Orals: Fri, 2 May | Room C
Satellite observations have frequently been used for river discharge estimation, particularly in ungauged catchments. The largest challenge for producing continuous time series of river discharge, e.g. with daily time steps, is typically the sporadic nature of satellite observations. Various methods, including spatio-temporal densification of satellite-derived water levels along river networks, have been proposed to address this issue. However, these estimates often suffer from high uncertainties.
Here, we present a novel approach, using both satellite-derived water levels (SWL) and reflectance indices (SRI) to estimate river discharge across 46 river stations in the Mediterranean region. We utilize Long Short-Term Memory (LSTM), known for their efficiency in modeling complex temporal relationships. While LSTM models have been widely applied in rainfall-runoff modeling within the hydrology community, few studies have explored satellite-derived river states as inputs due to their uncertainties and temporal discontinuities.
Gap filling was necessary for SWL and SRI datasets, originally available at intervals ranging from roughly 5 to 30 days. This was accomplished based on freely available discharge from the European Flood Awareness System (EFAS). For each catchment, we compiled daily dynamic variables. Besides the gap-filled SWL and SRI data, this included observed river discharge, as well as precipitation, temperature and potential evapotranspiration from global datasets.
For benchmarking purposes, we set up and calibrated lumped hydrological models for the same 46 catchments, using the same climate data as forcing. Results show that LSTM models outperformed lumped hydrological models in many catchments when using only climate variables as inputs, i.e. when being informed by the same dynamic data as the lumped rainfall-runoff models. The performance of LSTM models can be further improved with the inclusion of SRI and SWL. Shapley Additive Explanations (SHAP) analysis indicated that while climate variables are the most informative for discharge estimation, SRI and SWL also contribute significantly, but varying across individual stations.
The method integrates satellite-derived river states for improved river discharge estimation, while still allowing ingestion of climate input data. This goes beyond conventional hydrological models being forced by climate data only, or also existing densification algorithms for SWL, only using satellite observations
How to cite: Liu, J., Koch, J., Sivelle, V., Massari, C., Tarpanelli, A., and Schneider, R.: Deep Learning Estimation of River Discharge based on Satellite Observations in Mediterranean Catchments, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-16988, https://doi.org/10.5194/egusphere-egu25-16988, 2025.
The Global Climate Observing System (GCOS) identifies river discharge as an Essential Climate Variable (ECV), critical for understanding climate dynamics and managing water resources (GCOS, 2022). However, no satellite instrument currently exists to directly measure river discharge, which must instead be estimated indirectly. The ESA River Discharge Climate Change Initiative (CCI) precursor project (https://climate.esa.int/en/projects/river-discharge/) addresses this challenge by developing innovative methodologies based on satellite remote sensing data.
Four complementary approaches are being explored: (1) the use of long-term satellite radar altimeter time series of water surface elevations, combined with rating curves to estimate discharge; (2) The use of satellite imagery data to obtain river width, combined with rating curves to estimate discharge; (3) multispectral sensor data in the near-infrared (NIR) band, used to analyze river flow variability through the reflectance ratio between wet and dry pixels; and (4) a hybrid approach combining these two techniques. Radar altimeters offer the advantage of weather-independent measurements, while multispectral sensors provide higher temporal resolution but are limited by cloud cover.
This proof-of-concept study focuses on 54 locations across 18 river basins, spanning 2002–2022. The sites represent a variety of climatic zones, drainage areas (from 50,000 km² to the Amazon basin), levels of human activity, and availability of in situ data. The project showcases the potential for satellite-based global river discharge estimation, validated through comparisons with on-the-ground measurements.
This presentation will outline the methodologies employed, the computed discharge time series along with their validation during the first Phase of this precursor project (2023-2024), the objectives for the second phase, which has just started, and the progress achieved.
How to cite: Biancamaria, S. and Gal, L. and the CCI River Discharge: The River Discharge Climate Change Initiative precursor project, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11436, https://doi.org/10.5194/egusphere-egu25-11436, 2025.
Accurate river discharge monitoring is essential for understanding hydrological processes, yet the availability of in situ measurements is increasingly limited due to a declining number of operational gauges and temporal gaps in gauge records. Satellite altimetry offers a robust alternative to address these limitations. Here, we introduce the Satellite Altimetry-based Extension of the global-scale in situ river discharge Measurements (SAEM) dataset, which integrates data from multiple satellite altimetry missions to estimate river discharge and enhance global hydrological monitoring networks. Our analysis evaluated 47,000 discharge gauges and successfully derived height-based discharge estimates for 8,730 gauges, expanding the coverage of current remote sensing datasets by a factor of three. These gauges collectively represent approximately 88% of the globally gauged discharge volume. The SAEM dataset achieves a median Kling-Gupta Efficiency (KGE) of 0.48, demonstrating superior performance compared to existing global datasets.
In addition to discharge time series, SAEM offers three supplementary products: (1) a catalog of Virtual Stations (VSs) with metadata, including geographic coordinates, altimetry mission details, distances to discharge gauges, and quality flags; (2) for VSs with quality-controlled discharges, we provide IDs from L3 databases such as Hydroweb.Next (formerly Hydroweb), the Database of Hydrological Time Series of Inland Waters (DAHITI), the Global River Radar Altimeter Time Series (GRRATS), and HydroSat, and for VSs without corresponding time series in these L3 products, we have generated water level time series (SAEM WL) as an additional product; (3) rating curves that map water levels to discharge using the Nonparametric Stochastic Quantile Mapping Function approach. The SAEM dataset can enhance hydrological research, support water resource management, and allow addressing complex water-related challenges in the context of a changing climate.
How to cite: Saemian, P., Elmi, O., Stroud, M., Riggs, R., Kitambo, B. M., Papa, F., Allen, G. H., and Tourian, M. J.: SAEM: Satellite Altimetry-based Extension of global-scale in situ river discharge Measurements, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8247, https://doi.org/10.5194/egusphere-egu25-8247, 2025.
Accurate monitoring of river profiles and discharge is critical for understanding hydrological dynamics and water resource management. However, current nadir radar altimetry is constrained by orbital spacing and the meandering nature of rivers. This study explores the potential of off-nadir processing methods using Fully-Focused SAR (FFSAR) data from Sentinel-3A/-3B and Sentinel-6A, complemented by observations from the Surface Water and Ocean Topography (SWOT) mission.
An automated off-nadir processing algorithm was developed to estimate time-evolving river profiles in the cross-track direction. By applying off-nadir slant range corrections to retracked ranges, we expanded the effective cross-track measurement range to 6.6 km for Sentinel-3A/-3B and 9.3 km for Sentinel-6A. Validation against in-situ data from the Rhine, Danube, and Oder rivers demonstrated water level accuracy, with a standard deviation of difference (STDD) between 0.04 m and 0.09 m. Slope measurements exhibited a precision of 0.7–1.3 cm/km. Comparative analyses of river profiles over 60-km channels revealed STDD values of 0.14 m for Sentinel-6A and 0.19 m for Sentinel-3B.
Additionally, discharge in Rhine, Danube, and Oder rivers are computed from FFSAR off-nadir (2016-2024) and SWOT (2023.04-2024) using Metropolis-Manning (MetroMan) algorithms. Both of the discharge are evaluated against gauges. The results were evaluated using the NRSME and NSE metrics on the reach, showing good agreement between discharge from FFSAR, SWOT and gauges. This study is to prove that the discharge from SWOT can be extended to earlier periods using nadir altimetry data collected prior to 2023.
