• Land-atmosphere interactions and hydrological processes, including feedback mechanisms.
• Understanding the meteorological processes driving hydrological extremes.
• Tools, techniques, and expertise in forecasting hydro-meteorological extremes (e.g., river flooding, flash floods, droughts etc.).
• Fully integrated numerical earth system modelling.
• Quantification/propagation of uncertainties in hydro-meteorological model forecasts.
• The role of vegetation in hydro-meteorological extremes, in terms of transpiration, photosynthesis, phenology, etc.
• Energy cycles, complementing the hydrological cycles and related cryospheric processes.
• Hydro-meteorological prediction that includes impacts.
• Environmental variable monitoring by remote sensing and other observations.
• Quantification of (past/future) hydrological trends in observations and climate models
River floods are among the most disastrous and costly extreme weather events around the world. Atmospheric blocking events (persistent, slow-propagating and self-preserved weather systems that 'get stuck' and 'slow down' the atmospheric circulation) are a key feature of the persistent weather regimes over the Euro-Atlantic, playing a central role in shaping the weather and extreme weather of Europe. Notwithstanding this socioeconomic importance, the covariability between atmospheric blocking and river flooding has rarely been examined at the climate and continental scales. Our study explores the hydrological way in which atmospheric blocking propagates into river floods, and how this relationship varies across space and time. We analyse observed flood discharge across the continent together with atmospheric and terrestrial variables derived from reanalysis, resulting in >6000 rivers analysed over the last 60 years. We combine process-based hydrological analysis and climate analytics to obtain the influence of atmospheric blocking on floods in Europe across spatial and temporal scales, characterising the seasonal behavior, temporal dynamics over decades, and spatial patterns of this relationship. Our results clearly demonstrate the large-scale signature of blocking on flood behavior, and how atmospheric blocking has influenced the hydrometeorological drivers of river floods at the regional and continental scales, inducing robust patterns of variability across different flood attributes over the last 60 years. Despite strong decadal variability in the hydrological propagation of atmospheric blocking and blocking activity itself, our analyses reveal significantly changing impacts of atmospheric blocking on floods in Europe. Furthermore, our results demonstrate how the effects of atmospheric blocking interact locally with the regional hydrological characteristics. Overall, our findings link a dynamical mechanism with flood behavior within the climate system, improving the understanding of multiple features of flood variability in Europe over the last 60 years.
How to cite: Hernandez, D., Bertola, M., and Blöschl, G.: The hydrological propagation of atmospheric blocking into river floods in Europe, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-698, https://doi.org/10.5194/ems2026-698, 2026.
Atmospheric Rivers (ARs) are recognised as important drivers of extreme precipitation and flooding in many mid-latitude regions. However, their hydrological impacts over the complex terrain of the Italian peninsula remain relatively underexplored. In this study, we investigate the ground-level impacts associated with two extreme AR-related events affecting Italy: Vaia storm in October 2018 and Alex storm in October 2020. These events impacted different regions, characterised by diverse physiographic settings, thus providing an opportunity to analyse how AR-driven precipitation translates into river response under various orographic and hydrological conditions.
The analysis is based on distributed hydrological simulations performed using the CETEMPS Hydrological Model (CHyM) at high spatial resolution. CHyM control simulations were forced with hourly rainfall and air temperature observations from the Italian Civil Protection network and used as comparison for numerical experiments forced with hourly precipitation fields produced by the MOLOCH convection-permitting model at different spatial resolutions. Additional sensitivity experiments were conducted using meteorological simulations in which the AR contribution was largely weakened, allowing the assessment of the role of AR-related moisture transport in the hydrological response.
Hydrological impacts were evaluated using two stress indices derived from CHyM simulations: the Best Discharge-based Drainage index (BDD), which reflects river discharge conditions along the drainage network, and the CETEMPS Alarm Index (CAI), which highlights rapid runoff responses associated with intense precipitation.
The results highlight the important role of AR-driven precipitation in shaping the hydrological response during both events, also emphasizing the strong modulation exerted by regional orography, basin morphology, and drainage network structure. The comparison between simulations with and without AR forcing further illustrates the relevance of accurately representing AR-related processes in coupled meteorological–hydrological modelling frameworks.
Overall, this study contributes to improving the understanding of the links between AR dynamics and hydrological impacts and supports the development of more reliable approaches for forecasting AR-related hydro-meteorological hazards in complex terrain environments.
How to cite: Tomassetti, B., Di Antonio, L., Verdecchia, M., Colaiuda, V., Lombardi, A., Mastrangelo, D., and Davolio, S.: Ground impacts of Atmospheric Rivers over Italy: the extreme events of Vaia (2018) and Alex (2020) storms., EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-528, https://doi.org/10.5194/ems2026-528, 2026.
Challenges and risks related to flood prediction and warning, which became evident during and in the aftermath of devastating flood events in recent years in Germany, call for a new level of cooperation between weather and flood forecasting and their respective operational services. While these domains have historically been structurally quite separate in Germany, several collaborative projects are now laying the foundation for increased exchange and joint development.
The first phase of the project “Co-Design of Innovations between Weather and Flood Forecasting” (Italia–Deutschland science-4-services, IDEA-S4S), initiated directly in response to the lessons learned from the Ahrtal flood in 2021, addresses multiple topics. One focus is on improving mutual knowledge and understanding of forecast evaluation and verification, especially given the constantly increasing number of weather forecasting systems. Another emphasis has been placed on analyzing and training communication along the entire warning chain down to disaster management authorities, in particular regarding the handling of forecast uncertainty. The upcoming second phase of Co-Design aims to deepen this communication aspect by testing decision-making strategies that incorporate probabilistic criteria in multi-stage alert and readiness plans, in direct collaboration with emergency response units. Moreover, the second project phase will, among other topics, explore ML-based approaches to combine multiple ensemble models with different lead times into a best-guess ensemble as input for a hydrological model, thereby facilitating the use of the growing number of models in downstream flood applications.
The joint research project KI-HopE-De aims to test new machine-learning-based approaches and brings together meteorologists and hydrologists from both academia and operational services. First, a new high-resolution hourly hydrometeorological dataset (CAMELS-DE-1h) will be compiled and published for around 1,600 catchments across Germany. Second, various training strategies—from both hydrological and meteorological perspectives—for a regionally trained LSTM flood forecasting model will be developed to provide a benchmark model that is compared with the current operational models in Germany.
This contribution provides an overview of ongoing projects and highlights related presentations at the conference. We will share results and provide an outlook on upcoming work, and look forward to exchanging experiences with other initiatives at the intersection of meteorology and hydrology.
