The interception of rainfall by a surface, such as vegetation and soil, reduces the quantity of water participating in downstream processes. This rainfall interception loss is a non-negligible quantity that varies with the ecosystem, meteorological, and rainfall conditions. Rainfall interception models are needed to incorporate the three properties to estimate interception loss. However, the interception losses from these models are rarely validated directly. Rather their validation relies on the residual of the water balance. Eddy Covariance (EC) towers measure the fluxes of water vapour and present an opportunity to validate the modelled interception losses.
We present a pioneering interception study that compares evaporation from physically calibrated interception models to the energy balance closure corrected water vapour fluxes recorded intermittently by an EC tower. We generate parameters for the canopy interception models by the hierarchical Bayesian treatment of the Rutter, Rutter sparse, Gash, and Calder models with the data from automatic throughfall and rain gauges over 271 rain events (177 for Gash). We use these models to estimate the extent of soil cover, throughfall to the soil underneath the canopy, and interception losses. A physically calibrated soil evaporative capacitor model was then used to model the evaporation from the bare soil and soil beneath the canopy. The canopy interception models recreated the event-wise throughfall well, however they did not represent a substantial improvement on the benchmark of a simple percentage estimate. The combined soil and canopy model is considerably better than a simple percentage in recreating the magnitude and time series of interception loss. The method developed can be applied to any EC tower that measure throughfall allowing for broader insights to be generated into the capability and limitations of interception loss modelling.
How to cite: Kunadi, A., Silberstein, R., Leopold, M., and Thompson, S.: S&p 500(0): Utilizing Hierarchical Bayesian Modelling to advance parameter estimation of Canopy Interception Models using Eddy Covariance, Throughfall and Soil Hydraulic Measurements in a Mediterranean Ecosystem, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-17227, https://doi.org/10.5194/egusphere-egu25-17227, 2025.
Agrometeorological modelling provides a powerful tool to gain insights into the dynamics of agricultural ecosystems, especially where measurements are unavailable. For practical applications like forecasting, both temporal patterns and absolute levels of target variables have to be confidently predicted by the model. In our EU-funded project, we are working on the Europe-wide application of the model AMBAV. This physics-based model, which computes coherent water and energy balances in agricultural soil-vegetation-atmosphere systems, is developed at Germany's national meteorological service DWD (Deutscher Wetterdienst), and is in operational use for the area of Germany. The aim of this study is to evaluate the performance of AMBAV at sites in neighbouring countries using local measurements as reference.
We used soil descriptions and meteorological time series from selected DWD and ICOS ecosystem stations to run AMBAV simulations for grassland sites, resulting in multi-year time series at hourly resolution. To assess the model performance, the model predictions of soil moisture in various soil depths and latent heat flux densities were compared to locally measured time series. A set of statistical metrics including Pearson's correlation, the mean error and the ratio of standard deviations as well as the Kling-Gupta-efficiency was used to report on model performance.
The results show that the AMBAV model can be used to reliably predict soil moisture and latent heat flux densities at Central-European grassland sites. For soil moisture, correlations above 0.75 and mean errors within ± 0.08 m3 m-3 in soil depths down to 100 cm are achieved. Similarly high correlations are found for latent heat flux densities, while the magnitude of the mean error strongly depends on the corrections for energy balance closure typically applied to the measurement data. We also address seasonal variations of model performance in our evaluation. This work highlights strengths and weaknesses of the model AMBAV as well as the value of high-quality input and reference data. Our results encourage further investigation on a broad Europe-wide application of the AMBAV model.
How to cite: Brandtner, V., Herbst, M., Lucas-Moffat, A., Flechard, C., and Calvet, J.-C.: Modelling soil moisture and latent heat fluxes at grassland sites in Central Europe, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9310, https://doi.org/10.5194/egusphere-egu25-9310, 2025.
Stochastic rainfall models rely on accurate rainfall distributions. Since rainfall is generated by various processes, rainfall series are composed of events with different distributions. In such cases, the use of mixed distribution approaches has been recommended (e.g., Laaha 2023) based on event separation. However, separating rainfall events into generative types is not straightforward.
We propose clustering based on event characteristics derived from the rainfall series (such as peak magnitude, duration, average intensity, and relative time to peak) to stratify the rainfall series into types of rainfall events. These event characteristics are derived using Yevjevich’s theory of runs, which is commonly used in hydrological drought studies and is adapted here for rainfall event separation to exploit the temporal characteristics of rainfall events. Additionally, a binary lightning index is used to help distinguish between convective and stratiform events.