How to cite: Chen, J., Fenoglio, L., and Kusche, J.: Monitoring river profile and discharge with FFSAR off-nadir and SWOT, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13634, https://doi.org/10.5194/egusphere-egu25-13634, 2025.
Satellite nadir altimetry has been a powerful technique for understanding oceans and seas over the past few decades. However, over rivers and small inland water bodies it produces noisy observations, which can result in gaps or erroneous measurements in water level time series. In this study, we aim to identify and correct anomalous measurements through reprocessing at the Level 1B (L1B) stage of the satellite altimetry processing chain.
To this end, we first detect abnormal waveforms that lead to anomalous water level measurements by analyzing various parameters related to the satellite's altimeter like AGC parameter and tracker range, and also waveform shape features. These waveform features include the number and location of peaks, noise level, kurtosis, centre of gravity, and peakiness. Abnormal waveforms are identified through an analysis of the distribution of these features.
While previous studies focused solely on L2 measurements to retrack multi-peak and noisy waveforms, we propose a robust strategy to regenerate abnormal waveforms within the L1B SAR processing chain by eliminating unwanted backscattered power. This approach incorporates the Fully-Focused Synthetic Aperture Radar technique into the L1B processing chain, dividing the illumination time into smaller stacks comprising multiple beam looks.
Due to factors such as antenna side lobe gain, wide antenna footprints, and environmental unevenness, some beam looks may exhibit undesired patterns. Our proposed approach addresses this issue by comparing the power of individual stacks with an analytically-derived reference waveform and assigning weights to each stack based on their similarity to the reference waveform. This reduces the impact of unwanted components in the final waveform and enables the regeneration of detected abnormal waveforms for inland waters.
We applied the proposed method to Sentinel-3A, Sentinel-3B, and Sentinel-6MF measurements over 6 lakes and reservoirs of various sizes and validated the results against in-situ data. The validation demonstrates that the water height time series obtained from regenerated waveforms match significantly better with in-situ measurements. Specifically, the accuracy of the water level time series, measured in terms of RMSE, improved by around 60% for the selected case studies after applying retracking on newly generated waveforms.
How to cite: Sneeuw, N., Khalili, S., Tourian, M. J., Elmi, O., Engels, J., and Sörgel, U.: Recovering noisy measurements over inland water bodies by regenerating L1B SAR altimetry waveforms using a segment-weighted Fully-Focused - Synthetic Aperture Radar (swFF-SAR) processing scheme, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-5134, https://doi.org/10.5194/egusphere-egu25-5134, 2025.
Monitoring water levels in lakes and reservoirs forms a critical component of sustainable water resource management, particularly in regions where direct measurements are costly, time consuming or impossible. Traditionally, ground-based sensors are used as the primary means of water level observation. In recent past, remote sensing has emerged as a vital alternative for areas that are inaccessible, have sparse monitoring infrastructure or located in the transboundary regions. However, recent studies have highlighted limitations in temporal resolution required for immediate responses to water-related conflicts. We present here a novel methodology for enhancing the temporal resolution of water level time series derived from altimetry satellites by integrating data from other satellite types, such as optical (Harmonized Landsat Sentinel-2) and SAR (Sentinel-1), particularly for small and complex inland water bodies. Our approach leverages DEM-driven water masks with 1-meter intervals to systematically calculate reflectance values at various elevation levels, identifying water levels based on the most significant reflectance differences. Unlike static methods with fixed thresholds, our methodology dynamically adjusts thresholds according to regional and temporal variations, ensuring greater accuracy and adaptability. To mitigate the limitations of optical data, such as cloud coverage during the wet season, we integrated SAR data as a further enhancement to the developed approach. We tested this methodology on four reservoirs in South Korea—Chungju, Andong, Daecheong, and Juam—representing diverse hydrological characteristics. The results demonstrated significant improvements in the accuracy of water level estimation, even for highly variable and small water bodies. Further, the proposed method shows robustness across multiple satellite datasets while effectively addressing data gaps, providing a scalable and globally applicable framework for advancing water level monitoring. The approach underscores its potential to enhance hydrological assessment and water management, particularly in under-monitored regions.
How to cite: Han, K., Kim, S., Mehrotra, R., and Sharma, A.: Enhanced Water Level Monitoring for Small and Complex Inland Water Bodies Using Optical and SAR Retrievals, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-1200, https://doi.org/10.5194/egusphere-egu25-1200, 2025.
The water balance equation describes the exchange of water mass between land, ocean and atmosphere. Being able to close the water balance gives confidence in the ability to model and/or observe spatio-temporal variations in the water cycle and its components. At basin scale, the water balance equation (DTWS = P - ET - Q) compares derived total water storage (DTWS) with precipitation (P), evapotranspiration (ET) and runoff (Q). Many studies compare GRACE-based DWTS observations with P and ET datasets, and Q from Land Surface Model (LSM), due to the lack of in situ discharge observations. For some basins, human activities, glacier, reservoir and lake impact on the water cycle is not or poorly modeled by the LSM. In this case, the water budget may close due to compensation errors, for example between Q and ET.
In this study, we propose to evaluate the consistency of budget closure with Q computed from satellite altimetry data, which might have better accuracy than discharge from LSM. We will use the altimetry-based discharge products from the ESA CCI river discharge project (https://climate.esa.int/en/projects/river-discharge/), recently available. DTWS is evaluated from the CNES GRACE-GRACEFO L3 dataset. This dataset is an ensemble of 120 different solutions combining the state-of-the-art in terms of GRACE L2 data and corrections. The spread within the ensemble aims to cover the uncertainty in DTWS estimates. The dataset has a monthly resolution of 1 degree.
In order to evaluate the best combination of datasets to close the water balance, we will use more than 15 precipitation datasets using the FROGS database (https://frogs.ipsl.fr/) and more than 8 evapotranspiration datasets (GLDAS, ERA5-Land, GLEAM, SynthesizedET, SSEBop, MOD16, BESS V2, FLUXCOM). This ensemble-based approach will also enable to assess the dispersion of these precipitation and evaporation data for each basin. We evaluate the budget closure using different metrics (NSE, KGE, RMSD etc…) at 18 basins of different climate, latitude and size over 2002 to 2019.
Finally, we will compare the 18 water budget closures with those obtained with discharge computed from LSM, like GLDAS or ISBA/CTRIP, to assess the benefits of using altimeter-based discharge for the water budget closure.
How to cite: Lefebve, J., Biancamaria, S., Blazquez, A., Munier, S., and Zakharova, E.: Water budget closure assessment of 18 various basins combining GRACE and altimetry data, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-6660, https://doi.org/10.5194/egusphere-egu25-6660, 2025.
