow us to objectively answer this question. Recent findings indicate that the effect of soil dryness on human heat stress is positive in semiarid climates but negative in humid regions (Wouters et al., in prep.). Moreover, long-term changes in occurrence of dry spells are being explored worldwide using recent satellite data (Petrova et al., in prep.). An international workshop bringing together experts working on local land–atmosphere feedbacks will be held in November (20–22) in Ghent.

The third objective of DRY–2–DRY is to unravel the effects of land feedbacks on the expansion and concatenation of droughts and heatwaves. A new modelling framework that tracks the fate of heat in the atmosphere has been developed and applied to study the spatial propagation of drought and heatwave events. A recent study shows a common ingredient during the largest European mega-heatwaves of this century: drought conditions in the regions the wind blows from. Published in August 2019 in Nature Geoscience, our article demonstrates the concatenation of drought and heatwave events (Schumacher et al., 2019). Because drought conditions are projected to aggravate in midlatitudes, this could trigger even more extreme mega-heatwaves in the future, which should be considered in adaptation strategies. Due to its novelty, this study has already been highlighted in multiple newspaper articles, radio, blogs etc. Moreover, the approach developed is now being used to link ecosystem carbon uptakes to heat and moisture transport; this will help improve predictability of ecosystem states by understanding the regions that affect their supply of heat and moisture during the growing season (Schumacher et al., in prep.). Finally, a new study by Keune and Miralles (in review) challenges the notion of ‘watersheds’ that traditional hydrology builds upon. Findings indicate that due to the atmospheric transport of moisture between watersheds, these geographical units are far from being autarkic, but largely depend on the evaporation in neighbouring and even remote watersheds. This paradigm facilitates an objective assessment of drought propagation between watersheds and calls for global water governance.

The fourth and final objective of DRY–2–DRY is to quantify the influence of land cover and the value of land management in dampening drought and heatwave evolution. A modeling framework to assess land cover feedbacks has been developed, and comprises the combination of the atmospheric trajectory model (see the third objective) with a regional climate model. This enables the analysis of land use and land cover change on moisture and heat transport to remote regions. An interface has been developed and successfully tested in close collaboration with the developers of the regional climate model (Keune et al., in prep.). Meanwhile, the heat and moisture transport over tropical regions is being explored to quantify the role of rainforests in drought and heatwave occurrence in the tropics. In a first study, the impact of the Amazon rainforest in South American cities is analyzed by evaluating moisture and heat transport to these cities in relation to human heat stress indices (Smessaert et al., in prep.). The data needed to carry out the research to address this objective has already been compiled, and a large part of the efforts during the second phase of the project will be dedicated to understanding the influence of land cover on the propagation of droughts and heatwaves and their effects on ecosystems and human health.
Characterizing land–atmosphere feedbacks depends strongly on the capacity to accurately reproduce the spatial and temporal dynamics of terrestrial evaporation. For this reason, the first stages of DRY–2–DRY focused on advancing the state of science in satellite evaporation retrievals. With that goal in mind, the evaporation methodology hosted and developed by the ERC group ( has been further refined to incorporate novel SIF observations. Pagán et al. (2019) investigated for the first time the use of satellite observations of SIF to diagnose transpiration and contrasted the findings against ground-based measurements from eddy-covariance sites. Results demonstrated that SIF data can be empirically related to the ratio of transpiration to potential evaporation, being adequate to capture the effect of phenological changes and environmental stress on transpiration, and adequately reflecting the timing of this variability. Moreover, results indicated that state-of-the-art land surface models used by the IPCC often fail to capture the variability of terrestrial evaporation in periods of water stress. Further steps in this regards include the assimilation into our evaporation retrieval model of the full palette of data from the Sentinel satellite constellation. This will yield, for the first time, a high-resolution, continental-scale, near-real time data set of terrestrial evaporation that can be used in climate studies, but also for agricultural and water management.

The development of CLASS4GL, a tool designed to study land–atmosphere interactions at local scales, filled an existing scientific gap. Understanding the role of land–atmosphere feedbacks is necessary for improving early warnings, better climate projection and timely societal adaptation to extreme events, such as droughts and heatwaves. However, this understanding is typically hampered by the difficulties of attributing cause–effect relationships from complex coupled models and the irregular space–time distribution of in situ observations of the land–atmosphere system. Therefore, there is a need for simple deterministic appraisals that systematically discriminate land–atmosphere interactions from observed weather phenomena over large domains and climatological time spans. CLASS4GL provides that opportunity. It aims to foster a better process-understanding of the factors aggravating weather extremes locally, and can be used to scrutinize the representation of key processes in Earth system models and numerical weather prediction models. The ultimate goal of CLASS4GL remains to help improve local climate projections, provide earlier warning of extreme weather and enable a more effective development of climate adaptation strategies (Wouters et al., 2019). The tool is open-source and can be easily downloaded via First applications within the DRY–2–DRY project have already revealed novel findings regarding the effect of soil dryness on human heat stress during hot spells (Wouters et al., in prep) and the origins of the trends in dry spells that are being experienced worldwide (Petrova et al., in prep.).

The work performed to investigate the fate of heat and moisture in the atmosphere was enabled by the development of an elegant and novel framework that goes far beyond the state of the art in the field. The application of this framework to understand the inter-connections among watersheds via atmospheric moisture transport has challenged the very foundations of conventional hydrology (Keune and Miralles, in review). It demonstrated that these geographical units are far from being autarkic: they strongly depend on the moisture transported from neighbouring and remote catchments. This paradigm allows us to better manage the propagation of droughts across watersheds and calls for an effort to extend water governance beyond the scale of catchments. Likewise, the application of the methodological framework to explore trajectories of heat in the atmosphere has allowed us to push the frontier of knowledge on heatwave evolution. The study by Schumacher et al. (2019), recently published in Nature Geoscience, proved for the first time the connection between upwind droughts and downwind heatwaves. Ongoing work attempts to identify regional hotspots that are particularly prone to the occurrence of these concatenated drought–heatwave events. This will enable enhanced adaptation, and improve understanding of such extreme events. On the other hand, we aim to provide a first-time quantification of the potentially exacerbating impact of remote drought conditions on human health during heatwaves. This will gauge the influence of upwind areas in an intuitive manner, and provide evidence of the hazard that these concatenating events pose on human health. Considering the IPCC projections of droughts and heatwaves, focusing today on advancing our understanding on why these events frequently occur in synchronization appears the only means towards a successful societal adaptation.

Finally, the coupling of the atmospheric transport model to a regional climate model enables novel, continental-scale studies of land use change-induced impacts on drought and heatwave occurrence. Ongoing work on the role of rainforests in the occurrence of droughts and heatwaves will enhance our understanding of the climatic factors triggering human health impacts during these events, and the influence of deforestation. We are currently foreseeing the exploration of continental-scale deforestation and afforestation scenarios for the Amazon and Congo rainforests; this will allow us to identify corresponding feedback pathways through changes in heat and moisture transport. Findings will inform on how to better manage land cover to mitigate the impacts of droughts and heatwaves, both locally and at remote locations, and identify hotspot regions that are vulnerable to tropical deforestation.