How to cite: Bondy, J., Keller, J., Fundel, V., Blumenstein-Weingartz, I., Kiseleva, O., Wolff, S., Fundel, F., Lambert, A., Rauthe-Schöch, A., Khosravi, S., Valmassoi, A., Keller, J., Loritz, R., Ehret, U., Dolich, A., Demuth, N., and Badde, U.: Towards a new level of collaboration and communication between weather and flood forecasting: an overview of ongoing efforts in Germany, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-733, https://doi.org/10.5194/ems2026-733, 2026.
Small and medium-size catchments respond rapidly to intense rainfall and are therefore especially challenging for flood forecasting, as short warning times coincide with large meteorological and hydrological uncertainties. Within the KI-HopE-De (AI-based flood prediction in small river basins in Germany) project, which aims to develop AI-based flood forecasting for small catchments, we investigate how probabilistic numerical weather prediction can be transformed into hydrologically more useful inputs for machine-learning-based runoff prediction. The project focuses on short-range forecasts up to 48 hours and on catchments smaller than 500 km².
Our contribution addresses the meteorological side of this challenge through catchment-based post-processing of ensemble precipitation forecasts. We use hourly operational ICON-D2 ensemble precipitation forecasts with 3-hourly initializations and lead times up to 48 hours, aggregated from the native model grid to catchments. The resulting catchment-scale ensemble time series are post-processed using statistical and machine-learning-oriented methods designed to improve both calibration and hydrological relevance. In particular, we examine ensemble model output statistics (EMOS) and isotonic distributional regression (IDR) for marginal adjustment, and reconstruct temporal dependence using approaches such as ensemble copula coupling and the Schaake shuffle. This setup is designed to retain the benefits of probabilistic weather forecasts while generating physically and hydrologically more consistent forcing data for downstream flood-prediction models.
The presentation will show how these post-processing strategies perform at two connected levels. First, we will compare post-processed precipitation forecasts against the raw ensemble using probabilistic verification metrics and event-based diagnostics, with particular emphasis on extremes. Second, we will assess whether improved meteorological inputs translate into improved flood forecasts in selected severe-flood cases. In this way, the study highlights not only whether post-processing improves rainfall forecasts, but also whether these improvements are relevant for downstream hydrological AI applications in small, fast-responding catchments.
How to cite: Sharifi, E., Lerch, S., Perschke, M., and Knippertz, P.: Catchment-based post-processing of probabilistic weather forecasts for AI-supported flood prediction in Germany, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-507, https://doi.org/10.5194/ems2026-507, 2026.
In Germany, flooding in small catchments, driven by locally intense convective precipitation, has caused significant damage and in some cases fatalities in recent years. To ensure more accurate and timely operational flood forecasts and warnings, the regional German Flood Forecasting Centres require improved precipitation forecasts for short-term forecast ranges.
In this context, the ICON Rapid Update Cycle (RUC) of the well-established, high-resolution (2.2 km) regional German numerical weather prediction model ICON-D2 represents a promising approach. The RUC forecasts have been developed by the German Meteorological Service (Deutscher Wetterdienst, DWD) within the framework of SINFONY (Seamless Integrated Forecasting System). They are operationally available since summer 2024. Compared to the standard ICON-D2 model, the RUC— both deterministic and probabilistic — feature an improved two-moments microphysics scheme to better represent strong convective events, hourly updates (instead of three-hourly), a forecast horizon of 14 hours (compared to 48h), and earlier availability for users due to a reduced computation time.
Evaluating the impact of these “rapid update” characteristics on flood forecasting is one of the key objectives of the project “Co-Design of Innovations between Weather and Flood Forecasting”, which aims to strengthen the collaboration between the DWD and the regional German Flood Forecasting Centres. As part of this initiative, the Flood Forecast Centre of the State Institute for Environment Baden-Württemberg (HVZ LUBW) has been investigating the use of the RUC products for hourly operational discharge forecasting, using the distributed water balance model LARSIM (Large Area Runoff Simulation Model).
This contribution presents a verification of discharge forecasts for a large number of gauging stations in Baden-Württemberg during the summer 2024 flood events in south-western Germany. Discharge hindcasts driven by ICON-D2 and ICON-D2 RUC products were performed and systematically compared. The analysis includes both event-based case studies and an overall statistical evaluation.
How to cite: Kiseleva, O., Deutschländer, T., Rauthe-Schöch, A., Bondy, J., Bremicker, M., Badde, U., and Elfgang, D.: Towards Improved Flood Forecasting: Evaluation of Discharge Forecasts Driven by New Rapid-Update Precipitation Forecast Products of DWD, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-406, https://doi.org/10.5194/ems2026-406, 2026.
Accurate flood forecasting in small and medium-sized catchments remains a major
challenge due to rainfall prediction uncertainties, limited hydrological data at that
scale as well as short warning times related to rapid response times of such systems.
The joint German research project KI-HopE-DE (KI-gestützte Hochwasserprognose
für kleine Einzugsgebiete in Deutschland) aims to improve flood prediction in
Germany by testing new machine learning-based approaches and bringing together
meteorologists and hydrologists, both from academia and from operational services.
Within this framework, the German Weather Service (DWD) contributes by providing
high-resolution meteorological datasets to support data-driven modelling, as well as
by testing training strategies that stronger account for weather model particularities.
KI-HopE-De develops a regionally-trained Long Short-Term Memory (LSTM) using
data from 1,626 catchments across Germany. Besides the classical training based on
meteorological observations, in our study we explore the use of the novel ICON-
FORCE (Fine-scale Observation-based Reanalysis for Central Europe) reanalysis
dataset for optimizing the LSTM. This approach attempts to leverage the proximity of
the ICON-FORCE reanalysis data and the ICON-D2 forecast model later used for
inference and making forecasts. Meteorological variables derived from ICON-FORCE
reanalysis are used as input features to capture spatiotemporal dependencies and
interactions relevant for runoff generation. The study is designed as a large-sample
experiment to systematically assess the added value of high-resolution
meteorological reanalysis for data-driven flood prediction, with a particular focus on
model robustness and applicability across divers catchments.
By focusing on the interface between meteorology and hydrology, this work
contributes to ongoing efforts to better integrate atmospheric and hydrological
information in flood forecasting models.
How to cite: Khosravi, S., Valmassoi, A., Bondy, J., Dolich, A., Ehret, U., Loritz, R., and Keller, J.: Integrating meteorological and hydrological modeling in KI-HopE-DE: AI-basedflood prediction in small river basin in Germany using ICON-FORCE reanalysis, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-667, https://doi.org/10.5194/ems2026-667, 2026.