We compare two methods for event classification. The first method is model-based clustering using a Gamma mixture model. The second method is the robust partitioning method PAM, which uses Gower’s distance to handle the mixed data structure of the event characteristics. Both methods are optimized regarding the number of clusters using state-of-the-art criteria.
The analysis shows that clustering based on rainfall event characteristics and the lightning index is a simple yet effective method to reduce process heterogeneity in rainfall frequency analysis. These characteristics are obtained without additional weather data, which is a major strength of the approach. Finally, we compare the distributions of event types to discuss the value of mixed distribution approaches for stochastic rainfall modeling. In summary, this study encourages a better understanding of statistical assumptions in applied models and enriches the physical knowledge included in environmental statistics, such as stochastic rainfall models.
Reference:
Laaha G (2023). “A Mixed Distribution Approach for Low-Flow Frequency Analysis – Part 1: Concept, Performance, and Effect of Seasonality.” Hydrology and Earth System Sciences, 27(3), 689–701. ISSN 1607-7938. doi:10.5194/hess-27-689-2023.
How to cite: Laaha, G., Özcelik, N. B., Ortega Menjivar, L., Fischer, S., and Laimighofer, J.: Improving stochastic rainfall models through event classification, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19553, https://doi.org/10.5194/egusphere-egu25-19553, 2025.
High Mountainous Areas (HMAs) with extensive snow cover presents considerable modelling challenges due to their intricate topography and data scarcity. However, these regions are particularly of concern as they are experiencing rapid warming, resulting in accelerated snowmelt along with more intense rainfall events, complicating the notion of flood typology that has existed for the region since long. Here, we present a comprehensive framework to evaluate projected changes in flood typology in HMAs for future relative to the Historical period. The results suggest that there will be a notable increase in rainfall-induced floods, particularly of the short-duration variety, coupled with a decrease in snowmelt-induced floods as the future periods advance. Additionally, there is a noticeable shift in the mean timing of floods, suggesting a delay in their occurrence. Although these results are specific to the three regions studied, similar changes will likely occur in other snow-dominated basins across HMAs. These insights could empower policymakers to make informed decisions and enhance regional risk assessment and management strategies.
How to cite: Shah, S., Liu, Y., Kim, S., Sharma, A., and Fischer, S.: Shifting Mountain Flood Regimes under Global Warming, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-526, https://doi.org/10.5194/egusphere-egu25-526, 2025.
The identification of model mechanisms for representing hydrological (physical) process is a major scientific and practical problem in catchment scale hydrological modelling. We present a multiple hypothesis-testing approach to identify dominant hydrological mechanisms. The method combines: (i) Bayesian estimation of posterior probabilities of individual hydrological mechanisms given an ensemble of hydrological model structures; (ii) a test statistic that defines a “dominant” mechanism as the mechanism with (substantially) higher posterior probability than the sum of the alternative ones given observed data; (iii) a flexible modelling framework to generate hydrological models from combinations of available mechanisms. The uncertainty in the test statistic is approximated using a model bootstrap approach. The performance of the proposed framework is evaluated using synthetic and real data. We use 624 model structures from the Framework for Understanding Structural Errors (FUSE) and data from the Leizarán catchment in Basque Country (northern Spain). The synthetic experiments indicate that the mechanism identification method is reliable; as expected its statistical power (identifiability) declines as data/model errors increase. The "most identifiable" processes are those in the saturated zone and routing, and the "least identifiable" processes are interflow and percolation. The real data experiment yields results that are broadly consistent with the synthetic experiments, with dominant mechanisms identified for 4 of 7 processes. We expect that the proposed mechanism identification method will contribute to hydrological community efforts on improving process representation and model development.
How to cite: Prieto, C., Kavetski, D., Le Vine, N., Fenicia, F., Scheidegger, A., and Vitolo, C.: Identification of dominant mechanisms for representing hydrological processes in catchment scale models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2942, https://doi.org/10.5194/egusphere-egu25-2942, 2025.