Groundwater flowing through the fractured bedrock composing most mountain ranges has been increasingly recognized as a vital source of freshwater for both low-elevation communities and mountain ecosystems, maintaining streamflow and constituting a large portion of recharge to lowland aquifers used to support human activities. Despite the growing awareness of groundwater’s role in mountain hydrology and the potential impacts of climate change on mountain groundwater, it remains a challenge to study the dynamics of mountain aquifers, largely due to the low density of observational wells and challenges in characterizing the mountain block over large areas and depths. Here, we report on a new approach to characterize the flow and hydraulic properties of mountainous aquifers at a mountain range scale. We utilize high-precision Global Navigation Satellite Systems (GNSS) observations of vertical crustal displacement produced by the redistribution of freshwater on or near the Earth’s surface to estimate changes in groundwater storage within the Sierra Nevada and Cascades Range of the western United States with high spatial (10s of kilometer) and temporal (daily) resolution over the past two decades. We find that on average groundwater annual recharge is less than discharge, driving long-term declines in groundwater storage over the last 19 years. Furthermore, we find groundwater recharge to be up to 3x more variable than groundwater discharge in these mountainous areas, suggesting that mountain aquifers release a relatively constant amount of water to streams and adjacent lowland aquifers despite fluctuating recharge conditions. Utilizing identified periods of groundwater discharge, we characterize the hydraulic conductivity, storativity, and flow path length of these groundwater systems using fluid diffusion models in combination with our GNSS-inferred groundwater estimates. Our initial estimates of these parameters reveal relatively high values of bedrock conductivity (~1x10-3-1x10-4 m/s) relative to expected values based upon each region’s bedrock lithology, suggesting that areas with highly fractured bedrock as well as saprolite may exert a strong control on groundwater discharge at the mountain range scale. Furthermore, our results indicate that groundwater flow paths can span lengths on the order of 100s-1000s of meters, supporting the notion that groundwater can flow over extended areas supporting recharge at both a local and regional scales. Our work seeks to provide a new set of tools for hydrologists to investigate these often poorly understood systems.
How to cite: Swarr, M., Argus, D., Martens, H., Hoylman, Z., Oliver, B., and Gardner, W. P.: Geodetic Constraints on Mountain Bedrock Aquifer Flow and Diffusivity, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-14014, https://doi.org/10.5194/egusphere-egu25-14014, 2025.
The Gravity Recovery and Climate Experiment (GRACE) mission and its Follow-On (GRACE-FO) mission provide observations of terrestrial water storage (TWS) dynamics on regional to global scales. However, they lack high spatio-temporal resolution, making them challenging to interpret different gravity field products. A join collaboration between the National Aeronautics and Space Administration (NASA) and the European Space Agency (ESA), initiated a decade ago, is known as the Mass- change And Geosciences International Constellation (MAGIC). The ultimate aim of this collaboration to improve the current models and launch new high resolution missions in order to improve capacity for monitoring extreme events such as natural hazards, droughts and floods. The ultimate aim of this collaboration to improve the current models and launch new high resolution missions in order to improve capacity for monitoring extreme events such as natural hazards, droughts and floods. The primary objective of this study to examine the impact of improving spatial-temporal resolution of NGGM and MAGIC on rainfall estimation by developing multiple synthetic experiments on a European scale. The study employed SM2RAIN by inverting the soil water balance equation to estimate the rainfall accumulated between two consecutive TWS measurements. Initially, the ERA5L based TWSA at daily time scale was incorporated into SM2RAIN to check reliability of the model against ERA5L precipitation with spatial resolution of 100 km over Europe with range of latitudes 30 to 60°N and longitudes 10°W to 50°E.. The results shows SM2RAIN exhibited satisfactory performance at a daily temporal resolution, with mean values of R, RMSE, BIAS (0.85, 13.76, -0.29) against ERA5L precipitation. Based on statistical analysis, SM2RAIN-simulated rainfall shows good agreement across the most of Europe except in some areas of the northern Italy, northeastern states (Estonia, Latvia) and costal regions of Norway . Subsequently, synthetic experiments were developed by aggregating the daily ERA5 based TWS data into 5-day intervals which led to a decline in model performance against SM2RAIN-simulated rainfall as evidenced by all statistical measures with mean values of (0.73, 18.41 and -0.43) for CC, RMSE and BIAS respectively. In another experiment where inclusion of a target error 4.2 mm into 5-day TWS further reduce the model ability to access rainfall patterns, resulting in lower CC values across Europe, with the majority of areas showing below 0.3. At a threshold error 42 mm, the model’s performance of model significantly deteriorated in order to capture meaningful rainfall patterns with mean values of CC = 0.04 and RMSE 26.30. The results shows that degrading temporal resolution and larger error make the model quite difficult to capture and represent meaningful rainfall patterns, as the error completely overshadows the underlying dynamics captured in the SM2RAIN-simulated rainfall. The results of the study clearly highlight the benefit of NGGM and MAGIC in improving our capability to estimate various hydrological components relying on satellite data as inputs.
How to cite: Liaqat, M. U., Brocca, L., Leopardi, F., Camici, S., Ansari, R., and Garcia, J. G.: Beyond GRACE: Evaluating the Benefits of NGGM and MAGIC for Rainfall Estimation on a European scale, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15307, https://doi.org/10.5194/egusphere-egu25-15307, 2025.
The Cuvette Centrale wetland complex, located in the central depression of the Congo Basin, is a critical component of regional and global carbon and water cycles. The hydrological processes controlling these wetlands, of which 16.8 Mha are classified as peatlands, remain poorly understood, due to complex interactions between the Congo River, its tributaries, variable rainfall patterns, and anthropogenic influences. Here, we address this knowledge gap by interpreting the updates introduced by microwave data assimilation. The employed land surface data assimilation framework follows the setup of the 9-km Soil Moisture Active Passive (SMAP) Level-4 Soil Moisture algorithm that includes a land surface model specifically designed to simulate peatland hydrological processes (PEATCLSM).
First, we update PEATCLSM hydrological parameters for the Congo Basin peatlands, using a new event-based approach named: HYdrological PArameterization of in situ water level dynamics using SATellite-based precipitation (HYPASAT). Along with further adjustments to the PEATCLSM module, we significantly reduce the dry bias present in water level simulations with a previous model version. Second, we assimilate L-band brightness temperature (Tb) observations from the Soil Moisture and Ocean Salinity (SMOS) satellite mission for the period 2010 through 2022. We demonstrate that the assimilation of SMOS L-band Tb observations into PEATCLSM further enhances the accuracy of water level estimates, indicated by improved temporal correlations with in situ data. Finally, we present an analysis of the data assimilation state updates, which showed widespread systematic patterns that were linked to observed, but unmodeled, upstream river stage anomalies. The data assimilation results highlight the sensitivity of the hydrology of the Congo Basin peatlands to local and upstream rainfall variability, as well as river dynamics, and thus river management. Therefore, we emphasize the need for integrated hydrological and land management approaches in the peatland region.
How to cite: Apers, S., De Lannoy, G., Cobb, A. R., Dargie, G. R., Davenport, I., Reichle, R. H., and Bechtold, M.: Hydrological dynamics of the Cuvette Centrale peatlands: insights from enhanced land surface modeling and SMOS L-band data assimilation, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12589, https://doi.org/10.5194/egusphere-egu25-12589, 2025.
Remote sensing techniques are crucial for a continuous and comprehensive monitoring of inland waters. In particular, recent advances in satellite radar altimetry have allowed the observation of an increasing number of small and medium-sized lakes and reservoirs, even in complex topographies. The advent of nadir radar altimeters operating in Synthetic Aperture Radar (SAR) mode has significantly improved the resolution of observations in the along-track direction, from several kilometers in conventional pulse-limited altimeters to hundreds of meters in close-burst altimeters when applying unfocused SAR (UFSAR) processing, as is the case in the Sentinel-3 satellite constellation.