Traditionally, to simulate future river discharge on a global scale, a global hydrological model would be forced with atmospheric input data from a climate model. However, this approach does not consider any coupling effects between earth system model components such as the ocean, sea-ice, and most importantly in this case: the land-surface. Nowadays, these components are often integrated in earth system models (ESMs), allowing to directly study the interaction between precipitation, evaporation and surface and subsurface runoff. Since very recent, ESMs are capable of performing multi-decadal simulations globally at kilometer-scale (9 by 9 km spatial resolution), allowing to (partly) resolve convection, improve orographic precipitation, and include the impact of land surface heterogeneity, while also capturing all large-scale drivers. Given these accomplishments, we are asking the questions: 1) Can we assess future discharge changes by linking a kilometer-scale ESM directly to a river routing scheme? And 2) Do we see the impact of kilometer-scale resolved processes in the atmosphere and on land reflected in the discharge representation?
Here, we use kilometer-scale earth system simulations with the Integrated Forecasting System (IFS) model for past (1990-2020) and future climate (2020-2050; SSP3-7) as input for the river routing Catchment-Based Macro-scale Floodplain (CaMa-Flood) model. This provides global high-resolution historical and future river discharge simulations. Our modelling set-up allows to study how the water balance components react and interact; from atmospheric precipitation, to land-surface feedbacks including evaporation and runoff, to river routing generating streamflow. For the analysis, we first focus on the well-studied and well-observed Rhine basin. Thereafter, we extend our analysis to a diverse set of catchments of varying size globally. With the main goal to assess the impact of kilometer-scale resolved weather and land-atmosphere interactions on future discharge signals, and specifically during floods and droughts.
How to cite: Benedict, I., Spierenburg, A., Pedruzo Bagazgoitia, X., and M.C. Denissen, J.: Projecting future global discharge signals with kilometer-scale earth system simulations combined with river routing, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-660, https://doi.org/10.5194/ems2026-660, 2026.
In recent years, climate change has increased the spatiotemporal variability of precipitation, leading to a rise in both the frequency and severity of drought events. Consequently, the importance of water resource management and drought response has been steadily growing. Drought is not merely a result of precipitation deficiency; rather, it is a complex phenomenon that propagates from meteorological drought to agricultural, hydrological, and eventually socioeconomic drought. However, previous studies on drought propagation have largely focused on individual stages, which limits the comprehensive understanding of compound drought propagation under climate crisis conditions. In this study, regions that have experienced severe historical drought damage were selected to analyze the characteristics of drought propagation. By integrating meteorological and climate data with regional characteristic data, this study aims to develop a methodology for evaluating compound drought propagation. First, drought indices were calculated for each stage: the Standardized Precipitation Index (SPI) for meteorological drought, the Standardized Precipitation Evapotranspiration Index (SPEI) for agricultural drought, and the Streamflow Drought Index (SDI) for hydrological drought. Next, various spatiotemporal correlation analyses were conducted to investigate both spatial and temporal propagation characteristics of drought indices. Transition matrices for each drought stage were then constructed and analyzed using network analysis techniques. Furthermore, based on the identified propagation characteristics, transition probability analysis between drought stages was performed. This enabled the estimation of when meteorological drought conditions are likely to propagate into agricultural and hydrological drought stages. The proposed drought propagation-based analytical methodology quantitatively reflects both time lags and transition characteristics of drought events. It is expected to serve as a foundational tool for the development of drought early warning systems and to support decision-making in water resource management.
How to cite: Park, J. H., Bae, S. Y., and Kim, B. S.: Development of a Drought Propagation Methodology Using Meteorological and Climate Data, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-102, https://doi.org/10.5194/ems2026-102, 2026.
Southern Portugal is one of Europe’s climate change hotspots. Given its arid climate, droughts in this region are already long and frequent; however, recent studies estimate that their duration and frequency may increase, putting the region’s water availability at stake. Along its western coast lies the Sado River Basin (SRB), a large, entirely Portuguese catchment that is home to some of Portugal's most important agricultural hubs. Hence, considering how important water is in the SRB, and that precipitation events over the SRB may become scarcer, projecting the basin’s future water availability is a must. This study showcases and explores the estimated flow rates (FRs) of the Sado River in southern Portugal, focusing on the projected conditions expected in a near-term future period, between 2041 and 2060. These FRs were projected under the Shared Socioeconomic Pathways: 1–2.6 W/m2 (SSP1–2.6), 3–7.0 W/m2 (SSP3–7.0), and 5–8.5 W/m2 (SSP5–8.5), using downscaled and bias-adjusted General Circulation Model (GCM) ensemble projections from the Inter-Sectoral Impact Model Intercomparison Project (CHELSA-ISIMIP3b-Sado) as input for hydrological simulations with the Hydrological Simulation Program—FORTRAN (HSPF). To our knowledge, this research marks the first time that future climate projections from the 6th phase of the Coupled Model Intercomparison Project (CMIP6) were used as input to hydrological simulations of the SRB. CHELSA-ISIMIP3b-Sado projections have shown increases in temperature, and decreases in precipitation during autumn and spring months. Moreover, the HSPF simulation outputs have projected reductions in the Sado FRs, which could cause decreases in yearly accumulated riverine water volume (29% under SSP3–7.0 and 33% under SSP5–8.5), as well as increases in summer riverine water deficit (31% under SSP3-7.0). In addition, there is a projected increase in the variability of precipitation during the winter months, and as a consequence, meeting surface water demands of the Sado region may be delayed by up to 22 days. Therefore, in the coming decades, winter precipitation could become even more important for the recharge of reservoirs and the fulfilment of summer surface water needs in the SRB. Considering the predominance of the agricultural sector in the region, the results of this research should prompt local agricultural stakeholders and administrative institutions to enhance winter surface water storage and management in order to meet summer crop irrigation demands.
How to cite: Claro, A., Fonseca, A., Fernandes, A., Menz, C., Almeida, C., Fraga, H., and Santos, J.: How will climate change impact Southern Portugal flow rates? A CMIP6-HSPF modelling approach, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-381, https://doi.org/10.5194/ems2026-381, 2026.
