Robust climate impact assessments for rivers across Europe

Climate change is reshaping our rivers, with longer droughts and more intense floods. HydroImpact will make robust hydrological climate impact assessments accessible to water professionals, with a platform that quantifies these changes for their own catchment and application.

Get project updates Read the project summary

The project starts in May 2027, funded by BRIDGE.

Emission scenarios

Possible pathways of future greenhouse gas emissions

Global climate models

Global climate

Simulated for the whole planet on a coarse grid

Regional climate models

Regional climate

Simulated over Europe on a finer grid

Bias adjustment and downscaling

Local climate

Brought down to the catchment scale

Hydrological models

River impacts

Changes in streamflow, floods, low flows and seasonality

HydroImpact tests which chains suit each application and region, from the regional climate simulations onwards, and turns them into automated workflows.

Each step offers several models and methods, so the possible answers multiply along the chain. Unsuited choices and ignored uncertainty can lead to costly wrong decisions.

The challenge

Many decisions taken today depend on how rivers will change over the coming decades. Yet turning climate projections into actionable hydrological information remains difficult:

  • The modelling chain is complex. It involves huge climate datasets, and models that take time and expertise to set up.
  • Uncertainty accumulates along the chain. Leaving it out gives false precision, with protection that is too weak or needlessly costly.
  • The right workflow depends on the application. What works for droughts is not what a 100-year flood estimate needs.

Our approach

HydroImpact combines research and software. A systematic benchmark will first identify which modelling chains can be trusted for which application and region. An application-driven cloud platform will then run these validated workflows automatically and add AI-powered insights, so practitioners get robust, uncertainty-aware answers without building the chain themselves.

A systematic benchmark

Each application needs its own workflow, but no systematic benchmark shows which one suits which application and region. We will build it. On catchments across Europe, we will combine bias adjustment and downscaling methods with process-based, data-driven and hybrid hydrological models, and with calibrations that hold in a changing climate.

Only the chains that reproduce what an application needs, for example floods or low flows, will be kept. The spread that remains among them is the uncertainty, which we will quantify, attribute to its sources and explain.

How the platform will work

The platform will offer the validated workflows as a guided, code-free service. Everything runs in the cloud, so no climate data need to be downloaded and no models set up locally. An assessment takes three steps.

  1. Select

    You will choose the catchment on a map, the application (flood frequency, low flows, water availability, seasonality or reservoir inflows) and the time horizon.

  2. Compute

    HydroImpact will run the workflow validated for that application and region in the cloud, and quantify the uncertainty.

  3. Understand

    Results will come decision-ready: time series and indicators with their likely range, plain-language explanations of what changes and why, and reports or an API for your own tools.

Then comes the decision the results are for: sizing infrastructure for future floods, planning hydropower production and storage for the decades ahead, or preparing water-sharing plans for droughts.

Open and co-designed

Models and methods will be mostly open source, and results published in open-access, peer-reviewed articles.

Partners from engineering firms, hydropower companies, public authorities and insurance companies co-design and test the platform, which will cover Switzerland first, then Europe.

Project summary

Read the full summary from the funding proposal

Climate change has a profound impact on streamflow in mountain regions and their dependent lowland areas, with summer water availability decreasing and extreme events such as floods and droughts intensifying. The underlying changes in key hydrological processes, including snow accumulation and melt, evapotranspiration, and precipitation amount and intensity, have severe impacts on water supply and management.

Quantifying such changes requires a hydrological modeling workflow that involves multiple steps, including bias-adjustment and downscaling of climate data, calibration/training and evaluation of hydrological models, and modeling the streamflow response to future climate conditions. While the individual components of this workflow have been well studied, analyzing the effectiveness of different combinations of modeling decisions for targeted applications, such as assessing changes in flood frequency in a specific region, is time-consuming and often beyond practitioners’ expertise and technical capabilities. Consequently, most climate impact studies remain confined to academic settings, and impact assessments conducted by practitioners typically rely on incomplete workflows that consider small subsets of climate ensembles and hydrological models, which results in an underestimation of uncertainty.

In this project, we will make targeted hydrological climate impact assessments widely accessible to practitioners and engineers in Europe. To this end, we will build a cloud-based platform to facilitate, standardize, and enhance climate impact assessments in hydrology by integrating state-of-the-art modeling tools and employing a stakeholder co-design process. The aim of the platform is twofold: (i) ensure that tools and data are targeted to specific applications and (ii) offer reproducible workflows. To achieve this, our efforts will be structured along the following main axes:

  • Identification of suitable hydrological modeling workflows for a wide range of potential applications. At the heart of our project lies the development and an extensive evaluation of modeling workflows, i.e., combinations of different climate data processing techniques, including bias adjustment and downscaling, and hydrological models. We will test the suitability of these workflows for specific applications, e.g., assessing changes in water availability or flood frequency, and for different regions in Europe. Developing suitable workflows will require the development of (i) strategies to facilitate the transferability of climate data processing and modeling tools across diverse regions and climatic conditions; and (ii) uncertainty quantification techniques that can quantify different sources of uncertainty along the modeling chain.
  • Cloud-based platform development building on user needs. The platform will provide a unified and streamlined access to climate and hydrological data, enabling users to define their catchments of interest while automatically retrieving all relevant datasets for the planned application. The system will then identify, calibrate/train, and execute the most suitable hydrological modeling workflow. The resulting climate impact assessment will quantify projected changes in key hydrological indicators under future climate scenarios, including streamflow (e.g., mean discharge), flood and drought frequency and intensity, seasonality shifts, and associated uncertainties. Results will be delivered as summary statistics (e.g., return periods, exceedance probabilities), graphs, and time series which can be further processed by advanced users.
  • AI-driven results interpretation. To facilitate the interpretation of these results, we will develop an AI-powered analytics engine that will highlight key change patterns and provide actionable insights, making complex hydrological data and simulation results more accessible and intuitive to interpret for practitioners.

Our platform will revolutionize hydrological climate impact assessments by providing robust, reproducible, and user-friendly modeling pipelines and simulation analysis tools. It will empower diverse stakeholders (water resources managers, engineers and consultants, river basin authorities, hydropower operators, environmental agencies, risk managers and planners) to perform and interpret tailor-made hydrological climate impact assessments, therefore supporting adaptation to climate change. After project completion, the platform will be maintained and continuously enhanced by a startup to be founded during the project by the project partners, ensuring long-term impact and scalability.

Who is behind it

  • Pascal Horton

    Hydrology group, Institute of Geography and Oeschger Centre for Climate Change Research, University of Bern

  • Manuela Brunner

    WSL Institute for Snow and Avalanche Research SLF, Davos, and ETH Zurich

  • Mathieu Salzmann

    Swiss Data Science Center and EPFL

  • University of Bern
  • WSL Swiss Federal Institute for Forest, Snow and Landscape Research
  • ETH Zurich
  • EPFL
  • Swiss Data Science Center