Inversion methods for estimating geophysical parameters, such as Lake Water Level (LWL), from the backscattered altimetry signal are commonly called retrackers. These retrackers can be empirical, such as the widely used OCOG method or physics-based, i.e. a background waveform model is derived from the theoretical knowledge of the microwave scattering process and then fitted to the backscattered signal received on-board. Several retrackers of the second type have been developed for processing conventional radar observations, such as the Brown-type models, and also for UFSAR observations in the case, for example, of the SAMOSA model. However, one of the limitations of physics-based retrackers concerns the assumption that the radar footprint is completely covered by water, as is the case for the ocean. This assumption, which applies to large lakes, starts to degrade the accuracy of the retrieved geophysical parameters when monitoring smaller water bodies. For this reason, a retracker based on numerical simulations tailored to UFSAR observations was proposed for inland waters [1]. This latter model has the advantage of taking into account a priori knowledge of the lake contour (for example, the Prior Lake Database [2]), and, thus, only the in-water areas of the radar footprint contribute to the simulated waveform. A preliminary assessment of the performance of this retracker solution indicated a LWL accuracy better than 10 cm in most of the lakes [3].
Considerable effort has been put into making that retracker robust enough to generate demonstration altimetry products for the hydrological community. These Level-2 products, expected to be available in the Copernicus Data Space Ecosystem in early 2025, cover the entire Sentinel-3 mission time period (both A and B satellites) and provide information on more than 1200 lakes worldwide. This work will present the physical retracker basis and methodology, as well as the content and format of these new radar altimetry products, ready for use by scientific users. Finally, an extensive comparison with in-situ data will be performed to characterize the expected accuracy, with a special focus on time series for some specific lakes.
[1] Boy, F., et al., 2021. Improving Sentinel-3 SAR mode processing over lake using numerical simulations. IEEE Transactions on Geoscience and Remote Sensing, 60, 1-18.
[2] Wang, J., et al., 2023. The Surface Water and Ocean Topography Mission (SWOT) Prior Lake Database (PLD): Lake mask and operational auxiliaries. Authorea Preprints.
[3] Yanez, C., et al., 2023. Performance Assessment of Lake Water Level Estimation from Sentinel-3 SAR Data over 1000 Lakes and Reservoirs Worldwide. 2023 IEEE IGARSS, 2870-2873.
How to cite: Yanez, C., Wery, F., Calmettes, B., and Boy, F.: Introducing New Radar Altimetry Products from Sentinel-3 for Inland Water Monitoring, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2834, https://doi.org/10.5194/egusphere-egu25-2834, 2025.
Changes in lake water levels are closely related to climate change and can also reflect information about local human activities. Therefore, obtaining high temporal resolution time series of lake water levels is necessary for accurately analyzing hydrological changes. However, the existing methods mainly focus on the long-term changes in lake water levels, with less attention paid to short-term changes in lake water levels. In this paper, we proposed a new method to construct high temporal resolution lake water level time series by fusing multi-source altimetry satellite data based on Kalman filtering and using the MissForest algorithm to combine meteorological data (Kalman Fusion-MissForest water level, KF-MFWL). The accuracy of KF-MFWL was validated using gauge data , as well as compared with HYDROWEB and DAHITI. Finally, a dataset of daily lake water level time series for the Qinghai-Tibet Plateau from 2019 to 2021 has been compiled, and the driving factors influencing water level changes were analyzed. Our result shows that the KF-MFWL time series is comparable to that of HYDROWEB and DAHITI, but with a much higher temporal resolution. The annual rate of water level change for 264 lakes in the Qinghai-Tibet Plateau is 0.021m/y. Among them, the water level of 82 lakes has significantly increased with an average annual change rate of 0.171m/y, while that of 55 lakes exhibits a remarkable decrease with an average annual change rate of -0.145m/y. This study can provide an important data basis for water resource management in the Qinghai-Tibet Plateau region.
How to cite: An, Z., Jiang, W., and li, Z.: KF-MFWL: A High-Resolution Time Series Construction Algorithm for Lake Water Levels Based on Multi source Altimeter Satellites and Meteorological Data Fusion, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15098, https://doi.org/10.5194/egusphere-egu25-15098, 2025.
Switzerland is today rich in water, and retreating glaciers give way to new landscapes with lakes as an important element.
As part of the Collaborative Research Center (SFB 1502) funded by the German Research Foundation (DFG), a project is being carried out to analyze surface water storage change and river discharge using data from the latest generation of satellite altimetry. The goal is to monitor the impact of land use change on the water cycle, here on the exchange of water between rivers, lakes and reservoirs.
We distinguish two groups of lakes: natural lakes and reservoirs. The first group includes both ancient large lakes of small variations related to long-term changes in temperature and small lakes formed in the deglaciated area rapidly changing and related to glacier melting. The second group includes reservoirs with large water variations related to resource management, like hydropower and irrigation. We generate a lake inventory for modern times and trace them in the nadir-altimetry and wide-swath altimetry to monitor seasonal and intra-annual variability of surface between 2016 and 2024.
Fully Focused SAR nadir-altimeter processed data at 80 Hz, with along-track spacing of 85 meters are chosen together with SWOT swath-altimeter HR products. Lakes with area larger than 0.5 km**2 are used. Only ten of the more than eighty water bodies observed by SWOT in the region are detected by nadir-altimetry, showing that swath-altimetry is best suited for this application. Space-derived height and area time-series evaluated against in-situ, bathymetrie and Sentinel-1 images have higher accuracy in the natural Murnersee (1 cm bias and 3 cm standard deviation) than in reservoir Lake de Joux (31 cm bias and 13 cm stdd). The surface area has mean accuracy of 10%, highest change found is 100 m in hydroelectric reservoirs and 10 m in irrigation reservoirs. Most reservoirs are operated in a network.
We look at 70 water bodies with variations larger than 10 m, assuming that larger variations are related to water management. Annual minima are in May for hydroelectric and in November for irrigation reservoirs, while in natural lakes the annual maximum is in Summer. The amplitude of storage change in hydroelectric reservoirs is 70% higher than in irrigation reservoirs and is 80% higher than in natural lakes. The water budget in catchments is analysed comparing to land runoff and snowmelt from CLM model which is not including irrigation and hydropower.
This study hightlights the importance of the new satellite altimeter observations to study climate change, land and water use.
How to cite: Fenoglio-Marc, L., Chen, J., Naz, B., Frappart, F., and Kusche, J.: Monitoring surface water storage change in lake and reservoirs , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-18942, https://doi.org/10.5194/egusphere-egu25-18942, 2025.
Surface water storage in inland water bodies is crucial for understanding water storage dynamics, which directly impact the hydrological cycle. Traditional in situ methods face limitations in capturing these dynamics, especially for smaller and remote water bodies, highlighting the need for alternative approaches. Remote sensing techniques, particularly the combination of Global Navigation Satellite System reflectometry (GNSS-R) and radar altimetry, offer significant opportunities to overcome these challenges. By leveraging the unique capabilities of CYclone Global Navigation Satellite System (CYGNSS) and radar altimetry missions, it is possible to monitor water surface extent and elevation over time, enabling continuous estimation of surface water volume in both large and small water bodies.