Accurately predicting river water levels is essential for effective water resource management and reducing flood risks. Traditional hydrological models often struggle to capture the complex, nonlinear dynamics of river systems. In this study, we explore machine learning techniques to enhance water level predictions. Specifically, we focus on hybrid and ensemble models that combine the strengths of various algorithms to improve both accuracy and reliability. Our approach integrates methods such as Sequential Minimal Optimization for Regression (SMOreg), Rep-Tree, and Decision Table (DT) to predict water levels in the Rhine River. By leveraging hybrid models, we aim to uncover patterns in hydrological data that traditional methods may miss, leading to more precise predictions. The models were trained and validated using 10 years of historical data from the Worms station, incorporating meteorological and hydrological variables as inputs. This study demonstrates that hybrid and ensemble machine learning models offer a robust and reliable solution for predicting river water levels. It underscores the potential of advanced data-driven approaches to support sustainable water resource management and mitigate the impacts of flooding.
How to cite: Sheikh Khozani, Z. and Ionita, M.: Advancing River Water Level Prediction Using Hybrid and Ensemble Machine Learning Models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-5576, https://doi.org/10.5194/egusphere-egu25-5576, 2025.
In the context of global climate change and rapid urbanization, the risk of urban flood disasters caused by heavy rainfall is continuously increasing. To effectively address this challenge, this study developed a stormwater flooding simulation model at the urban community scale by coupling a one-dimensional pipe network hydrodynamic model (SWMM) with a two-dimensional surface water hydrodynamic model (LISFLOOD-FP), with particular emphasis on the impact of temporal dynamics on simulation outcomes.
The coupling process integrates time step synchronization, data transmission, and updating model configuration to ensure accurate and dynamic simulations. Initially, an appropriate time step for the SWMM model was selected to ensure its output data provided a suitable temporal resolution for LISFLOOD-FP, optimizing data exchange frequency and detail. At the end of each time step, the overflow node coordinates and volumes from SWMM were converted into the required .bci and .bdy file formats and promptly transmitted to LISFLOOD-FP as input conditions for the next time step, ensuring real-time interaction between the models. Meanwhile, the LISFLOOD-FP configuration files (.par) were updated in real-time based on the latest SWMM output data, incorporating the most recent overflow information as boundary conditions. This continuous feedback loop allowed LISFLOOD-FP to dynamically adjust its simulations, enhancing the precision of inundation predictions.
Validation using actual precipitation data from July 11, 2023, and design storms with various return periods (1a, 5a, 10a, 20a, 50a, and 100a) demonstrated high accuracy in simulating stormwater network loads, inundation extents, and depths. High-risk areas were primarily located at the boundaries of academic zones, the southern side of residential areas, and their intersections. The study concludes that the real-time coupled simulation method, grounded in temporal sequence, not only enhances the precision of inundation predictions but also fully accounts for the complexity of urban stormwater systems, providing robust support for urban planning and disaster mitigation strategies.
How to cite: Chen, C., Li, Y., Wang, L., and Hu, T.: Temporal dynamics of coupling 1D and 2D hydrodynamic models for urban community rainstorm flooding simulation, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9561, https://doi.org/10.5194/egusphere-egu25-9561, 2025.
Deficit in water budget has a negative effect on society, environment and economy. In order to mitigate these negative effects, it is important to known how precipitation falling on the basin is shared by streamflow, evapotranspiration, baseflow and infiltration; i.e., What is the share of precipitation transferred to flow, returned to the atmosphere through evapotranspiration or contributed to the basin storage system through infiltration? Sharing water resources among drinking and utility water, agriculture, animal husbandry, industry, tourism, etc., is necessary and important for the sustainability. In a simple water budget, the basic and traditional approach can be accompanied by the methods such as Budyko curve, a method that can be used in water budget calculation in hydrology. The Budyko curve consists of a nonlinear relationship between the evaporation rate and the dryness index and it defines the water- and energy-based limits of evapotranspiration. In this study, a water budget approach based on the Budyko curve is proposed by considering the monthly rainfall-runoff relation. Firstly, the aim of the study is to determine whether Kucuk Menderes River Basin in western Turkey complies with the Budyko curve; and if yes, secondly, whether it changes over time, and if it does, how it changes. The monthly total precipitation, and monthly average temperature of meteorological stations and monthly average streamflow were used. Assuming that there will be no change in the basin water storage in the long-term, actual evapotranspiration will be taken as the difference between the streamflow discharge and precipitation, and the potential evapotranspiration will be calculated by the empirical Thornthwaite method. From the application on the upstream and downstream reaches, the river basin was found consistent with the Budyko framework in general. However, according to the calculations made by using the Budyko curve, higher flows than expected were obtained for the upstream reach and lower flows than expected for the downstream reach of the basin. Based on the results of the case study, the potential of Budyko curve as a method to use in planning and management of river basin was demonstrated. The study will also investigate the possibility of extending the use of the Budyko curve to ungauged or nested basins.
