Climate change and risks 3 min

Anticipating drought severity via river-flow data

INRAE coordinates an operational tool designed to inform the French authorities of the severity of forthcoming droughts, enabling the latter to plan appropriate restrictions on water use. We put three questions to hydrologist Vazken Andréassian, head of the INRAE research unit at the heart of the project.

Published on 09 August 2022

illustration Anticipating drought severity via river-flow data
© INRAE J.Weber

What are the benefits of the tool your researchers have developed?

Vazken Andreassian: Serious droughts can cause major reductions in river flows. In some regions, even access to drinking water can be affected, meaning that supplies must be brought in by road. It is therefore essential to be able to predict these events so that they can be managed as well as possible. The French public services keep a very close eye on river flows using data from multiple measuring stations sited along our watercourses. The collected data are sent to a national database (HYDRO).

The tool we have developed (PREMHYCE 1) enables us to use these data, along with predictive modelling, to forecast river flows in the short (0-15) and medium terms (16-90 days), making it possible to anticipate what restrictions on water usage may need to be imposed as a function of the predicted streamflows. Decisions on this are taken by the prefects of the départements, who then issue restriction orders (see the 2022 map).

How does the tool work?

V. A.  In practical terms, the tool operates as a website that combines several predictive models2. The shorter-term forecasts (up to 15 days) are based on rainfall predictions. Beyond 15 days, a climatic approach is used – streamflow predictions are based on many years of observed relationships between climate and river flows. The data we select to feed the models (chosen from around a thousand measuring stations) are representative of the status of each watercourse and provide full coverage of an area. For example, we won’t pick a measuring station that is just downstream from a dam since the flows it records will reflect the water-release schedules of the dam’s managers rather than local rainfall.  We send the predictions from the various models, along with our uncertainty assessments, to each regional public body (DREAL 3), which then analyses these short and medium-term results so it can provide the best information possible to decision-makers at the prefectures.

Why compare the data on river flows from 1976 and 2022?

V. A.: The media regularly ask this question. The comparison is well founded, in that 1976 is the historic drought year that has most in common with the current year, in terms of both the duration and extent of the drought, which is affecting the whole of France.

Looking right now, at the beginning of August, at the data on river flows in 1976 and 2022, we can confirm that the two droughts are both comparable and serious. This year’s drought has been more marked in the east of France compared with 1976 but has been a little less severe in the west. France has repeatedly experienced severe episodes of drought (the previous two were in 1921 and 1949). We need to be vigilant, because the IPCC predicts a likely increase in the frequency of such major droughts and, above all, we need to anticipate their impact and do something about it!

  1. The PREMHYCE tool, see inset.
  2. These are the meteorological, climate and hydrological models produced by INRAE, the BRGM (French geological survey), EDF (national electricity supplier), the University of Lorraine and the French forecasting service Météo-France
  3. DREAL: regional directorate for environment, development and housing.
Comparison of the drought events of 1976 and 2022 based on observed flows in the final week of July

The PREMHYCE programme and its partners

PREMHYCE (low-flow forecasting with hydrological models, comparison and assessment) is a research programme coordinated by INRAE, with support from the OFB (French biodiversity agency) and the French Ministry for Ecological Transition. It brings together teams from the BRGM (French geological survey), EDF (national electricity provider), Météo-France (weather forecasting service) and the University of Lorraine. PREMHYCE produces flow forecasting for several hundred reporting points in France, for up to 90 days ahead.

The work of PREMHYCE has been extended through a research project funded by the ANR, the French national research funding body, known as CIPRHES (integrated chain for the hydrometeorological forecasting of low flows and droughts). INRAE is coordinating this project, working with its existing PREMHYCE partners. The project aims to further develop novel approaches to integrated hydrometeorological modelling for low-flow forecasting, based on the PREMHYCE model for ensemble streamflow prediction. Contact:

Pascale Molliertranslated by Teresa Bridgeman


François Bourgin, François Tilmant, Vazken AndréassianContinental Hydrosystems – Resources, Risks, Restoration (UR1462 HYCAR)



Learn more


More than half of the world’s rivers are non-perennial: researchers mapped them for the first time

PRESS RELEASE - A new study of 64 million kilometres of rivers around the world reveals that between 51% and 60% of them stop flowing periodically or dry for part of the year. These findings, published on 16 June 2021 in Nature, stem from the first-ever effort to empirically quantify and map the global distribution of non-perennial rivers and streams. This study, conducted by researchers from INRAE and McGill University, aims to strengthen recognition of the prevalence and ecological, and the need for improved management of these unique ecosystems.

07 June 2021

Climate change and risks

The launch of the first open-source smartphone application to monitor drying events in river networks

PRESS RELEASE - River networks are among Earth’s most threatened hotspots of biodiversity and are essential for human wellbeing. Climate change and increasing human water use are causing more rivers and streams to dry, but these drying river networks have received little attention. DRYvER is a Horizon 2020 project which will collect, analyse and model data from drying river networks across two continents—Europe and South America—to create a novel global meta-system approach that incorporates hydrology, socioeconomics, ecology and biogeochemistry in order to craft strategies, tools, guidelines, and recommendations for adaptive management of river networks.

28 September 2021

Climate change and risks

Water resources, global changes: perspectives for hydrology research

Climate and environmental changes, modelling, risks, new societal challenges… using a broad consultation process, the international scientific community identified the 23 most important unsolved problems in the field of hydrobiology. Interview with Vazken Andréassian, director of the Research Unit on Hydrosystems and Bioprocesses, Resources, Risks, and Restoration (HYCAR) of the INRAE research centre of Île-de-France - Jouy-en-Josas-Antony who participated in this community consultation.

07 July 2020