Downscaled key statistics for 24-hr daily rainfall, the wet-day frequency and wet-day mean precipitation, can provide a basis for input to hydrological models with the help of a simple weather generator. A demonstration is provided by the SPRINGS project (https://www.springsproject.eu/) that models the link between climate and public health. It focuses on some specific pilot regions, which include Ghana, Romania, Tanzania and Italy, and the lessons learned may also benefit other parts of the world. For example, water quality and quantity is modelled for the study of the dispersion of waterborne pathogens that lead to diarrhoea in the Volta catchment of Ghana and provide challenges for the water supply in Timisoara, Romania. A model chain employed by the SPRINGS project is presented together with preliminary results, starting from global climate models and ending with specific regional policy advice. The work includes empirical-statistical downscaling (ESD) of large multi-model CMIP5/6 ensembles of global climate models and the estimation of the number of days with heavy 24-hr rainfall. It is important to produce robust and reliable information that can provide a basis for decision-making. While robust and reliable downscaled results imply large multi-model ensembles, various downscaling strategies, and downscaling statistical parameters, hydrological models typically need time series of hourly or daily weather. Hence, an optimal chain of models, involving climate, downscaling, hydrological, epidemiological and economical models, needs to make some compromises. Here, a strategy for cross-disciplinary coproduction of knowledge is presented, with a particular emphasis on climate and hydrology while keeping in mind the needs for practical policy-making.
How to cite: Benestad, R.: A model chain for studying the link between climate and the dispersion of waterborne pathogens involving downscaling and hydrological models, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-472, https://doi.org/10.5194/ems2026-472, 2026.
Rain-on-snow (ROS) events occur when liquid precipitation falls onto snowpack, leading to accelerated snowmelt and the formation of ice layers. These events decrease albedo, intensifying the snow-albedo feedback, and can trigger avalanches and increase flood risk due to the combined effects of rainfall and snowmelt. As climate change shifts precipitation patterns from snow to rain, ROS events are becoming more frequent and intense, making their accurate detection increasingly important. Passive microwave satellite data offer promising potential for detecting ROS events. However, most previous studies have focused on a limited number of frequency channels, typically 19 and 37 GHz. Recently, L-band (1.4 GHz) has gained attention, but a comprehensive evaluation across the full microwave spectrum is still lacking. This study aims to address this gap by assessing the suitability of multiple microwave channels for ROS detection. We analyze the feasibility of identifying ROS events using brightness temperature (Tb) data from the Soil Moisture Active Passive (SMAP) and Advanced Microwave Scanning Radiometer 2 (AMSR-2) satellites (1.4-89.0 GHz), evaluated against in situ observations. Additionally, we use the Snow Microwave Radiative Transfer (SMRT) model to simulate the impact of ROS events on Tb across a wide range of microwave frequencies. The results reveal distinct Tb changes during ROS, confirming microwave sensitivity to liquid water in snow. Using the Normalized Polarization Ratio (NPR), we find that L‑band NPR is particularly effective for identifying ROS events, consistent with SMRT simulations. Our analysis shows that a high NPR indicates a ROS event. We further apply this method across the Arctic and compare the results with ERA5 reanalysis, showing promising agreement. Overall, our findings advance the development of microwave-based approaches for detecting and monitoring ROS events under changing Arctic climate conditions.
How to cite: Kouki, K., Colliander, A., and Riihelä, A.: Detecting rain-on-snow events in the Arctic using passive microwave satellite data, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-162, https://doi.org/10.5194/ems2026-162, 2026.
Rainfall interception loss (Ei) by vegetation plays a crucial role in the regional and global water cycle, as intercepted rainfall does not enter the terrestrial hydrological cycle. Moreover, Ei can account for up to 10–50% of gross precipitation in forested ecosystems. Accurate estimation of Ei is therefore essential for applications such as drought monitoring and prediction, flood prevention, and river discharge prediction.
Several approaches exist to estimate Ei, including empirical methods, numerical models such as the Rutter model, and analytical models such as the sparse Gash and van Dijk–Bruijnzeel models. The latter have proven particularly useful due to their simplicity and their assumption of one storm per day, which enables application at the global scale with minimal input data. When properly calibrated, these models provide reliable long-term estimates of Ei in forested environments. However, when applied at the scale of individual rainfall events without recalibration, they tend to overestimate (respectively underestimate) Ei for low (respectively high) observed values. These discrepancies may arise from uncertainties in parameter estimation, input data errors, conceptual model errors, or observational uncertainties.
To investigate the potential conceptual model errors in the sparse Gash and van Dijk–Bruijnzeel models, new analytical model formulations are introduced, which explicitly represent canopy drainage, based on the exponential drainage function of the Rutter model. Additionally, to systematically investigate the sources of error, a modular modelling framework is employed. This framework allows for multiple configurations of input data, variable implementations (e.g., canopy cover and storage, precipitation intensity and evaporation rate) and the aforementioned model structures. The objective is to develop a robust interception model capable of providing accurate daily estimates at the global scale using readily available data, while remaining applicable to both long-term observational datasets and individual rainfall events.
All combinations of input data, parameter estimation approaches, and model structures are evaluated using both literature-based parameter values and optimised parameter sets. Model calibration and validation are first conducted on a large subset of long-term observational water balance studies reported in literature. The optimised parameter sets are subsequently validated against an independent subset of long-term datasets, as well as event-based water balance observations. In addition, the modular framework enables a systematic assessment of the epistemic uncertainty associated with each modelling choice.
Finally, a global daily dataset of Ei (1980–present) at 0.1° spatial resolution is presented, derived from the best-performing model configuration. This dataset provides new opportunities to investigate the interactions between Ei and climate- and weather-related processes. Furthermore, this work is framed within the activities of the new ESA CCI Land Evaporation project and is a first example of the usage of the modular framework, which will be expanded towards a multiphysics land evaporation model.
How to cite: Delbare, K., Baez-Villanueva, O. M., Bonte, O., Schellekens, J., Zhong, F., and Miralles, D. G.: Improving event-scale and long-term estimates of rainfall interception using a modular analytical modelling framework, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-597, https://doi.org/10.5194/ems2026-597, 2026.
Soil texture characteristics play an important role in determining the amount of soil moisture and the availability of water for evapotranspiration. As a result, they might affect the components of the water and the surface energy balance. An anthropogenic practice that also influences these components is irrigation, enhancing soil moisture, which subsequently leads to an increase in latent heat flux. Moreover, in some irrigation parameterizations, soil texture characteristics such as field capacity and permanent wilting point determine the irrigation amount, potentially affecting the overall impact of irrigation on surface variables. Some studies evaluated the impact of different soil texture maps globally and regionally in climate models, while other studies investigated the impact of irrigation on surface and atmospheric variables. However, the combined effect of soil parameters and irrigation on Earth system modeling, particularly in the context of uniform soil texture maps, has not been investigated yet.