How to cite: Cavus, Y., Eris, E., Akyuz, D. E., and Aksoy, H.: Budyko Curve as a Tool for Long-term Water Budget Calculation, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-795, https://doi.org/10.5194/egusphere-egu25-795, 2025.
Floods rank among the most financially devastating natural hazards, imposing significant risks on communities, insurers, and policymakers. In recent years, intensive advancements in artificial intelligence and machine learning research have enhanced the ability to mitigate these risks but also for predicting vulnerable areas, claim amounts, and patterns of flood impact. In this context, our study explores the potential of machine learning models to predict flood insurance claims based on historical streamflow data, actual flood claim records, and regional characteristics. To this respect, we integrate the US-CAMELS dataset, which provides detailed streamflow timeseries, with Federal Emergency Management Agency’s National Flood Insurance Program (NFIP) Redacted Claims dataset, containing millions of flood-related insurance claims across the contiguous USA. This integration yields a composite dataset featuring streamflow metrics—such as intensity, duration, and recurrence—alongside FEMA variables, including claim history, flood frequency, and policy characteristics.
Our approach employs machine learning models to predict outcomes such as expected aggregated insurance claims and the likelihood of claim occurrences across different regions, while simultaneously evaluating model’s performance. Through this methodology, we identify critical predictors of flood-related insurance claims, providing valuable insights for risk assessment, enhancing the non-structural elements of early warning systems and economic resilience in flood-prone areas, thus, contributing to the development of proactive and data-driven insurance strategies.
How to cite: Manolis, G. T., Papoulakos, K., Tepetidis, N., Iliopoulou, T., Dimitriadis, P., Tsaknias, D., and Koutsoyiannis, D.: Machine learning models for predicting flood insurance claims through recorded streamflow and historical-claims data integration, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-4837, https://doi.org/10.5194/egusphere-egu25-4837, 2025.
Satellite precipitation products, such as early runs of the Integrated Multi-satellitE Retrievals for GPM (IMERG-E), provide comprehensive spatial and temporal coverage, offering significant improvements in rainfall monitoring in remote regions. These products are essential for enhancing the accuracy of flood forecasting. However, compared to the ground-based observations, IMERG data not free from biases. In this study, we apply Cumulative Distribution Function (CDF) matching combined with Support Vector Machines (SVM) to correct biases over the Mahanadi River Basin in eastern India. The Kernel-based SVM is used to capture the nonlinear relationships between the 0.1º× 0.5h IMERG-E and 0.25º×1-day IMD gridded observations. Subsequently, this method is compared with the traditional CDF bias-correction techniques, aiming to improve the IMERG-E precipitation estimates. The corrected IMERG estimates can contribute to more reliable flood forecasting, supporting informed decision-making processes in flood risk management.
How to cite: Mohanty, R. S., Chatterjee, C., and Sahoo, B.: Enhancing accuracy of IMERG-E satellite rainfall products for Mahanadi River Basin using bias-correction methods, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15074, https://doi.org/10.5194/egusphere-egu25-15074, 2025.
A coupled model integrating dual attention mechanism into BiGRU-RED for multi-step-ahead streamflow forecasting
The accurate streamflow forecast is of utmost importance in the efficient administration of water resources. In this research, we introduced the DA-BiGRU-RED model, an approach that incorporated the dual attention (DA) mechanism involving both feature and temporal attention into the Bidirectional Gated Recurrent Unit (BiGRU) with a recursive encoder-decoder (RED) structure. The feature attention was derived by allocating weights to the hidden states of the BiGRU in the encoder, enhancing the model’s capability to efficiently capture crucial features of the input variables. Concurrently, the temporal attention mechanism was established by jointly weighting the hidden states of the BiGRU in the encoder and decoder, enabling the extraction of temporal message from the input variables. This dual-attention mechanism empowered our model to effectively extract essential information from various kinds and temporal instances of input data, thereby improving the accuracy of multi-step streamflow forecasting. Furthermore, to assess forecasting uncertainty, we employed MC dropout based on Bayesian statistical theory. To gauge the effectiveness of our proposed model, we applied it for 1-day, 3-day, 5-day, and 7-day ahead forecasting in the Heihe River basin in Northwest China. Our model consistently outperformed both the BiGRU-ED and BiGRU models, as evidenced by Nash-Sutcliffe coefficient (NSE) values exceeding 0.69 in nearly all prediction scenarios. Additionally, the uncertainty assessment revealed that the DA-BiGRU-RED model exhibited the highest PUCI values, underscoring its efficacy in extracting key features and temporal information from input variables and providing more accurate and robust forecasting results.
