Strengthening Flood and Drought Preparedness with HydroLand
Climate DT User Story
The Climate Change Adaptation Digital Twin (Climate DT) is implemented as part of the European Commission’s Destination Earth initiative, by a partnership led by the CSC – IT Center for Science and leading institutions across Europe, in close collaboration with ECMWF. Through a co-design process, the Climate DT integrates tailored, user-oriented information for sectors affected by climate change, helping them address related challenges more effectively. Hydrology is one of these sectors.
Understanding how global warming affects terrestrial hydrology is essential for assessing flood and drought risk and future water availability. How climate and land surface interact varies widely with local conditions and land characteristics, which makes impact assessment especially difficult as extreme events grow more frequent. Context-specific information is therefore needed to support decision-making at local and regional scales.
The HydroLand application, developed by the Helmholtz Centre for Environmental Research – UFZ and integrated in the Climate DT workflows, addresses this need. It transforms high-resolution climate simulations into variables and indicators of climate change impacts on terrestrial hydrology, providing stakeholders with hydrological information at a scale that supports targeted decisions on both flood peaks and droughts.
1. About terrestrial hydrology
Terrestrial hydrology examines the movement, distribution, and quality of water across the land component of the Earth system, including soils, rivers, lakes, and groundwater. It describes how precipitation and evapotranspiration govern water flows at, across, and beneath the land surface through processes such as rainfall and snowmelt, runoff, infiltration, evapotranspiration, and groundwater flow.
These processes interact with each other in non-linear ways. While dry soil absorbs rainfall, the same rain falling on ground already saturated from earlier rainfall can cause substantial runoff and even lead to sudden, fast-rising floods. Other local factors such as slope, soil type and land cover shift the outcome further. Because so much depends on the state of the land when rain arrives, assessing changes in flood and drought risks requires models that can bring all of these factors together. The results are highly relevant for water supply, agriculture, hydropower and ecosystem protection.
2. HydroLand: Simulating the terrestrial water cycle
HydroLand runs the mesoscale Hydrological Model, developed at the UFZ. Driven directly by the Climate DT’s climate simulations, mHM represents how water moves through and beneath the land surface and produces a set of hydrological information: evapotranspiration, soil moisture, runoff, and river discharge.
These variables are produced globally at the same high spatial resolution (i.e., 5km) and time frequency (i.e., hourly) as the climate models embedded in the Climate DT system. At this level of detail, the model can simulate processes that unfold quickly and over short distances, for example the rise of a river in the hours after an intense rainstorm, or the difference in water availability between one valley and another.
HydroLand can also be used to test how changes on the land itself would alter these processes. Land cover and soil properties are described explicitly in mHM, so a simulation can be repeated with these conditions modified. This allows HydroLand to answer what-if questions such as: “If a region lost 30% of its forest cover, how would that affect the water stored in its soils and the response of its rivers to heavy rainfall?”
3. HydroLand: working with users for tailored indicators
HydroLand actively integrates users’ needs throughout its development process. The application targets water suppliers, catchment and reservoir managers. Through this co-design process, HydroLand produces key indicators tailored to the water sector, and is fully integrated into the Climate DT’s system alongside the climate models.
These indicators address floods and droughts and are based on climate data produced by the Climate DT through the two types of simulations: storylines, which reconstruct specific extreme-events under present and future scenarios, and multi-decadal climate projections, which provide long-term climate trends.
The Aridity Index is the ratio of the mean annual actual evapotranspiration (aET) to the mean annual precipitation (pre) for each grid cell. Values approaching 1 indicate that most precipitation is returned to the atmosphere as evapotranspiration, reflecting drier conditions, while lower values indicate greater water availability. It is computed from mHM output by aggregating sub-daily data into monthly and annual sums, averaging over all available years and then taking the ratio.