In this study, we assess the impact of two uniform soil texture maps on the surface energy and water budget components over the EURO-CORDEX domain during the years 2017–2018, both in isolation and in combination with irrigation. We choose uniform loam and sand soil type maps to evaluate the potential extent of change in the interactions between soil, climate and irrigation. Moreover, these soil textures are more common in agricultural land. We conduct a set of five simulations with the ICON-nwp model at 3 km resolution, a control run and four experiments. The control run includes the default soil map and no irrigation. Then, we perform two sets of experiments utilizing uniform soil maps, one representing loam and the other sand, both with and without irrigation. The surface energy balance decomposition (SEB) method (Thiery et al., 2017) allows us to determine which components of the surface energy balance influence any temperature changes. Some preliminary results indicate that the uniform sandy soil experiment intensifies the heat wave of 2018. In the same year, experiments with irrigation show that the temperature cooling is stronger over the uniform sandy soil experiment in the Alps, East Europe and Mid-Europe. In contrast, the temperature cooling is stronger over the uniform loamy soil experiment in moisture-limited regions such as the Iberian Peninsula and the Mediterranean. The SEB identifies the variables that mostly influence the temperature changes across different prudence regions.
How to cite: Roque, J., Siebert, S., and Valmassoi, A.: Soil matters: Evaluating soil texture and irrigation impacts in Earth system simulations, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-589, https://doi.org/10.5194/ems2026-589, 2026.
The probable maximum flood (PMF) is a key factor in dam safety and other hydrologic risk assessment contexts. It is often calculated using the probable maximum precipitation (PMP) in conjunction with streamflow prediction models. However, many such models are available, and structural differences between models and their differing data requirements could lead to significant differences in the PMF estimates they generate. We tested this hypothesis by estimating the PMF for the Alouette River in southwest British Columbia, Canada, using three models selected in part for their diversity: WRF-Hydro (complex spatially distributed process-based model), UBCWM in Raven (intermediate-complexity semi-distributed process-based model) and a machine learning model (lumped empirical model). WRF-Hydro provides the most physically complete representation and is also the most resource-intensive and data-demanding. UBCWM offers an effective balance between realism and efficiency, while the ML approach, though less physically interpretable, is computationally inexpensive. The models were developed while holding the meteorological inputs constant, although all three models handle the meteorological input very differently and have different input requirements. In addition to temperature and PMP sequences, WRF-Hydro required wind speed, radiation and specific humidity inputs which provided an additional challenge of how these inputs might change under a probable maximum storm. All three model approaches produced plausible PMF magnitudes, yet these PMF estimates varied by more than 400 m3/s or 28%, in terms of maximum peak flow and by about 8,000 m3 in terms of event volume. We also utilised IES-PEST++ to quantify parameter estimation uncertainty for the process-based models, which fell in the range of 12% to 16%. Since the PMF estimates produced by all three models appear reasonable for practical adoption—and the true value cannot be empirically verified — these results underscore the importance of explicitly accounting for uncertainty in PMF estimation. Additionally, machine learning models are not commonly applied in PMF studies due to well-known concerns around extrapolation to unseen data, we show that even a relatively simple ML model can extrapolate usefully to values more than 40% greater than the data it was trained on. The outcomes of this study provide a quantitative measure of the implications of hydrologic model and parameter selection uncertainty on PMF estimation and imply that a multi-model ensemble could be an effective pathway toward capturing that uncertainty in flood risk assessment processes.
How to cite: Legarth, E., Stull, R., and Fleming, S.: The role of hydrologic model choice in probable maximum flood (PMF) estimation uncertainty , EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-40, https://doi.org/10.5194/ems2026-40, 2026.
The quality and reliability of hydrological forecasts strongly rely on the input in terms of precipitation forecasts. A joint project named "Co-Design of innovations between weather and flood forecasting" strengthens the collaboration between the German Meteorological Service (DWD) and regional flood forecasting centers. One major task aims at condensing, tailoring and communicating the complex meteorological verification to the requirements of hydrologists, especially in the face of an increasing number of weather forecast systems.
The currently most advanced precipitation forecast models of DWD are ICON-D2 EPS and ICON-RUC EPS (Rapid Update Cycle). While ICON-D2 is a well-established limited-area forecast model, the ICON-RUC has a similar setup but is initiated every hour (as opposed to every three hours) and is based on a 2-moments microphysics scheme to better account for strong convective events. In contrast, its forecast horizon is presently limited to 14 hours while the ICON-D2 covers two days.
A question relevant to downstream users is if statistical tendencies can be identified with regard to forecast skills at different weather situations, seasons or spatiotemporal scales. Furthermore, the added value of the ensemble runs compared to the deterministic runs will be assessed in order to promote their importance and use, particularly regarding extreme events.
Here, we assess the performance of different meteorological models for different event severities, depending on the geographic location (Germany: Coast, North, South, low mountain ranges, and Alps), season and/or weather situation. A neighborhood verification approach will be used at first while other verification methods will be assessed and potentially customized as required.
How to cite: Blumenstein-Weingartz, I., Bondy, J., Fundel, F., Hoff, M., Fundel, V., and Keller, J.: Customizing and communicating meteorological verification to flood forecasting centers, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-579, https://doi.org/10.5194/ems2026-579, 2026.
Climate change–induced increases in precipitation and extreme weather events are causing severe urban flooding problems worldwide. In particular, in South Korea, urban flooding has been occurring frequently due to extreme meteorological phenomena such as localized heavy rainfall. To minimize damage caused by urban flooding, it is essential to provide flood risk information that can be practically utilized at disaster sites.
Therefore, this study developed an integrated urban flood information system using radar-based meteorological data. The system consists of five stages: data collection, rainfall prediction, flood prediction, flood risk assessment, and disaster response support. In the data collection stage, real-time meteorological data, CCTV imagery, and urban infrastructure data are gathered. For rainfall prediction, point-based observations and radar data are first applied to a Random Forest–based regression model to perform quantitative precipitation estimation (QPE) bias correction. The generated QPE is then used as input to a deep learning model, NowcastNet, to produce quantitative precipitation forecasts (QPF) at 10-minute intervals for up to 3 hours ahead. For flood prediction, the predicted QPF is applied to a hybrid U-ConvLSTM model that combines U-Net and ConvLSTM architectures to estimate urban flood occurrence patterns and inundation depth. Flood risk is assessed by integrating QPF and flood prediction results with urban flood impact factors, including pedestrian areas, transportation facilities, agricultural and livestock facilities, industrial facilities, infrastructure, and public amenities, to calculate spatially distributed flood risk. Based on the estimated risk, flood risk levels (Attention, Caution, Warning, and Severe) are determined for each administrative district. For disaster response support, a Large Language Model (LLM) is employed to generate response protocols corresponding to each flood risk level. By providing AI-generated response guidelines to users, the system enables rapid and effective disaster management.