How to cite: Zhou, T. and Huang, C.: A coupled model integrating dual attention mechanism into BiGRU-RED for multi-step-ahead streamflow forecasting, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7862, https://doi.org/10.5194/egusphere-egu25-7862, 2025.
Glacial Lake Outburst Floods (GLOFs) often occur at high elevations, resulting in catastrophic flash floods while leaving little or no time for effectively managing them. They are characterized by destructive velocities of several m/s, enabling the transport of a large amount of debris and sediments. Due to a combination of the solid and liquid forces, these flows are characterized by greater momentum and continue recruiting debris and sediments along the path. As a result, they achieve farther run out distances compared to clear water flows. Hence, they deviate largely from pure water mechanics or Newtonian hydraulics. Modelling them using non-Newtonian physics and sediment transport may pave the way for an accurate simulation of these floods. Therefore, in this study, we use various non-Newtonian schemes in a freely available hydraulic model to simulate the recent 2023 Sikkim GLOF flood. We perform Monte Carlo uncertainty-based calibration to estimate the behavioural parametric values for different non-Newtonian approaches in the model. As an outcome of this study, we identify the best-performing non-Newtonian approach and its calibrated parametric ranges. Thus, the study serves as a guideline towards accurate modelling of GLOFs. The framework adopted in this study can be used elsewhere to simulate GLOFs realistically. As they are expected to become more frequent due to climate change, a pre-assessment modelling strategy for gauging the potential damage can be established for the existing glacial lakes.
How to cite: Nallasamy, N. D., Ganapathy, A., Cook, K., Rajak, R., and Hovius, N.: An effective non-Newtonian approach in simulating GLOF Floods, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11834, https://doi.org/10.5194/egusphere-egu25-11834, 2025.
Water resource scarcity-to-sustainability is a glaring issue with climate change, increases in the population, and industrial development. Hydropower reservoirs provide a sustainable energy production solution against all these odds. Further, these are highly integrated into modern energy grids and are becoming more susceptible to variability in climatic and consumption changes.
so this topic become a global issue today due to the pressure of increasing climate change effects, rapid population growth, and industrialization. Demand for hydropower reservoirs coupled with the energy production and ecosystem services that they provide, has put a lot of stress on these sources. The present approach employs a novel stochastic model using system dynamics to optimize water allocation from hydropower reservoirs while ensuring ecosystem sustainability as well as environmental conservation.
The focus of the study is on Karun Basin. Karun Basin is one of Iran's most important water basins and contains 27 reservoirs of varying sizes and operational priorities. Dynamic interactions among the system components were modeled using the Vensim software platform, taking into consideration 49 scenarios of domestic, industrial, agricultural, and environmental demands as well as expected changes in future climatic and socio-economic conditions this modeling effort undertakes using.
Some important performance indicators, such as temporal and volumetric reliability, were calculated to assess the sustainability of the system for all possible management strategies. It is evident from the results that the traditional policies of water allocation are not sufficient to address modern challenges like increasing demand and climatic variability in available supplies. The proposed model optimizes key allocation strategies that maximize energy production and minimize environmental impacts and is applied to recognize endowments by decision-makers. Most importantly, three critical vulnerabilities were identified in the worst-case scenario for the major reservoirs- Beheshtabad, Vanak, and Shahid. This study highlights the importance of adaptive management in risk reduction and resilience of the system.
The present work demonstrates how system dynamics can be used as a decision-support tool for dual purposes: achieving a balance between energy production and environmental sustainability-potentially because the system may be actionable. It has paid attention to one of the matters increasingly being discussed as an important area of action to advance the integration of ecosystem services in water resource management policies. Future work should expand the model using an online environmental monitoring extension and various other applications in similarly water-stressed areas. The stochastic model and sustainability metrics have developed the approach that policymakers and practitioners can take to meet the increasing challenges of water and energy management. Future research will help improve the model through real-time data collection and investigate its applicability in other water-stressed regions.