The Discharge Indicators characterize flood and drought events in river discharge relative to a percentile threshold calculated from the simulated discharge record at each grid cell. Three indicators are derived: frequency (number of events per grid cell), duration (total number of timesteps per year during which an event is active), and intensity (cumulative deviation of discharge from the threshold during events). Events are classified as floods when discharge exceeds the threshold (percentile ≥ 90 and 95) and as droughts when it falls below it (percentile < 5 and 10).
Understanding HydroLand indicators
Floods
Both flood and drought indicators are based on percentile thresholds of river discharge, calculated per location:
- The threshold is the 30-year mean of the annual percentile discharge:
- A high-flow percentile (e.g. P90, P95 or P99) for floods
- A low-flow percentile (e.g. P1, P5 or P10) for droughts
- The thresholds are calculated separately for the reference period (1990-2014) and the scenario period (2015-2049).
Events are detected against the thresholds:
- A flood event starts when discharge exceeds the high-flow thresholds (e.g. P95) and ends when it falls back below.
- A drought event starts when discharge falls below the low-flow threshold (e.g. P10) and ends when it rises above it.
Flood Event Definition
Droughts
4. Scenario studies demonstrating HydroLand’s capabilities in Germany
4.1 Black Forest –Co-designing indicators to assess land-use change
In the Dreisam catchment in the Black Forest (Germany), forest cover plays a central role in regulating surface runoff and river discharge. Changes in that cover, through conversion to agricultural or urban land for example, affect local hydrology with downstream consequences for flood risk and water quality.
To evaluate these impacts, the Climate DT partnership worked with users to build a set of “what if” land-use scenarios representing forest loss of 30%, 60% and 85%. Designing the scenarios jointly keeps them anchored in the decisions the users actually face in local planning and decision-making.
HydroLand simulates each scenario and returns the resulting changes in runoff, soil moisture and river discharge. Users can feed those outputs then into their own water quality models to trace the downstream effects, and their feedback on which outputs prove useful in practice guides how HydroLand’s indicators are refined. The result is a consistent framework for testing how future land management decisions would play out, under both average and extreme conditions.
4.2 Southern Germany – Using HydroLand Indicators to Explore Extreme Rainfall Impacts
Between 31 May and 3 June 2024, southern Germany experienced an intense period of rainfall, with many areas receiving more than 150 mm of precipitation in just a few days. This event was driven by an atmospheric low-pressure system that moved from the Mediterranean, around the eastern side of the Alps, and into central Europe. Rivers in the Swabian Jura and the Alpine foothills south of the Danube rose above levels expected to occur only about once a century, causing widespread damage, including dam failures, evacuations, and transport disruption.
Using a storyline simulation from the Climate DT, HydroLand calculated key hydrological variables to explore how a similar event would unfold in a world 2°C warmer. The application highlights key changes to the streamflow (river discharge), total surface run-off and the total soil moisture in the affected region.
In this storyline, the warmer atmosphere increases precipitation amounts, and river discharge rises in the tributaries feeding the Danube. Comparable initial soil moisture conditions at the beginning of the event imply comparable infiltration. As a result, additional precipitation runs off rather than infiltrates, resulting in higher flood peaks. This example demonstrates how HydroLand can translate climate simulations into actionable insights, highlighting potential changes in hydrological extremes and helping anticipate the combined effects of rainfall patterns and catchment responses under climate change.
5. Future outlook
Building on this experience, the developers of the Hydroland application aim to extend the collaboration to a wider range of users across Europe, bringing in diverse sectors and hydrological contexts to broaden and strengthen its applications. Through this ongoing engagement, HydroLand and the Climate DT are moving toward operational workflows that deliver regular, actionable hydrological indicators, supporting decision-making for flood and drought preparedness across Europe and beyond. Additionally, HydroLand aims at aligning with other services on the DestinE platform focussing on hydrologic information like the European Drought Monitor. This will allow users not only to monitor current drought events, but also understand how they would unfold in a warmer climate.