HS2.2.3 | Advancing process representation for hydrological modelling across spatio-temporal scales
EDI
Advancing process representation for hydrological modelling across spatio-temporal scales
Convener: Elham R. Freund | Co-conveners: Simon Stisen, Björn Guse, Luis Samaniego, Sina Khatami
Orals
| Thu, 18 Apr, 10:45–12:30 (CEST), 14:00–18:00 (CEST)
 
Room 3.16/17
Posters on site
| Attendance Fri, 19 Apr, 10:45–12:30 (CEST) | Display Fri, 19 Apr, 08:30–12:30
 
Hall A
Orals |
Thu, 10:45
Fri, 10:45
Understanding and representing hydrological processes is the basis for developing and improving hydrological and Earth system models. Relevant hydrological data are becoming globally available at an unprecedented rate, opening new avenues for modelling (model parametrization, evaluation, and application) and process representation. As a result, a variety of models are developed and trained by new quantitative and qualitative data at various temporal and spatial scales.
In this session, we welcome contributions on novel frameworks for model development, evaluation and parametrization across spatio-temporal scales.

Potential contributions could (but are not limited to):
(1) introduce new global and regional data products into the modeling process;
(2) upscale experimental knowledge from smaller to larger scale for better usage in catchment models;
(3) advance seamless modeling of spatial patterns in hydrology and land models using distributed earth observations;
(4) improve model structure by representing often neglected processes in hydrological models such as human impacts, river regulations, irrigation, as well as vegetation dynamics;
(5) provide novel concepts for improving the characterization of internal and external model fluxes and their spatio-temporal dynamics;
(6) introduce new approaches for model calibration and evaluation, especially to improve process representation, and/or to improve model predictions under changing conditions;
(7) develop novel approaches and performance metrics for evaluating and constraining models in space and time

This session is organized as part of the grass-root modelling initiative on "Improving the Theoretical Underpinnings of Hydrologic Models" launched in 2016.

Orals: Thu, 18 Apr | Room 3.16/17

Chairperson: Elham R. Freund
10:45–10:50
10:50–11:10
|
EGU24-21063
|
solicited
|
Highlight
|
On-site presentation
Lieke Melsen

Science, despite its status as objective and searching for truth, is inherently a social activity. Research is conducted by scientists that collaborate, work in teams, get advised by their supervisor, get funding to study particular questions, know one another from earlier projects, and so on. In these social interactions, we together define what we consider important to study, or what we deem unimportant. This occurs at multiple levels: Funding agencies, for example, have the power to determine which research questions should be addressed. As hydrological modelling community, we have implicitly agreed that discharge is the main variable of interest - focusing on other fluxes or states is often presented as an advancement.  And at the modelling team level, we often (implicitly) agree on a modelling vision. From interviews with modellers from different teams, it for example became apparent that one team had the modelling vision to `keep things as simple as possible’. Given this vision, the modeller was inclined to choose the simpler parameterization over a more complex one to describe the same process. In another team, ‘scale invariance’ was considered more important, and therefore process representations were selected based on their scalability. Therefore, if we want to “advance” process-representation in models across spatial and temporal scales, the theme of this session, we should acknowledge that different researchers have different perceptions of what advancing comprehends, that there is no objective measure to define advancement, and that the first step probably is, that we have to clarify and express our modelling vision.

How to cite: Melsen, L.: The sociology of modelling: how we shape a perception together, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-21063, https://doi.org/10.5194/egusphere-egu24-21063, 2024.

11:10–11:20
|
EGU24-16034
|
ECS
|
On-site presentation
Wouter Knoben, Martyn Clark, Ying Fan, Hilary McMillan, Jordan Read, and Katie van Werkhoven

The North American continent is home to a wide range of different hydro-climates. A key research gap is that there is currently limited understanding on the spatial variability of dominant hydrologic processes across these different hydro-climates. This limited understanding makes it difficult to select computational models that faithfully represent the hydrologic processes across such large domains, yet faithful representation of the different hydro-climatic behaviors is critical for accurate numerical prediction.