Keywords: water resource management, Karun Basin, hydropower optimization, system dynamics, stochastic modeling, climate adaptation, sustainability indicators, Vensim
How to cite: Pirani, N., Jabbari kalkhoran, B., Masoumi ganjgah, F., Karimzadeh, R., Ghadimi hellabad, R., and Sabouri, M.: Adaptive Water Resource Management for Hydropower and Ecosystem Resilience: A Case Study Using System Dynamics, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11507, https://doi.org/10.5194/egusphere-egu25-11507, 2025.
Hydrodynamical models on watersheds with high flash-flood risk are useful tools. Although it is important to know how reliable a model is for calculating aspects of the overland flow.
Steep-sloped, natural watersheds have sudden geometrical changes. The slope alteration on the overland flow area, the chaotic meandering of the natural stream beds cause complex turbulent flow.
In average, hydrodynamical models with solvers based on the full depth integrated Shallow Water Equations (SWE) can be accepted as the most accurate methods. On the other hand, the SWE based models can easily become unstable due to the changes in geometry. If a simulation becomes unstable, to increase stability simplifications can be necessary. The Local Inertia Approximation (LIA) is a simplified method, where nonlinear advection is neglected. Although with the simplification the proper representation of the turbulent flow can be lost. Therefore, additional methods are required to regain the proper turbulent effect without the sacrifice of the model’s stability.
The Large Eddy Simulation (LES) can be a suitable additional part for the overland flow model. The method’s task is to increase the model’s accuracy on the calculation of the turbulent flow.
This research uses artificial watersheds to compare different modelling methods thru stable calculations. The goal is showing how simplified hydrodynamical models’ accuracy can be amplified on calculating turbulent overland flow.
Keywords: numerical modelling, hydrodynamics, shallow water equation, local inertia, overland flow, watershed model
How to cite: Ámon, G. and Bene, K.: Accuracy of the Local Inertia Approximation solver and the importance of Large Eddy Simulation in 2D hydrodynamical models on steep-sloped watersheds, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13519, https://doi.org/10.5194/egusphere-egu25-13519, 2025.
Soil moisture plays a critical role in various hydrological processes, controlling groundwater recharge, infiltration, and the generation of overland flow. Additionally, it stands as a key determinant for the water supply essential for sustaining vegetation and agricultural crops. Soil moisture measurements using Time-Domain Reflectometry (TDR) sensors are labor-intensive and expensive, often requiring extensive setup and maintenance. Furthermore, because of the small support volume of these sensors, two of them placed at the same depth just a few meters apart can produce readings that vary by 20 volume percentage or more. These uncertainties are not random; rather, they reflect the small scale variability of e.g. soil texture, which largely controls soil water retention and thus soil moisture dynamics. To effectively model soil moisture and address these uncertainties, we need deep learning (DL) models that can not only make accurate predictions but also learn to represent the underlying variability. Here, we present a model that combines Long Short-Term Memory Networks (LSTMs) with Gaussian Mixture Models (GMMs), trained on a large dataset of uniquely placed in-situ soil moisture observations collected from the Attert experimental basins in Luxembourg. By training on in-situ soil moisture observations, our model aims to generalize soil moisture dynamics across spatial dimensions, temporal scales, and depths. Unlike traditional models that predict a single, deterministic value, the proposed network outputs weighted probabilistic distributions, providing a promising way to capture small scale soil moisture variability. With this approach we aim to evaluate the predictive performance of different Gaussian-Mixture LSTM (GM-LSTM) setups for soil moisture dynamics and to quantify observational variabilities and uncertainties. By temporal predictions and the assessment of variability we have shown that the developed model setup is capable to model the dynamical fluctuations of soil moisture, as well as to replicate the variability within the cluster site locations. In addition, we observed seasonal variations in the probabilistic model outputs, with lower uncertainty during dry periods and higher variability during wet phases, highlighting the ability of data-driven approaches to uncover relationships and offer additional insights into the dynamics of soil moisture systems. In summary, by using GM-LSTMs, we demonstrated that this modeling approach is capable of simultaneously predicting soil moisture dynamics while accounting for local-scale variability, which is important for improving drought monitoring and agricultural productivity.
How to cite: Bischof, B., Loritz, R., and Zehe, E.: Modeling Soil Moisture Dynamics and Variability Using Gaussian Mixture-Based Long Short-Term Memory Networks, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12180, https://doi.org/10.5194/egusphere-egu25-12180, 2025.