This study employs the CYGNSS satellite constellation to generate water masks from Delay Doppler Maps (DDMs) for Gandhisagar reservoir, Ghaghra river in Ayodhya, and Chilka lake. CYGNSS can distinguish smooth water surfaces from rough terrestrial surfaces as the DDMs generated are dominated by coherent reflections. This makes it a valuable tool for inland water body detection. An algorithm is developed to classify DDMs into coherent, incoherent, and mixed categories using a deep convolutional neural network based on the InceptionResNetV2 architecture, achieving a classification accuracy of 97.46\%. The water masks generated by CYGNSS will be compared against Pekel Global Surface Water masks and Sentinel-1 data using a thresholding method to ascertain the performance.
The elevations of the water body are estimated from radar altimetry satellites Sentinel-3 and Sentinel-6, and also from Surface Water and Ocean Topography (SWOT) mission. These estimates are then compared with in situ Water Resources Information System India (WRIS-India) data provided by the Central Water Commission, Government of India. By combining water surface area from CYGNSS and elevation data from satellite altimetry missions surface water volume change is calculated. This approach provides a framework for assessing volumetric changes in inland water bodies by combining multiple datasets.
How to cite: Suresh, R. V., Garkoti, A., and Devaraju, B.: Estimation of surface water volume using CYGNSS and radar altimetry, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19256, https://doi.org/10.5194/egusphere-egu25-19256, 2025.
The water budget of the Great Salt Lake (GSL) relies on surface water and groundwater inflows from snowmelt in the Wasatch Mountain Block (WMB). Existing estimates of direct groundwater inflows to GSL are essentially derived from water budget residuals (i.e. inflow needed to balance the water budget) and are therefore subject to large uncertainty. Independent measures of groundwater inflows are needed to verify and improve water budgets and to evaluate the complex interplay between lake water and groundwater. Groundwater modeling, stream chemistry, streamflow modeling, and stream hydrograph analyses indicate that groundwater inflow (both directly into GSL and into streams within the GSL watershed) have been underestimated. Recent research has documented that most snowmelt infiltrates soils and recharges groundwater in the WMB before contributing to surface water supplies in the Salt Lake Valley. However, subsurface water storage and its role in water budget calculations remain difficult to quantify based on traditional hydrologic observations. Geophysical observations (GPS and satellite- and terrestrial-gravity) provide independent constraints on the flow and storage of water mass in the Mountain Block- Valley hydrological system. We demonstrate that geophysics data combined with land surface energy balance models, stream hydrograph data, and snowpack are suited to quantify the amount and time scales of water storage in seasonal snow, soil moisture, groundwater, and surface water storage in reservoirs and the GSL.
How to cite: van Dam, T.: Understanding Flow and Storage between the Wasatch Mountain Block and the Salt Lake Valley using GPS, Satellite Gravity, and Terrestrial Gravity, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-20595, https://doi.org/10.5194/egusphere-egu25-20595, 2025.
Tropical lowland lake-floodplain systems are increasingly threatened by climate change effects and other human-induced pressures. Determining the effect of these pressures on the water balance is challenging because of a lack of hydrological monitoring data, which impedes water management decisions. A collection of optical remote sensing and Synthetic Aperture Radar (SAR) scenes is used in combination with supervised classification algorithms and topographical data to derive lake volumes for the period 1984–2023, which are analyzed for trends and correlation with satellite-derived climate data. Although lake volumes show strong interannual variability, no significant historical trend is identified. A precipitation response time of approximately two months is observed, suggesting a considerable contribution of groundwater to the lake’s water balance. Minimum lake volumes found for the period 2014–2017 coincide with a prolonged period of below-average precipitation, indicating the effect of decreased groundwater recharge. Dry season lake volumes show weak correlation with cumulative precipitation in comparison to rainy season lake volumes, further indicating the importance of groundwater inflow for the dry season water balance. Results suggest that climate change effects and anthropogenic activities may have little short-term impact on the lake’s dry season volume, while altering groundwater recharge may have more significant long-term effects.
How to cite: de Klein, T., Bense, V., and Mustafa, S.: Estimation of water storage changes in a tropical lake-floodplain system through remote sensing, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-4299, https://doi.org/10.5194/egusphere-egu25-4299, 2025.
The increasing availability of Earth Observation (EO) satellites equipped with active microwave sensors suitable for flood mapping has improved flood monitoring capabilities. However, current observation frequencies still fall short of adequately characterizing inundation dynamics, particularly during critical moments such as the flood peak or maximum inundation extent. This limitation represents a significant research challenge in flood remote sensing. Advances in multimodal satellite hydrology datasets, coupled with the deep learning (DL) revolution, offer new opportunities to address the frequency gap in flood observations. TransFuse presents a scalable data fusion framework that combines DL with EO data to achieve daily, high-resolution flood inundation mapping. This proof-of-concept study highlights the potential of Vision Transformers (ViT) to predict flood inundation at the spatial resolution of Sentinel-1 (S1) imagery. The approach integrates time series data from coarse but temporally frequent datasets, such as soil moisture and precipitation from NASA’s SMAP and GPM missions, with static predictors like topography and land use. A ViT model was trained using flood maps derived from S1 imagery processed by a Random Forest Classifier, allowing the prediction of high-resolution flood inundation. Additionally, a classical UNET convolutional neural network (CNN) was used as a benchmark to compare model performance. Two case studies were used to evaluate this methodology: the December 2019 flood event in southwest France at the confluence of the Adour and Luy rivers, and the Christmas floods of 2023 on Germany’s Hase River. Predicted high-resolution flood maps were validated against independent flood masks derived from S1 images outside the training dataset. Results demonstrate that both ViT and CNN-UNET models effectively generalize the hydrological and hydraulic relationships that drive flood inundation, even in areas with complex topographies. Notably, the ViT model outperformed the CNN, achieving approximately 20% higher accuracy in both case studies. Further testing in diverse catchments with varying land-use, hydrology, and elevation profiles is recommended to assess model sensitivity under differing conditions. The proposed methodology can revolutionize flood monitoring by enabling daily observation of spatial inundation dynamics. This capability could support the development of improved parametric hazard re/insurance products, helping to address the flood protection gap faced by vulnerable populations worldwide.
How to cite: Dasgupta, A., Hosch, P. C., Sahu, R., and Waske, B.: TransFuse: Advancing Frequent Flood Monitoring using Vision Transformers and Earth Observation, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8312, https://doi.org/10.5194/egusphere-egu25-8312, 2025.
The Modrac Reservoir is a reservoir located mostly in the municipality of Lukavac, although it also touches the outskirts of the cities of Tuzla and Živinica. The Modrac Dam was built in 1964 on the Spreča River, which simultaneously formed the reservoir of the same name, with the main purpose of securing the necessary quantities of water for production processes of industrial capacities in the area of the municipalities of Lukavac and Tuzla.
Multi-purpose reservoir Modrac is a key water management facility of special importance for the life of the population, industry and tourism of the Tuzla Canton and as such for the entire Tuzla Canton but also BiH represents an inestimable value that requires special treatment, permanent investment and quality maintenance and management. By applying modern technologies used for operation and management of multi-purpose reservoirs and associated hydroelectric power plants, it is possible today to manage such water management systems in the most efficient way, to the benefit of all participating factors. Along with water supply and tourism, one of the key purposes and functions of the Modrac reservoir is flood protection in downstream areas.