Here we present progress on a synthesis of dominant hydrologic processes under different combinations of climate-terrain-human forcings, engaging the broader community of catchment and Critical Zone scientists. The product from this research will be a continental “Hydrologic Mosaic”, with each landscape in the mosaic described by a set of perceptual and conceptual models. In this first step, we produce a continental map of hydrologic landscapes defined through the juxtaposition of hydroclimate, terrain and geology, and vegetation, land use, and management. We will define hydrologically meaningful indicators of terrestrial hydrology that concisely describe a location’s (i) hydroclimate (e.g., aridity, snow fraction, energy/water seasonality), (ii) topography and geology (e.g. depth to bedrock, soil porosity, topographic slope), and (iii) vegetation, land use and management (e.g., vegetation type, agricultural drainage, reservoir size), and calculate values for these indicators for each location on the continent. We then use clustering analysis to create a manageable number of representative hydrologic landscapes.

This work functions as a starting point in a wider project, where these initial hydrologic landscapes will be refined through interactions with regional experts. Together, we will develop perceptual (sketches and descriptions) and conceptual (box-and-arrow diagrams) of the dominant processes in each hydrologic landscape. These conceptual diagrams will contribute to large-domain modeling efforts by allowing targeted model selection and comparison efforts for each hydrologic landscape.

How to cite: Knoben, W., Clark, M., Fan, Y., McMillan, H., Read, J., and van Werkhoven, K.: Developing perceptual models of hydrologic behavior across the North American continent, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16034, https://doi.org/10.5194/egusphere-egu24-16034, 2024.

11:20–11:30
|
EGU24-973
|
ECS
|
On-site presentation
Ana Clara de Sousa Matos, Marvin Höge, Thiago Victor Medeiros do Nascimento, Gustavo de Oliveira Corrêa, Francisco Eustáquio Oliveira e Silva, and Fabrizio Fenicia

Brazil faced a severe water crisis during the mid-2010s, resulting in water scarcity and water rationing in various cities. The Belo Horizonte Metropolitan Region was seriously affected. It is located in the southeastern part of the country and home to roughly 5 million people (Costa et al., 2015) and mining industry. The region’s water supply relies on a complex and integrated system, which combines a water abstraction at the Velhas river and three reservoirs. One of these reservoirs, named Serra Azul, reached a minimum of only 5,4 % of its total capacity during the crisis. Here, we demonstrate tools for improving the water management in this area, by developing a hydrological model suitable for mountainous regions with tropical climates. Our case study was the Serra Azul reservoir’s well-gauged catchment. We selected 12 gauges that cover several head waters and rivers section in the 260 km² area.  We used these discharge data (3-5 years), and available static catchments' attributes (e.g. subsurface properties), to adapt a flexible framework for conceptual hydrological modeling. Hence, we identified a suitable model structure using SUPERFLEX (Fenicia et al., 2014). The findings show that by including soil type, lithology and land cover as explanatory variables in the model, we obtained significant improvements in performance, e.g. the correlation between the base flow index estimated for observed and simulated time-series increased from 0.40 to 0.76. We also accounted for groundwater contributions to the streamflow, modelling the relation between the percentage of porous aquifer within each catchment and its flow magnitude. Thereby, we improved the average NSE and timeseries correlation considerably.  Overall, we successfully set up a parsimonious hydrologic model for water resources management in a region that is notoriously difficult to predict, where anthropic activities such as mining and agriculture have a decisive impact on the water cycle.

 

References:

Costa et al. Caracterização e Quadros de Análise Comparativa da Governança Metropolitana no Brasil: análise comparativa das funções públicas de interesse comum (Componente 2)-RM do Rio de Janeiro (Relatório de Pesquisa). (2015). Rio de Janeiro: Institute for Applied Economic Research–Ipea.

Fenicia et al. "Catchment properties, function, and conceptual model representation: is there a correspondence?." Hydrological Processes 28.4 (2014): 2451-2467.

How to cite: de Sousa Matos, A. C., Höge, M., Medeiros do Nascimento, T. V., de Oliveira Corrêa, G., Oliveira e Silva, F. E., and Fenicia, F.: Model development for a water supply catchment in southeast Brazil, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-973, https://doi.org/10.5194/egusphere-egu24-973, 2024.

11:30–11:40
|
EGU24-4385
|
ECS
|
On-site presentation
Cristina Prieto, Nataliya Le Vine, Dmitri Kavetski, Fabrizio Fenicia, Andreas Scheidegger, and Claudia Vitolo

Hydrological modelling of ungauged catchments, which lack observed streamflow data, is an important practical goal in hydrology. A major challenge is to identify a model structure that reflects the hydrological processes relevant to the catchment of interest. Paraphrasing a well-known adage, “all models are wrong, but some model-mechanisms (process representations) might be useful.”