All collected and generated information is delivered to users through tables, graphs, and visual outputs within the system interface. The developed system will be further validated through real-world application testing in a Living Lab environment to identify potential issues and enable continuous improvement. Through ongoing advancement, the system is expected to significantly contribute to enhancing urban resilience against extreme weather events and strengthening disaster response capabilities.
Acknowledgements
This work was supported by the Technology Innovation Program (RS202400398858, Development of AI-based urban flood damage risk prediction and evaluation technology for practical use) funded By the Ministry of the Interior and Safety (MOIS, Korea)
How to cite: Kim, D., Lee, Y., Jin, J., and You, J.: Development of an Integrated Urban Flood Information System Using Weather Radar, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-250, https://doi.org/10.5194/ems2026-250, 2026.
As the frequency and intensity of short-duration, intense rainfall increase due to climate change, the risk of urban flooding in areas adjacent to rivers is growing. In particular, urban underground spaces (such as underpasses, underground parking garages, and semi-basement residential units) are structurally vulnerable to water accumulation, and sudden flooding in these areas can lead directly to loss of life. Recent recurring incidents of flooding in underground spaces further highlight the need for proactive drainage management. Drainage pump stations are critical flood prevention infrastructure in low-lying urban watersheds; however, existing rule-based operation methods have limitations in that they respond reactively only after water levels reach a threshold, making it difficult to effectively cope with rapidly changing rainfall patterns. This study proposes an integrated prediction-operation framework that enables real-time, proactive pump operation. Using the Huff time distribution method, a total of 900 rainfall scenarios—each with a total rainfall of 250 mm and a duration of 6 hours—were generated. These were then used as input data in the EPA-SWMM model to simulate runoff. We trained a Bi-LSTM model using the results of the 900 rainfall and runoff simulations to develop an inflow prediction model, and then developed an AI model that operates pumps based on the predicted inflow for the next 8 hours. To analyze performance under extreme conditions, scenarios were simulated where inflow reached 70% and 100% of the total drainage capacity (2,232 m³/s). The analysis compared the current rule-based pump station operation with the AI-based pump station operation developed in this study. The results showed that, unlike the rule-based method which responds reactively after the water level threshold is reached, the AI-based method effectively suppressed peak water levels by proactively operating pumps before the water level rose, utilizing the 8-hour inflow forecast. Furthermore, the maximum water level was reduced by an average of 0.61 m (16.8%) compared to rule-based operation. Notably, under extreme conditions with 100% discharge capacity, a maximum reduction rate of 25.6% was achieved, confirming that the effectiveness of the AI system is further enhanced during extreme rainfall events.
How to cite: An, S. W., Bae, S. Y., and Kim, B. S.: A Study on the Optimization of Real Time Urban Pumping Station Operation Based on Prediction, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-430, https://doi.org/10.5194/ems2026-430, 2026.
This study aims to develop an AI-based long-term daily runoff model that can jointly consider meteorological data and the topographic data of a watershed, and to verify its applicability and performance. To this end, AI models such as LSTM (Long Short-Term Memory) and Bi-LSTM (Bidirectional-LSTM), which can effectively learn the temporal dependency of time-series data, were applied. In selecting the target watershed for model application, several factors had to be considered: the watershed should have representativeness with a certain minimum size, and at least 10 years of meteorological data as well as daily runoff data should be available. Considering these factors comprehensively, the Yongdam Dam watershed, located in the Geum River basin in South Korea and having more than 10 years of long-term data, was selected as the target watershed in this study. As input data, meteorological data that directly affect runoff response, such as daily precipitation and daily temperature, were used. In addition, to reflect the climatic and topographic characteristics of the watershed, the CAMELS (Catchment Attributes and Meteorology for Large-sample Studies) dataset was processed and used for AI model training. Through this, this study was not limited to building a model for a single watershed, but enabled the AI model to learn from data from multiple watersheds, and constructed a long-term daily runoff model that integratively learns meteorological data and topographic data. To evaluate the performance of the trained model, the simulation results of the AI model developed in this study were numerically and visually compared and verified using various hydrological performance indicators and visualizations, including NSE (Nash-Sutcliffe Efficiency) and RMSE (Root Mean Squared Error). This study considered not only meteorological data but also topographic characteristics in AI-based long-term daily runoff simulation, and it is expected that the model can be applied to future mid- to long-term water resources and disaster management.
Acknowledgements
This research was supported by the Specialized university program for confluence analysis of Weather and Climate Data of the Korea Meteorological Institute (KMI) funded by the Korean government (KMA).
How to cite: Kim, B. S., Choi, S. C., and Chu, K. S.: Development and Verification of an AI-based Long-Term Daily Runoff Model using Meteorological and Topographic Data, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-65, https://doi.org/10.5194/ems2026-65, 2026.
Western Norway is one of the regions with the most precipitation in Europe and is regularly hit by floods. As a result of human-induced climate change the frequency of floods in this region is expected to increase. It is therefore important to have good prediction systems in place in order to predict and mitigate the consequences of future floods. However, the hydrometeorological modelling of Western Norway is complicated because of the complex terrain and weather conditions. This project aims to develop a hydrometeorological ensemble model that is optimized for the Western Norwegian geography and climate. The primary objective of this project is to develop a finite element model based on the discontinuous Galerkin Method with the MEPS meteorological model as meteorological forcing. To obtain a high-resolution model while keeping computational costs relatively low, we employ a coupled 1D–2D hydrological modelling approach. In this model rivers are represented by 1 dimensional channels and the land surface with a 2 dimensional triangular grid, which have a resolution of O(101-102m). As we want to prioritize resources to areas with the most complex topography a higher resolution is used in areas with larger topographic gradients. As the goal is to develop a model that is suitable for real-life forecasting we will utilize the ensemble-Kalman filter for data-assimilation, as this is a very effective method. Hydrological processes such as snow melt and precipitation phase are calculated from simple linear parameterizations, while the infiltration is modelled with the curve number method as these are very efficient and have shown to perform well both in research and in real-life forecasting systems. In the ensemble generation step we take the uncertainties in the meteorological forcing, initial conditions and model parameters into account in order to make the resulting ensembles represent the "real" uncertainty as well as possible. The first test case of the model will be the Gaular water catchment, as it has very good coverage by both hydrological and meteorological measurement stations and was hit by a big flood in 2014. The eventual goal is to extend the model to a larger area of Western Norway.