This paper will present an operational prognostic local model of the Spreča River, which includes the Spreča River basin from its source to the Karanovac hydrological station. The developed local hydrological prognostic model of the river Spreča was created with a total of 6 sub-basins, the total size of the modeled basin is about 1,900 km2.
In the subject hydrodynamic model, a total of more than 180 km of watercourses were modeled as 13 river sections, 2 Q2D branches and 8 connecting channels within the Q2D sections. The geometry of the modeled sections is defined with approximately 230 cross-sections.
The subject paper will present the calibration of the hydrological and hydrodynamic model of the Spreča River for the associated catchment up to HS Karanovac, for water levels and flows, and was carried out for the period 2020-2023. Based on the prognostic model, a prognostic system for predicting floods in real time was created, which will also be presented in this paper.
The local operating system of the Spreča River, as well as other systems managed by the Agency for the Sava River Water Area, are compatible with the prognostic system developed in Croatia, which enables a simple exchange of input data.
Keywords: Modrac reservoir, Spreča river, hydrological-hydrodynamic model, prognostic model, operational system, flood forecasting
How to cite: Kovčić, O., Deduš, B., Kvesić, D., Ramuščak, R., and Jahić, E.: Operational forecasting model of the prevention and accumulation river basin Modrac, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-6403, https://doi.org/10.5194/egusphere-egu25-6403, 2025.
Climate change is significantly transforming familiar environments and affecting daily life. In this context, continuous monitoring of river discharge in space and time is crucial for planning human activities related to water use, preventing or mitigating losses due to extreme flood events, and reducing the effects of water scarcity.
Conventional in-situ monitoring stations have limitations such as low spatial density, incomplete time coverage and delays in data availability. These challenges hinder continuous spatio-temporal monitoring of river discharge. In response, researchers and space agencies have developed innovative satellite-based approaches to estimate runoff and river discharge using only satellite observations. In this perspective, the European Space Agency (ESA) has supported the STREAM (SaTellite-based Runoff Evaluation And Mapping) and STREAM-NEXT projects, which integrate satellite data on precipitation, soil moisture, terrestrial water storage anomalies, altimetric water levels, and snow cover into a simplified hydrological model, STREAM, to provide long-term independent global-scale gridded runoff and river discharge time series.
The STREAM model has been applied to over 40 river basins globally, including some of the largest such as the Mississippi-Missouri, Amazon, Danube, Murray-Darling, and Niger. It has demonstrated a strong capability to replicate observed river discharge even in heavily human-impacted basins where flow is regulated by dams and reservoirs. In addition, the model has shown its efficiency in simulating runoff and river discharge in Arctic basins (e.g. Lena, Mackenzie, Ob, Yenisey, and Yukon) where flows are controlled by glacier melt, and in small basins where the spatial resolution is still too coarse to describe the characteristics of the basins accurately.
The positive results obtained have paved the way for regionalizing the parameters of the STREAM model to make it applicable on a global scale. Through the calibration of the STREAM model across the 40 pilot catchments, it was possible to obtain a large set of parameters that were linked, through specific relationships, to various features including climate, soil characteristics, vegetation and topographic attributes. This approach yielded regionalized STREAM parameters. This study aims to evaluate the efficacy of the STREAM runoff and river discharge estimates, derived from regionalized parameters, across a diverse range of basins. To this end, a comparative analysis will be conducted between observed and simulated river discharge, as well as between simulated and modeled land surface runoff estimates.
This work aims to highlight how the use of readily available data, analyzed using a conceptual regionalized hydrological model, can improve the estimation of river discharge and the development of runoff maps, even in basins where complex interactions between natural processes and human activities prevail.
How to cite: Leopardi, F., Brocca, L., Saltalippi, C., Dari, J., Nielsen, K., Saemian, P., Sneeuw, N., Tourian, M., Restano, M., Benveniste, J., and Camici, S.: Toward a global scale runoff estimation through satellite observations: the STREAM model , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7377, https://doi.org/10.5194/egusphere-egu25-7377, 2025.
Posters on site: Fri, 2 May, 08:30–10:15 | Hall A
Rivers are a critical component of the global water cycle, serving as dynamic pathways for freshwater flow and storage. However, global discharge data is limited, particularly in regions with sparse in-situ measurements. This study introduces a hybrid modeling framework that leverages advanced satellite observations combined with machine learning and deep learning algorithms to estimate river discharge.The framework combines Sentinel-2 optical imagery, Sentinel-1 Synthetic Aperture Radar (SAR) data, and satellite altimetry data from Sentinel-3 and Sentinel-6 leveraging their complementary strengths. The input variables for the model include total water surface area and water indices derived from Sentinel-1 and Sentinel-2, while satellite altimetry provides water level time series. Sentinel-1 effectively compensates for the limitations of optical sensors under cloudy conditions. Moreover, satellite altimetry data are particularly evaluable in areas where lateral water expansion is constrained by topography and SAR or optical are unable to detect variations. The hybrid model, combining Long Short-Term Memory (LSTM) networks and Random Forest Regression (RFR), estimates river discharge with satellite-derived measurements. In effort to account for varying river morphologies, reach boundaries and river centerlines from the SWOT River Database (SWORD) are incorporated, ensuring robust adaptability to diverse conditions.The model is calibrated and validated against in-situ measurements on corresponding dates, using in-situ discharge data from the Mississippi River (USA), Kizilirmak River (Türkiye), and Po River (Italy). Designed to achieve high accuracy across diverse climatic and topographical settings, the proposed framework offers a scalable solution for estimating river discharge. By integrating satellite observations with a hybrid methodology, this approach has significant potential for enhancing global hydrological assessments.
How to cite: Tural, C. Y., Yilmaz, K. K., and Tarpanelli, A.: Satellite-based Framework for River Discharge Estimation: A Hybrid Approach Integrating Sentinel-1, Sentinel-2 and Altimetry Data , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-967, https://doi.org/10.5194/egusphere-egu25-967, 2025.
Satellite altimetry data has become essential for studying the dynamics of water bodies, especially in regions with limited or inaccessible data. Traditional low-resolution mode (LRM) satellites’ accuracy cannot be guaranteed when it comes to assessing water levels in small- (< 200 m in width) and medium-sized (200–800 m in width) rivers. Synthetic aperture radar (SAR) altimeters, exemplified by Sentinel-3 A, have shown great potential for inland water altimetry. Nevertheless, developing algorithms to retrack the raw data remains an essential requirement in this context. This is attributed to the width of small-sized rivers, which is often narrower than the along-track resolution of both LRM and SAR altimeters. In addition, new altimeters may have long revisit cycles and different spatial coverage and cannot yield historical data necessary in some situations.
To address these challenges, this study proposed a conditional threshold retracker (CTR). The CTR algorithm is well-designed and facilitates accurate water level monitoring. Moreover, we proposed an enhanced footprint filter (EFF), thus significantly bolstering the number of available cycles. Our findings demonstrate that the developed method substantially enhances the temporal and spatial resolution of both LRM and SAR altimetry satellites during water level monitoring in rivers of different climate types. The width of the thirteen selected rivers is on the order of 85–630 m. The CTR significantly improved the water level monitoring accuracy by 68 %– 78 %. Furthermore, the EFF increased the number of water level cycles by approximately 49 %–68 %. These findings have practical implications for obtaining accurate water level data, estimating river discharge and improving hydraulic model calibration.
