• Energy Research

Intermittent rivers and ephemeral streams (IRES) account for more than half of the world's rivers.However, few studies have investigated the evolution of IRES under climate change. Aiming toovercome this problem, DRYvER proposed to provide daily hydrological projections, daily flowconditions and flow intermittence indicators in 6 European Drying River Networks (DRNs).The current work aims to produce reach-scale daily hydrological projections available for the period1985-2100 for each DRN. To this end, coarse spatial resolution daily projections from Global ClimateModels (GCMs) are downscaled to obtain high resolution projections over the period 1971-2100.Secondly, the high-resolution projections are used as input to the JAMS/J2000 model to obtain dailycatchment-scale hydrological projections. Several GCMs are used as well as 3 shared socio-economicpathways (SSPs) to capture the uncertainty due to climate modelling and greenhouse gas emissionscenarios.The results show that the methodology is able to reproduce the historical hydrological behaviour ofthe DRNs in terms of seasonality, with some difficulties in Morava and Vantaanjoki regarding summerdischarge. Regarding the future periods, the responses of the six catchments were clearly different,showing an impact of climate change closely related to their location. Spring, summer and autumndischarges show a decrease for all catchments and all SSPs considered. For winter discharge, two of thecatchments show a slight increase, but the other four also show a decrease of varying intensity.

Intermittent rivers and ephemeral streams (IRES) account for more than half of the world's rivers.However, few studies have investigated the evolution of IRES under climate change. Aiming toovercome this problem, DRYvER proposed to provide daily hydrological projections, daily flowconditions and flow intermittence indicators in 6 European Drying River Networks (DRNs).The current work aims to produce reach-scale daily hydrological projections available for the period1985-2100 for each DRN. To this end, coarse spatial resolution daily projections from Global ClimateModels (GCMs) are downscaled to obtain high resolution projections over the period 1971-2100.Secondly, the high-resolution projections are used as input to the JAMS/J2000 model to obtain dailycatchment-scale hydrological projections. Several GCMs are used as well as 3 shared socio-economicpathways (SSPs) to capture the uncertainty due to climate modelling and greenhouse gas emissionscenarios.The results show that the methodology is able to reproduce the historical hydrological behaviour ofthe DRNs in terms of seasonality, with some difficulties in Morava and Vantaanjoki regarding summerdischarge. Regarding the future periods, the responses of the six catchments were clearly different,showing an impact of climate change closely related to their location. Spring, summer and autumndischarges show a decrease for all catchments and all SSPs considered. For winter discharge, two of thecatchments show a slight increase, but the other four also show a decrease of varying intensity.

The 2022 European drought affected the whole of France, leading to very severe summer low-flows. The outstanding characteristics of this event is questioned within a long-term historical context. It first makes use of 600+ daily streamflow series gauging near-natural catchments from all French regions. The historical background is provided by the 25-member ensemble hydrological reanalysis FYRE Hydro which covers the period 1871-2012 for the above stations. FYRE Hydro originates from a 25-member ensemble streamflow simulation using the GR6J lumped conceptual model. These simulations integrate three types of uncertainty: (1) the meteorological uncertainty through the use of the 25 members of the high-resolution FYRE Climate meteorological reanalysis (Devers et al., 2021) as forcings, (2) the uncertainty in streamflow measurement used to calibrate the hydrological models, and (3) the hydrological model error based on relative discrepancies between observed and simulated streamflow (Bourgin et al., 2014). An ensemble Kalman filter furthermore combined these ensemble simulations with available historical series together with their uncertainties to produce the FYRE hydro reanalysis. Streamflow observations from 2022 are compared to severe drought years in the FYRE Hydro reanalysis as identified by Caillouet et al. (2021). Results show that 150-year records were broken over a large number of stations for various low-flow indicators, confirming the exceptional nature of this hydrological drought event. Bourgin, F., Ramos, M., Thirel, G., and Andréassian, V.: Investigating the interactions between data assimilation and post-processing in hydrological ensemble forecasting, Journal of Hydrology, 519, 2775 – 2784, 85 https://doi.org/https://doi.org/10.1016/j.jhydrol.2014.07.054, 2014Caillouet, L., Vidal, J.-P., Sauquet, E., Devers, A., Lauvernet, C., Graff, B., and Vannier, O.: Inter-comparison of extreme low-flow events in France since 1871, LHB: Hydroscience Journal, 107, 1-9 https://doi.org/10.1080/00186368.2021.1914463, 2021Devers, A., Vidal, J.-P., Lauvernet, C., and Vannier, O.: FYRE Climate: a high-resolution reanalysis of daily precipitation and temperature in France from 1871 to 2012, Clim. Past, 17, 1857–1879, https://doi.org/10.5194/cp-17-1857-2021, 202