How to cite: Attema, R.: A High-resolution Ensemble Run-off Prediction Model for Western Norway, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-286, https://doi.org/10.5194/ems2026-286, 2026.
Physical hydrological models provide interpretable and physically consistent data, yet their performance can degrade due to parameter uncertainty, forcing biases, or structural simplifications (e.g., simplified parameterizations of groundwater storage, snowmelt or evapotranspiration). In parallel, Long Short-Term Memory (LSTM) recurrent neural networks have demonstrated a high capacity for learning non-linear relationships and temporal dependencies in hydro-meteorological series. This work evaluates the potential of combining both approaches through a hybrid scheme, analyzing predictive performance gains and investigating the contexts where such improvements are limited.
The central hypothesis is that integrating physical model outputs as inputs to the LSTM, alongside meteorological forcings (maximum/minimum temperature and precipitation) and static catchment attributes, should enhance streamflow prediction by incorporating hydrological states and dynamical constraints that are difficult to infer from meteorology alone. Initial results show a clear improvement over pure physical models and superior performance of regional LSTMs compared to local versions (training with all headwater catchments vs with only one). Additionally, the inclusion of static parameters significantly increases the goodness of fit.
However, the added value of incorporating physical models into the hybrid approach is, on average, modest and primarily observed for low flows. To investigate this apparent lack of general improvement, two diagnostic tests were conducted. First, a conditional redundancy test was performed, removing meteorological influence by fitting a linear model (Ridge) to both observed streamflow and physical model output to evaluate residual dependency. Second, a knowledge distillation test was carried out, training an LSTM to reproduce physical model outputs using only meteorological forcings. Results indicate that the LSTM emulates the physical model almost perfectly, suggesting that much of the information provided by the simulator is not independent for the network but is instead derivable from the meteorological forcings.
Finally, regional dependency was analyzed by classifying catchments into five clusters based on climatic and physiographic characteristics. In humid/temperate clusters, high emulation and zero hybrid gain were observed. In dry/transitional clusters, the simulator may introduce bias or noise and, in a transitional cluster, with significant storage, catchments where hybrid modeling consistently improves were identified. Together, these results reveal a predictable "physical signature" that allows for anticipating in which basins the hybrid approach will provide consistent enhancements.
Acknowledgements: This research has been carried out within the framework of project PID2021-126401OB-I00, funded by MICIU/AEI/10.13039/501100011033 and by ERDF, EU.
How to cite: Tacoronte, N., García-Valdecasas Ojeda, M., Castro-Díez, Y., Esteban-Parra, M. J., and Gámiz-Fortis, S. R.: Hybrid hydrological modeling with LSTM and physical simulators, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-303, https://doi.org/10.5194/ems2026-303, 2026.
Evapotranspiration (ET) is a key component of the coupled water and energy cycles, linking land–atmosphere exchange with hydrological response. However, generating spatially continuous and reliable ET datasets remains challenging. This is especially true in regions such as Ireland, where persistent cloud cover limits satellite retrievals and ground-based observations are sparse. Enhancing spatial ET estimation is therefore essential for a range of applications, from hydrological monitoring and water resource management to agricultural planning and climate adaptation.
This study presents a framework for gridded ET estimation over Ireland that combines data-driven correction with physically based land-surface modelling. The framework builds on an adaptive bias correction (AB) method and a dynamic multi-product combination (COM) method developed previously to improve ET estimates from multiple products at 21 station locations across Ireland. While these methods improve ET estimation at point locations, extending them to spatially continuous maps requires suitable interpolation and modelling approaches.
The framework has two complementary components. The first uses spatial interpolation by kriging to extend bias-corrected ET estimates beyond station locations and generate gridded ET maps across Ireland, with particular attention to performance in regions with limited observational coverage. The second uses the Surface Externalisée (SURFEX) land surface model to generate physically based ET estimates, providing an independent process-based representation of ET. Both components are evaluated against eddy-covariance measurements from the Irish National Agricultural Soil Carbon Observatory (NASCO) flux tower network, an extensive network covering a range of land uses across Ireland.
This work forms part of the broader Evapotranspiration for Ireland (ET4I) project, which aims to develop high-resolution gridded ET maps nationwide. The evaluation presented here represents a step towards more reliable national-scale ET mapping in Ireland and offers a basis for improved hydrological, agricultural, and environmental applications in cloud-prone, data-limited environments.
How to cite: Muhammad, H., O'Loughlin, F., Lanigan, G., Finkele, K., and Sweeney, C.: Towards Gridded Evapotranspiration Estimation for Ireland: A Mapping and Modelling Approach, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-699, https://doi.org/10.5194/ems2026-699, 2026.
Extreme rainfall and urban heat events are intensifying across Europe under climate change, with compound occurrences where heatwaves are followed by convective precipitation posing particular challenges for urban flood risk and heat stress management. While the urban heat island effect and its influence on local atmospheric dynamics are well documented, the net impact of urban land cover on compound hydrometeorological extremes remains poorly understood, particularly at the convection-permitting scales needed to resolve city-scale processes.
This study aims to quantify the net effect of urban land cover on a compound hydrometeorological extreme that struck the Netherlands in July 2025: a sustained heatwave followed by a cold front-driven convective rainfall episode over the Amsterdam metropolitan region. Using the Weather Research and Forecasting (WRF) model, this is achieved by comparing simulations against a no-urban control case in which urban land cover is removed from the domain. To ensure this comparison is robust and not contingent on a single model configuration, the analysis is conducted across two dimensions of uncertainty. First, three urban landscape configurations of increasing heterogeneity are tested: the standard MODIS land use classification, a Local Climate Zone (LCZ) dataset from World Urban Database and Access Portal Tools (WUDAPT), and a high-resolution data derived from realistic 3D urban morphology, allowing the urban signal to be assessed independently of how the city is represented. Second, a physical scheme ensemble combining three turbulence treatments (YSU and MYJ planetary boundary layer schemes, and a Large Eddy Simulation approach without PBL parameterization) and three microphysics schemes (WSM6, Thompson, and Morrison double-moment) is used to constrain the sensitivity of the results to physical parameterization choices.