How to cite: Hu, X.: Improving water level monitoring in small to medium-sized rivers: An enhanced footprint filter-based conditional threshold retracker approach, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-4658, https://doi.org/10.5194/egusphere-egu25-4658, 2025.
River discharge is a fundamental quantity that is required to improve our understanding of the hydrological cycle and to inform flood, drought, and water resources management (Gerten et al., 2008; Rajsekhar and Gorelick, 2017; Rao et al., 2020). Discharge monitoring plays a vital role in detecting climatic and environmental change because discharge is an integrated variable reflecting the coevolution of many processes within a basin (Hansford et al., 2020). However, ground-based measurements of discharge are often expensive and not available for many rivers globally. Therefore, spaceborne measurements are pursued as alternatives.
Recent studies have proposed various methods based on hydraulic equations to estimate discharge from multiple remotely sensed variables, such as water surface elevation (WSE), river width, and slope (Durand et al., 2016). However, such methods generally demand instantaneous observations of several variables. Some other methods rely on one single variable, such as width, WSE, or raw signal reflectance, provided that in-situ discharge data are available to build empirical relationships.
One widely used approach involves stage-discharge rating curves. Like ground-based methods, these curves estimate discharge by relating river stage (water level or WSE) measured by altimetry to discharge values previously recorded at gauging stations. This approach is straightforward to implement. This study leverages the power of Sentinel-3 altimetry to augment discharge estimates at the global scale.
We aim to achieve this through two key objectives:
- Developing a global network of rating curves: We will create a comprehensive dataset of stage-discharge rating curves using Sentinel-3 altimetry data.
- Investigating key influencing factors: We will investigate how river characteristics impact the reliability of these curves. Understanding these factors is crucial for optimizing their effectiveness.
How to cite: Jiang, L. and Zhao, Y.: Stage-discharge rating curves using satellite radar altimetry, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7758, https://doi.org/10.5194/egusphere-egu25-7758, 2025.
The interaction between rivers and coastal water bodies is critical to hydrological and ecological systems, particularly under the accelerating impacts of climate change. This study investigates the hydrodynamics of the Po river and its interactions with the Adriatic Sea during extreme events such as backwater effects during floods and saline intrusion during droughts. Using high-resolution data from Surface Water and Ocean Topography (SWOT) mission, integrated with in situ measurements, detailed LiDAR datasets, and a hybrid 1D-2D modeling approach in HEC-RAS, the research advances understanding of river-coast dynamics and their responses to climate-induced pressures.
The SWOT satellite, launched in December 2022, employs cutting-edge Ka-band Radar Interferometry (KaRIn) technology. The mission provides a variety of hydrological products for the surface water dynamics, with a revisit cycle of 21 days. For the inland rivers, the products include high-accuracy observations of water surface elevation, width, and slope, over a 120 km swath, allowing for improved rating curves and flow duration analysis. Stretching over 650 kilometers and flowing through eight Italian regions, the Po river is a lifeline for the northern region.
HEC-RAS is used to simulate riverine and floodplain dynamics, combining the computational efficiency of 1D modeling for long river reaches with the spatial detail of 2D modeling in areas with complex flow patterns, such as floodplains and river-coast interfaces. LiDAR-derived digital elevation models (DEMs) provide the foundation for defining cross-sectional profiles and updating hydraulic geometry, enabling precise representation of terrain and channel morphology.
The research follows a multi-phase methodology: SWOT data are processed to derive water surface elevations and extents, validated using in situ measurements and compared with HEC-RAS simulations. The study emphasizes extreme conditions, quantifying backwater effects during high flows and the severity of saline intrusion under low-flow scenarios. The integration of SWOT data with the HEC-RAS model allows for a detailed analysis of hydrodynamic processes, supporting the development of risk prediction models and improving water resource management strategies.
How to cite: Kurdnezhad, F., Domeneghetti, A., and Tarpanelli, A.: Leveraging SWOT data to analyze river hydrodynamic and coastal interactions during extreme events: A case study of the Po river, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10012, https://doi.org/10.5194/egusphere-egu25-10012, 2025.
The Amazon estuary conveys the largest amount of freshwater to the world ocean (20% of the global runoff). Over the past few years, its discharge exhibited record-breaking anomalies, be it flood events (June 2021 and June 2022) or dry spells (drought of November 2023 and October/November 2024). Assessing quantitatively the imprint of these extremes over the estuarine water level is challenging though, due to the ubiquitous and vigorous tidal signal propagating upstream from the Atlantic Ocean which prevents remote discharge estimates in the estuarine part of the river. We used the multi-mission nadir altimetry dataset composed of J3+S6A, S3A, S3B. Altogether, the satellite tracks encompass the whole estuary from its upstream limit 900 km inland down to the mouths of the Amazon terminal delta, making possible to map synoptically the spatio-temporal evolution of the estuarine water level and compute the separation of the tidal and hydrological contributions into the water surface height. The approach relies on an accurate de-aliasing of the tide in the altimetry records, based on a cross-scale hydrodynamic model of the Amazon estuary purposely developed and duly validated. This model uses the SCHISM ocean circulation code, with resolution of the order of 250 m inside the estuary. It allowed inferring a time-varying tidal atlas, which is utilized to remove the tidal signal from the altimetric anomalies. The altimetric residuals depicts the spatio-temporal pattern of water level anomalies in response to discharge variations, both during the flood and drought periods. For instance, the 2021 and 2022 extreme floods induced an anomaly that lasted about 1 month each time, with water level peaks about 50 cm above the seasonal climatology, extending over the upper 500 km of the 900 km-long estuary. Downstream-ward of this, the imprint of the extreme floods decayed sharply, and reached insignificant magnitude throughout the downstream-most 300 km of the estuary (corresponding roughly to the terminal delta). A mirror conclusion can be drawn for the 2023 drought, with 1 m negative anomaly below the seasonal, mostly restricted to the upper 300 km of the estuary at the peak of the event in November 2023, and with a weak signal further downstream. The magnitude of these anomalies largely exceeds the bounds of the accuracy of our altimetric dataset. We present that it is now possible to derive reliable discharge estimates in the estuarine reach of the Amazon river by converting these tidal-free water levels from altimetry measurements through a classical rating curve, including for the extreme events.
How to cite: calmant, S., Arjailles, V., durand, F., coulet, P., santos, L., testut, L., moreira, D., paris, A., and paiva, R.: Separation of the hydrological and tidal components in water heights to estimate discharge in the downstream, tidal, Amazon, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-18170, https://doi.org/10.5194/egusphere-egu25-18170, 2025.
In recent years, advancements in radar altimetry, particularly the Synthetic Aperture Radar (SAR) approach, have revealed fine-scale features over coastal ocean and inland waters. SWOT mission has offered unprecedented details and accuracy in observing the nuance of water surface gradient since its launch at the end of 2022. It provides a great opportunity to monitor the ungauged rivers and waterbodies timely and repeatedly. This study aims to utilize SWOT L2 Lake and Pixel Cloud products to monitor multiple lakes, ponds, and reservoirs in Taiwan. In our fieldwork consisting of 14 small ponds and 12 major reservoirs, it has been verified that the surface height and its temporal changes could be observed by SWOT at an accuracy of submeter level during cycles 3-26. After our reprocessing by clustering of pixel clouds within the predefined water masks, the accuracy can be further improved to <10 cm level. It is concluded that SWOT offers an alternative view of hydrological parameters, which can play a critical role in future water resources management.