In complex geographic environments, spatial relationships are often distorted by natural barriers and irregular terrain as well as irregular sampling. These sub-optimal conditions often present challenges for geostatistical modeling, especially in multivariate data. Traditional covariance and variogram models relying on Euclidean distance may fail to capture such complexities, necessitating the use of alternative approaches. Building on previous research, this study investigates the performance of various covariance and variogram models applied to multivariate geostatistical data from a mine in Ireland. The dataset, consisting of published concentrations of co-located metals, provides a good opportunity to explore the utility of advanced spatial modeling techniques in continuation of our previous (univariate) studies.
The analysis employs Gaussian anamorphosis with the Kernel Cumulative Density Estimator (KCDE) to normalize the Multivariate data, with the use of a look-up Table for the back-transform of the predicted grid values. The resulting data follow the Normal Distribution N(0,1) and thus the transformed data are gaussian and the values are of the same range. A range of theoretical variogram models, (for example the Exponential, Gaussian, Spherical) as well as the previously introduced Harmonic Covariance Estimation (HCE) model, is applied to assess their suitability for co-kriging in a multivariate context. Emphasis is placed on ensuring positive-definiteness of the resulting covariance matrices through Eigenvalue analysis [1]. With more than two variables, the invertibility of the augmented covariance matrix is not ensured, even for Euclidean distances. Positive definiteness is further complicated with the utilization of non-Euclidean distance metrics such as Manhattan, Minkowski, or Chebyshev distances.
The primary goal of this study is to evaluate the comparative performance of the HCE model against traditional variogram models in modeling multivariate spatial relationships and the positive definiteness of the multivariate covariance matrices. Furthermore, the accuracy of co-kriging predictions with the various established models and HCE, both before and after the back-transform will be tested and discussed. This research extends the understanding of non-Euclidean geostatistical modeling in multivariate contexts, with potential applications in other regions with complex terrains or spatiotemporal phenomena.
The research project is implemented in the framework of H.F.R.I call “Basic research Financing (Horizontal support of all Sciences)” under the National Recovery and Resilience Plan “Greece 2.0” funded by the European Union – NextGenerationEU (H.F.R.I. Project Number: 16537)
1) Curriero, F.C.: On the use of non-euclidean distance measures in geostatistics. Mathematical Geology 38, 907–926 (2006)
How to cite: Koltsidopoulou, M. D., Pavlides, A., Chrysanthi, M., and Varouchakis, E. A.: Exploring Positive-Definiteness in Multivariate Geostatistics with Non-Euclidean Distances, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12375, https://doi.org/10.5194/egusphere-egu25-12375, 2025.
Posters virtual: Tue, 29 Apr, 14:00–15:45 | vPoster spot A
EGU25-1783 | ECS | Posters virtual | VPS9
Modeling the impact of lateral flow on terrestrial water balance components and surface energy fluxes using WRF-Hydro with multi-configuration ensembles: a study over Krishna River CatchmentTue, 29 Apr, 14:00–15:45 (CEST) | vPA.1
Spatio-temporal variability of the terrestrial hydrological processes (land heat and water storage anomalies) has important implications in the climate predictability through their effects on surface energy and water fluxes. The changes in seasonal precipitation patterns associated with the Indian Summer Monsoon can alter the hydrological processes; for a given catchment, which in turn can influence the exchange of water and energy at the land surface-atmosphere interface. Hence the reliable prediction of the basin-scale water cycle components in a physically based high-resolution hydrological model equipped with sophisticated Land Surface Models (LSMs) is of prime requirement. The modern LSMs can provide detailed representations of important biophysical, biogeochemical and hydrological processes of varying spatial and temporal scales by incorporating the necessary feedbacks between the land and the atmosphere. When coupled to a physically based fully distributed hydrological model, it can affect the soil moisture patterns means of recycling the surface and sub-surface runoff (lateral terrestrial flow). However, despite the role of lateral terrestrial hydrological processes for the improved simulation of soil moistures, the sensitivity studies involving the land surface and sub-surface feedbacks are less pronounced especially for a tropical humid region with complex physiographic settings (presence of complex topography) under monsoon regimes (strong synoptic forcings). Therefore, in the present study, we examined a process based diagnosis regarding the role of the lateral flow on the terrestrial hydrological processes (Evapotranspiration, surface and sub-surface runoff, stream flow) and surface energy fluxes (latent heat, sensible heat) by using a multi-configured modeling framework of offline WRF-Hydro with Noah-Multi parameterizations (MP) LSM to enable systematic evaluation of the multiple physical parameterizations of hydrologic process representation; the validation has been done with the reanalysis dataset, a remotely sensed product and ground based observations.