We extend a method previously introduced for mechanism identification in gauged basins, by formulating the Bayesian inference equations in the space of (regionalized) flow indices principal components and by accounting for posterior parameter uncertainty. We use a flexible hydrological model to generate candidate mechanisms and model structures. Then, we use statistical hypothesis testing to identify the "dominant" (more a posteriori probable) hydrological mechanism. We assume that the error in the regionalization of flow indices principal components dominates the error of the hydrological model structure.

The method is illustrated in 92 catchments from northern Spain. We treat 16 out of the 92 catchments as ungauged. We use 624 model-structures from FUSE (flexible hydrological model framework). The case study includes real data and synthetic experiments.

The findings show that routing is among the most identifiable processes, whereas percolation and unsaturated zone processes are the least identifiable. The probability of making an identification (correct or wrong), remains stable at ~25%, both in the real and in the synthetic experiments. In the synthetic experiments, where the “true” mechanism is known, we can evaluate the reliability, i.e., the probability of identifying the true mechanism when the method makes an identification. Reliability varies between 60%-95% depending on the magnitude of the combined regionalization and hydrological error. The study contributes perspectives on hydrological mechanism identification under data-scarce conditions.

Prieto et al. (2022) An Exploration of Bayesian Identification of Dominant Hydrological Mechanisms in Ungauged Catchments, WRR58(3), doi:10.1029/2021WR030705.

How to cite: Prieto, C., Le Vine, N., Kavetski, D., Fenicia, F., Scheidegger, A., and Vitolo, C.: Can we identify dominant hydrological mechanisms in ungauged catchments?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4385, https://doi.org/10.5194/egusphere-egu24-4385, 2024.

11:40–11:45
11:45–11:55
|
EGU24-19807
|
ECS
|
On-site presentation
|
Ioannis Sofokleous, Adriana Bruggeman, Corrado Camera, Hakan Djuma, Mohsen Amini Fasakhodi, and George Zittis

We tested the widely used atmospheric WRF (Weather Research and Forecasting) model in a coupled mode to the hydrological model WRF-Hydro. The coupled WRF/WRF-Hydro model adds the simulation of horizontal surface and subsurface flow of water relative to the standalone WRF. We conducted simulations for the Mediterranean island of Cyprus and 31 small mountainous river basins for the hydrological year 2011-2012. We found higher soil moisture (20%), more evapotranspiration (33%) and a small increase in rainfall (3%) for the coupled WRF/WRF-Hydro model, compared to the WRF model without horizontal flows. We also forced WRF-Hydro with observed rainfall and five different set-ups of WRF and examined the modelled streamflow. The WRF set-ups were adapted from combinations of different microphysics, cumulus cloud, planetary boundary layer and surface layer schemes. We found that WRF-Hydro with observed rain underestimated the average streamflow by 6%, during a two-year simulation (2011-2013). The best of the five WRF set-ups showed a 19% underestimation of the average streamflow, thus, an optimized ensemble of WRF set-ups is needed to model the streamflow. Our study suggests that the coupling of WRF with the WRF-Hydro model can improve land-atmosphere simulations. We will also present the calibration of parameters of the land surface component of the coupled model with observations of soil moisture and transpiration that could further enhance the ability of the model to represent the different parts of the combined terrestrial-atmospheric water cycle.

How to cite: Sofokleous, I., Bruggeman, A., Camera, C., Djuma, H., Amini Fasakhodi, M., and Zittis, G.: Simulations of energy and water balances with WRF and WRF-Hydro models: the role of model coupling and parameterizations, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19807, https://doi.org/10.5194/egusphere-egu24-19807, 2024.