How to cite: Tseng, K.-H.: Monitoring Surface Water Bodies in Taiwan by SWOT Lake And Pixel Cloud Products, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8839, https://doi.org/10.5194/egusphere-egu25-8839, 2025.
Land surface water bodies are important to ensure water security for agricultural, industrial, domestic, and environmental sectors. Especially in dryland areas, such as the Loess Plateau in China, changes in land surface water bodies as a response to climate change and human activities have been the subject of great concern. Many dams and reservoirs have been constructed on the Loess Plateau to combat serious soil erosion and water resource shortages. These projects are widely recognized as an effective measure to enhance soil conservation, but little is known about the dynamics of surface water bodies. In this study, we employ a long-term satellite water product to detect the spatial-temporal variability in surface water at the regional scale on the Loess Plateau and identify the potential cause of climate change and human activities. The results show that the area of permanent water has increased by approximately 800 km2 during the past two decades. Surface water expansion is primarily associated with small water bodies (< 1 km2), as their number has roughly doubled, while the number and area of large water bodies have remained stable. We found that surface water expansion has little correlation with precipitation variation but is highly correlated with water withdrawal for agricultural, industrial, and other sectors. Thus, the surface water expansion on the Loess Plateau is primarily contributed by hydraulic project construction as a response to the increasing water demand. The above findings imply the positive role of hydraulic projects, but it is essential to note that the continuous expansion of surface water might not be sustainable because of constraints from natural conditions.
How to cite: Liu, Y. and Xie, X.: Surface water expansion due to increasing water demand on theLoess Plateau, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9406, https://doi.org/10.5194/egusphere-egu25-9406, 2025.
Sentinel-1 C-band Synthetic Aperture Radar (SAR) provides a new means for indirectly monitoring water reservoir level and storage by mapping water cover at resolutions suitable for large water bodies. Monitoring these fluctuations is essential for informed water resource management even in otherwise gauges reservoirs for the purpose of verification, detection of anomalies due to changes in floor morphology, etc. However, classification of water on SAR images can be ambiguous, to which end several non-parametric methods such as Otsu, Kittler–Illingworth (Kavats et al., 2022), k-means (Cheng et al., 2022), and entropy-based image thresholding (Sekertekin, 2021) have been proposed. Here we evaluate the capability of these methods to accurately reconstruct reservoir biweekly water level and storage. The analysis is performed on Sentinel-1 Ground Range Detected (GRD) imagery, acquired in the VV polarization mode from the COPERNICUS/S1_GRD image collection from October 2014 till today using Google Earth Engine (GEE). Processing is performed using the GEE JavaScript API and executed through the R programming environment using the rgee package. Water level and storage are derived from water cover using respective level-area and level-storage curves. The methods are applied to two reservoirs located in Greece and the Czech Republic, which are characterised by distinct seasonal water availability and demand leading to the respective reservoir level fluctuations. Results are validated by comparing against official measurements, indicating satisfactory fit. These findings highlight the potential of the proposed methods automated continuous reservoir monitoring, especially in regions facing increasing climatic variability as climate change is expected to increase the intensity of droughts and seasonal fluctuations in water availability. The study contributes to improving methodologies for assessing water dynamics in diverse climatic environments and supports the development of more efficient strategies for water resource management.
Acknowledgements
This research was conducted during ERASMUS+ KA131 mobility (contract number 1023). This work has received funding from REACT4MED: Inclusive Outscaling of Agro-Ecosystem Restoration Actions for the Mediterranean. The REACT4MED Project (grant agreement 2122) is funded by PRIMA, a program supported by Horizon 2020.
References
Cheng, L., Li, Y., Zhang, X., & Xie, M. (2022). An Analysis of the Optimal Features for Sentinel-1 Oil Spill Datasets Based on an Improved J–M/K-Means Algorithm. Remote Sensing, 14(17), 4290. https://doi.org/10.3390/rs14174290
Kavats, O., Khramov, D., & Sergieieva, K. (2022). Surface Water Mapping from SAR Images Using Optimal Threshold Selection Method and Reference Water Mask. Water, 14(24), 4030. https://doi.org/10.3390/w14244030
Sekertekin, A. (2021). A Survey on Global Thresholding Methods for Mapping Open Water Body Using Sentinel-2 Satellite Imagery and Normalized Difference Water Index. Archives of Computational Methods in Engineering, 28(3), 1335–1347. https://doi.org/10.1007/s11831-020-09416-2
How to cite: Daliakopoulos, I., Kadlec, J., and Skaloš, J.: Reconstruction of reservoir water level and storage using Sentinel-1 C-SAR, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19181, https://doi.org/10.5194/egusphere-egu25-19181, 2025.
Floods and droughts are among the most destructive hydrological extremes, creating severe socio-economic disruptions worldwide. Many regions, especially those in the Global South, remain highly vulnerable due to inadequate forecasting precision caused by sparse observational networks and limited model capabilities. The European Commission-funded SEED-FD (Strengthening Extreme Events Detection for Floods and Droughts) project under Horizon Europe aims to address these gaps by leveraging advanced Earth observation (EO) and non-EO datasets to strengthen forecasting systems for floods and droughts.
The primary objective of SEED-FD is to enhance the accuracy and global usability of the Copernicus Emergency Management Service (CEMS) Early Warning Systems (EWS). This involves refining key elements of the CEMS hydrological forecasting framework, including the LISFLOOD model’s hydrological processes and calibration strategies, integrating innovative machine learning and data assimilation techniques to improve predictions, and creating new global forecast products. A key focus is on incorporating nontraditional observational data, such as precipitation, soil moisture, and streamflow measurements from EO sources, as well as river discharge data obtained from microstations.
The project adopts a two-step strategy: initial algorithm and method validation in data-rich regions (Danube and Bhima basins) to establish proof of concept, followed by scaling and application in three diverse and vulnerable regions—the Paraná River Basin (Brazil), the Niger River Basin (West Africa), and the Juba-Shebelle Basin (Horn of Africa).
This presentation will cover mid-term findings from SEED-FD, emphasizing progress in hydrological model calibration, improved process representation, data assimilation, and machine learning-based post-processing. These advancements have demonstrated enhanced prediction reliability in the Danube and Bhima basins and offer valuable lessons for scaling solutions to other vulnerable regions.
How to cite: Pedinotti, V. and Matthews, G. and the Vanessa Pedinotti and Gwyneth Matthews: Enhancing Global Flood and Drought Forecasting with SEED-FD: Integrating Remote Sensing for Hydrological Insights, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11723, https://doi.org/10.5194/egusphere-egu25-11723, 2025.
Additional speakers
- Jérôme Benveniste, France
- Angelica Tarpanelli, National Research Council, Italy
- Karina Nielsen, Technical University of Denmark, Denmark
- Fernando Jaramillo, Sweden
- Ioannis Louloudakis, Hellenic Mediterranean University, Greece