How to cite: Sarkar, S. and Lakshmikanthan, P.: Modeling the impact of lateral flow on terrestrial water balance components and surface energy fluxes using WRF-Hydro with multi-configuration ensembles: a study over Krishna River Catchment, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-1783, https://doi.org/10.5194/egusphere-egu25-1783, 2025.
EGU25-2663 | Posters virtual | VPS9
Spatial and Temporal Extreme Modeling of Daily Maximum Precipitation Based on a Generalized Additive ModelTue, 29 Apr, 14:00–15:45 (CEST) | vPA.25
South Korea experiences significant regional variation in precipitation due to its unique topographical features. Over the years, the intensification of summer rainfall concentration has led to recurring damage from floods and torrential downpours. To mitigate such impacts, the Korea Meteorological Administration monitors precipitation using observed data from weather stations and estimated values for non-observed locations. The Generalized Extreme Value (GEV) distribution is commonly employed to model annual maximum precipitation, enabling the calculation of return levels that serve as foundational data for flood prevention. This study aims to estimate the spatially generalized additive GEV distribution of daily maximum precipitation using data from 54 Automatic Synoptic Observation System (ASOS) stations between 1972 and 2024. Spatial elements (latitude, longitude, altitude) and temporal elements (year) were incorporated into the model. The location, scale, and shape parameters of the GEV distribution were estimated using the maximum likelihood method, with smoothing functions accounting for spatial and temporal factors. The results indicate that the location and scale parameters, influenced by latitude and longitude, are relatively lower in central regions, while the shape parameter, influenced by altitude, shows similar trends. Furthermore, return levels for 50-year and 100-year return periods are notably higher in mountainous regions. Goodness-of-fit tests, such as the Anderson-Darling test, were performed on the GEV distributions of 53 ASOS stations, excluding one. However, 12 stations located in island regions, high-altitude areas, or regions affected by typhoons exhibited distributions that were difficult to explain spatially. These findings are expected to aid in the development of efficient water resource management strategies and regional flood prevention measures based on the distribution characteristics of precipitation.
How to cite: Lee, B., Hwang, Y., and Yoon, S.: Spatial and Temporal Extreme Modeling of Daily Maximum Precipitation Based on a Generalized Additive Model, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2663, https://doi.org/10.5194/egusphere-egu25-2663, 2025.
EGU25-2660 | Posters virtual | VPS9
Transformed Technique for Applying the Generalized Extreme Value Distribution to Block MinimaTue, 29 Apr, 14:00–15:45 (CEST) | vPA.28
This study proposes a novel approach for analyzing block minima data within the Generalized Extreme Value Distribution (GEVD) framework by incorporating the Negative Power Transformation (NPT). The NPT method, which includes a hyper-parameter to adjust data bounds (effectively reducing to the Reciprocal Transformation (RT) when the hyper-parameter is 1), aims to improve the accuracy and robustness of long-term return level (RL) estimations. Traditional transformation methods often exhibit limitations in accurately predicting RLs for extended return periods. Through extensive Monte Carlo simulations, we demonstrate that the NPT-GEVD method outperforms conventional approaches in terms of bias, standard error (SE), and root mean square error (RMSE) for return periods of 25, 50, and 100 years. Notably, the NPT-GEVD consistently provides reliable RL estimates across various parameterizations and sample sizes, particularly when using L-moments for estimation with smaller datasets. The efficacy of the NPT-GEVD method is further validated through its application to inter-arrival time (IAT) rainfall data from South Korea. The analysis revealed that RLs for detecting the time to exceed hourly cumulative rainfall thresholds of 60 mm, 90 mm, and 110 mm varied significantly across locations, ranging from 30 minutes to over 4 hours. This research underscores the significance of advanced transformation techniques in enhancing the accuracy and reliability of environmental risk assessments. The NPT-GEVD method offers valuable insights for improving flood prediction and mitigation strategies in the face of climate change.
How to cite: Yoon, S. and Prahadchai, T.: Transformed Technique for Applying the Generalized Extreme Value Distribution to Block Minima, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2660, https://doi.org/10.5194/egusphere-egu25-2660, 2025.