Model performance is evaluated against hourly observations from the KNMI station at Schiphol, covering 2 m air temperature, 10 m wind speed, specific and relative humidity, and incoming solar radiation, with precipitation evaluated against radar observations. Preliminary results from the baseline configuration show strong agreement with observed diurnal temperature cycles (R2 = 0.90, RMSE = 1.85°C ), while systematic biases emerge in wind speed and near-surface humidity, and simulated convective precipitation is delayed and underestimated relative to observations.
These initial findings establish a credible baseline from which the full ensemble analysis will provide a rigorous and representation-independent estimate of how urban land cover modulates heatwave intensity, convective triggering, and precipitation distribution during compound hydrometeorological extremes.
How to cite: Smook, C., Droste, A., Schleiss, M., and Chen, X.: Urban Land Cover Effects on a Compound Heatwave and Convective Rainfall Event: A WRF Ensemble Study over Amsterdam, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-473, https://doi.org/10.5194/ems2026-473, 2026.
Despite growing concern over climate‑driven hydrological extremes, nationwide and regime‑specific assessments of future drought and river‑discharge changes in Japan remain limited. In this study, we examined the impacts of climate change on river discharge at major hydropower dam sites in Japan using hydrological simulations based on a 5-km resolution ensemble climate dataset. We applied clustering to the seasonal evolution of river discharge and evaluated the impact of each cluster, thereby revealing strong regime-dependent differences that challenge the feasibility of uniform adaptation. By integrating analyses of interannual changes in hydrological variables, we found a pronounced decline in water availability along with increases in both flood and drought frequencies. Notably, the number of consecutive hydrological drought increased by approximately 1.3 times under 2K warming and 1.7 times under 4K warming relative to present-day climate conditions. Amplified drought conditions emerged in regimes typical of the Sea of Japan side of western Japan and heavy-snow mountainous regions, where concurrent increases in evapotranspiration and decreases in precipitation were likely to intensify drought risk. Seasonally, hydrological drought increased most in December–January on the Pacific side and in August–September on the Sea of Japan side. These regional and seasonal contrasts were closely linked to global sea surface temperature warming patterns that modulate atmospheric large-scale circulation, thereby governing drought severity in Japan. Collectively, our findings highlight the need to tailor water-use adaptation strategies to specific regional and seasonal river discharge regimes under future climate warmings, underscoring the importance of regime-aware planning for hydropower and water resource management.
How to cite: Ohba, M., Arai, R., yanagihara, H., and Kawazoe, S.: Climate Change Impacts on River Discharge Regimes and Hydrological Drought in Japan: Insights from Nationwide Hydropower Climate Projections, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-49, https://doi.org/10.5194/ems2026-49, 2026.
Conventional single-threshold methods for detecting hydrological extremes conflate absolute water deficits with relative seasonal anomalies in seasonally fluctuating lakes, a distinction that is becoming increasingly critical as these systems undergo unprecedented regime shifts driven by climate change and water infrastructure expansion. This study develops a dual-threshold framework that integrates fixed and variable thresholds to independently quantify these dimensions, enabling robust attribution of regime shifts in seasonally fluctuating lakes. The fixed threshold captures absolute extremes based on historical percentiles, while the variable threshold accounts for seasonal hydrological norms, thereby disentangling water deficits driven by external forcing from those arising from seasonal variability. Applied to Poyang Lake (China’s largest freshwater lake and a Ramsar-listed wetland) using 1960–2020 hydrometeorological records and 1984–2020 Landsat imagery, the framework reveals a fundamental transition from flood-dominated to drought-dominated extremes following the 2003 operation of the Three Gorges Dam upstream: extreme floods decreased from 7 events in 1960–2002 (43 years) to 2 events in 2003–2020 (18 years), with delayed midsummer timing and contracted inundation extents, while extreme droughts increased from 3 to 6 events over the same periods, manifesting as persistent autumn-winter phenomena. SHAP-based attribution demonstrates that absolute extremes are governed by basin-scale hydrodynamic forcing (Yangtze backwater and tributary inflow), whereas relative anomalies reflect enhanced atmosphere–hydrology coupling with potential evaporation as a key driver. By decoupling hydrodynamic and climatic controls on dual scales, this work advances mechanistic understanding of asymmetric extremes and provides a transferable diagnostic tool for adaptive management in seasonally fluctuating lakes under non-stationary conditions worldwide. The framework offers a systematic approach applicable to other large lakes facing similar hydrological alterations.
How to cite: Dai, X. and Yang, G.: Attenuated Floods yet Amplified Droughts: Dual-Threshold Attribution in Poyang Lake, China, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-436, https://doi.org/10.5194/ems2026-436, 2026.
Floodplain methane (CH₄) emissions constitute a substantial component of the global CH₄ budget. Nevertheless, their response to the increasing frequency of extreme drought events remains insufficiently understood, largely owing to the pronounced variability associated with alternating wet–dry hydrological regimes. To address this research gap, we conducted two years of in-situ CH₄ flux measurements across alternating hydrological cycles (2022–2023) in the Poyang Lake floodplain wetland, a period during which the region experienced an extended drought. Our results reveal that CH₄ emissions during non-flooding periods (1.82 ± 1.36 mg CH₄ m⁻² h⁻¹, mean ± standard deviation) were significantly higher than those during flooding periods (1.26 ± 0.96 mg CH₄ m⁻² h⁻¹). Notably, under drought conditions, CH₄ fluxes during the autumn growing season (2.04 ± 1.43 mg CH₄ m⁻² h⁻¹) were 35% greater than those observed in the spring growing season (1.51 ± 1.21 mg CH₄ m⁻² h⁻¹). Further analysis indicates that, apart from air temperature, CH₄ fluxes were primarily regulated by vegetation during non-flooding periods, whereas during flooding periods they were predominantly governed by water level fluctuations and inundation duration, factors that modulate key biogeochemical processes. The enhanced temperature sensitivity of CH₄ emissions emerged as a critical mechanism underlying the elevated autumn emissions relative to spring, a pattern directly attributable to the shortened flooding duration in the Poyang Lake floodplain. These findings highlight the pivotal role of extreme drought in altering hydrological regimes and CH₄ emission dynamics in floodplain wetlands, with important implications for predicting wetland responses under future climate change scenarios.
How to cite: Wan, R., Wang, X., Yang, G., Zhao, X., and Li, B.: Hydrologic Processes Drive Methane Fluctuations in a Large Subtropical Floodplain, EMS Annual Meeting 2026, Utrecht, Netherlands, 6–11 Sep 2026, EMS2026-433, https://doi.org/10.5194/ems2026-433, 2026.
Supporters & sponsors