The 2022 European drought affected the whole of France, leading to very severe summer low-flows. The outstanding characteristics of this event is questioned within a long-term historical context. It first makes use of 600+ daily streamflow series gauging near-natural catchments from all French regions. The historical background is provided by the 25-member ensemble hydrological reanalysis FYRE Hydro which covers the period 1871-2012 for the above stations. FYRE Hydro originates from a 25-member ensemble streamflow simulation using the GR6J lumped conceptual model. These simulations integrate three types of uncertainty: (1) the meteorological uncertainty through the use of the 25 members of the high-resolution FYRE Climate meteorological reanalysis (Devers et al., 2021) as forcings, (2) the uncertainty in streamflow measurement used to calibrate the hydrological models, and (3) the hydrological model error based on relative discrepancies between observed and simulated streamflow (Bourgin et al., 2014). An ensemble Kalman filter furthermore combined these ensemble simulations with available historical series together with their uncertainties to produce the FYRE hydro reanalysis. Streamflow observations from 2022 are compared to severe drought years in the FYRE Hydro reanalysis as identified by Caillouet et al. (2021). Results show that 150-year records were broken over a large number of stations for various low-flow indicators, confirming the exceptional nature of this hydrological drought event. Bourgin, F., Ramos, M., Thirel, G., and Andréassian, V.: Investigating the interactions between data assimilation and post-processing in hydrological ensemble forecasting, Journal of Hydrology, 519, 2775 – 2784, 85 https://doi.org/https://doi.org/10.1016/j.jhydrol.2014.07.054, 2014Caillouet, L., Vidal, J.-P., Sauquet, E., Devers, A., Lauvernet, C., Graff, B., and Vannier, O.: Inter-comparison of extreme low-flow events in France since 1871, LHB: Hydroscience Journal, 107, 1-9 https://doi.org/10.1080/00186368.2021.1914463, 2021Devers, A., Vidal, J.-P., Lauvernet, C., and Vannier, O.: FYRE Climate: a high-resolution reanalysis of daily precipitation and temperature in France from 1871 to 2012, Clim. Past, 17, 1857–1879, https://doi.org/10.5194/cp-17-1857-2021, 202

Key messages from the modelling results of flow intermittence projections under climate change scenarios in the six European river networks studied as part of the H2020 DRYvER project (https://www.dryver.eu/). The dataset contains one global summary fact sheet as well as one fact sheet per studied river network : Albarine (France), Bükkösdi (Hungary), Butiznica (Croatia), Genal (Spain), Lepsämänjoki (Finland), and Velicka (Czechia).

Key messages from the modelling results of flow intermittence projections under climate change scenarios in the six European river networks studied as part of the H2020 DRYvER project (https://www.dryver.eu/). The dataset contains one global summary fact sheet as well as one fact sheet per studied river network : Albarine (France), Bükkösdi (Hungary), Butiznica (Croatia), Genal (Spain), Lepsämänjoki (Finland), and Velicka (Czechia).

Hydrological model outputs at daily time step for the projection simulations (1985-2100) in the Velicka DRN (Czechia). Simulated discharge, baseflow an state of flow (flowing/dry) is spatially distributed at the reach scale. Other hydroclimatic variables (temperature, precipitation, rainfall, snowfall, potential evapotranspiration, actual evapotranspiration, vegetation interception, snow water equivalent, saturation of the soil layer, saturation of the grounwater layer) are spatially aggregated at the catchment scale. Hydrological projections were produced for 3 SSP scenarios and 5 climate models.

Hydrological model outputs at daily time step for the projection simulations (1985-2100) in the Velicka DRN (Czechia). Simulated discharge, baseflow an state of flow (flowing/dry) is spatially distributed at the reach scale. Other hydroclimatic variables (temperature, precipitation, rainfall, snowfall, potential evapotranspiration, actual evapotranspiration, vegetation interception, snow water equivalent, saturation of the soil layer, saturation of the grounwater layer) are spatially aggregated at the catchment scale. Hydrological projections were produced for 3 SSP scenarios and 5 climate models.

Hydrological model outputs at daily time step for the projection simulations (1985-2100) in the Genal DRN (Spain). Simulated discharge, baseflow an state of flow (flowing/dry) is spatially distributed at the reach scale. Other hydroclimatic variables (temperature, precipitation, rainfall, snowfall, potential evapotranspiration, actual evapotranspiration, vegetation interception, snow water equivalent, saturation of the soil layer, saturation of the grounwater layer) are spatially aggregated at the catchment scale. Hydrological projections were produced for 3 SSP scenarios and 5 climate models.