11:55–12:05
|
EGU24-2729
|
On-site presentation
Yuanxin Song, Yanjun Zhang, Shanqi Li, and Jiahui Yang

Abstract: Subsurface stormflow is the dominant runoff generation mechanism during certain flash flood events. The collapse characteristics and threshold behavior are critical to the transition from the slow runoff stage to the rapid runoff stage and the relevant study is not only essential for hydrology theory but also for flash flood disaster prevention. In recent years, efforts have been made to explore the common principles of hydrological processes under strong spatial heterogeneity, but findings from field experiments and numerical studies were difficult to apply to the modeling process. Furthermore, we have yet to develop a deeper understanding of the mechanisms for the impact of complex factors on subsurface stormflow and lack a comprehensive understanding of the formation mechanism of threshold.
This presentation discusses how we plan to address this research gap. Firstly, around the phenomenon of “burst-block-burst” in the subsurface stormflow runoff generation process, rainfall-runoff simulation experiments were carried out and factor analysis was conducted to determine the main influencing factors of subsurface stormflow runoff generation. The main influencing factors include soil texture factors, collapse state factors, initial state factors, and other factors, and the influence of these four types of factors decreases in turn. In the second step, we constructed a field hydrological station in the Huanggou Watershed located in Hubei Province, China, collected the rainfall-runoff data, and found that the subsurface stormflow process shows a three-stage-double-threshold behavior: the water storage stage, the initial flow stage, and the rapid flow stage. In the third step, synthesizing the main influencing factors, the three-stage double-threshold process was quantified. Further, the three-stage subsurface stormflow-based model (TSSM) was developed and applied to the Huanggou Hillslope and the Huanggou Watershed. The results show that TSSM performed well, with NSEs of 0.82 and 0.67 in the calibration and verification periods of the Huanggou slope, and NSEs of 0.76 and 0.74 in the calibration and verification periods of the Huanggou Watershed, respectively.
This study elucidated the collapse characteristics and threshold behavior of subsurface stormflow and developed an effective simulation model, which contributes to increasing our understanding of three-stage subsurface stormflow and is beneficial for hydrologists to develop more realistic hydrological models.

Keywords: Subsurface stormflow; Collapse characteristics; Threshold behavior; Three-stage subsurface stormflow mechanism; TSSM

How to cite: Song, Y., Zhang, Y., Li, S., and Yang, J.: Study on the collapse characteristics and threshold behavior of the subsurface stormflow mechanism, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2729, https://doi.org/10.5194/egusphere-egu24-2729, 2024.

12:05–12:15
|
EGU24-13230
|
On-site presentation
Tricia Stadnyk, Mohamed Ismaiel Ahmed, Martyn Clark, and Alain Pietroniro

Hydrologic modelling in the low-lying, flat prairie or arctic pothole regions is challenging because of variable contributing areas that modify the transformation of local runoff into streamflow. Most hydrological and land surface models fail in predicting prairie hydrology due to overlooking or inadequately representing the variable contributing area dynamics. In this study, we develop an open-source, model-agnostic version of a revised formulation of the recently developed Hysteretic Depressional Storage (HDS) model. This revised formulation accounts for the hysteretic relationship of pothole depressions and its effects on streamflow generation. The revised HDS model is implemented and tested with two different hydrological models of varying complexity (MESH and HYPE). The modified hydrological models are tested on a number of prairie pothole basins in Canada. Results show improved simulations of the streamflow response in the tested basins. Importantly, the modified models replicate the known hysteretic relationships between depressional storage and contributing areas in that region. The open-source HDS implementation approach is designed for use in hydrologic or land surface modelling systems, enabling improvements in simulating the complex hydrology and streamflow regimes globally.

How to cite: Stadnyk, T., Ahmed, M. I., Clark, M., and Pietroniro, A.: An Improved Representation of The Variable Contributing Area Concept in Hydrologic and Land Surface Models, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13230, https://doi.org/10.5194/egusphere-egu24-13230, 2024.

12:15–12:25
|
EGU24-5982
|
ECS
|
On-site presentation
Zahra Eslami, Khodayar Abdollahi, and James Kirchner

Hydrological losses play a significant role in determining the runoff coefficient, influencing the amount of precipitation that ultimately contributes to surface runoff. These losses are influenced by various characteristics, including rainfall, the physical and geographical attributes of the watershed (such as slope and land use), and the soil moisture content within the watershed.

Here, we evaluate the effect of the variability of the loss function on the amount of simulated runoff and the runoff coefficient in specific watersheds in Iran. The investigation entails the assessment of two distinct conditions. First, the runoff coefficient is calculated under the assumption of a constant loss, utilizing the φ index. Second, a variable loss function, derived from a soil moisture algorithm, is employed to determine the runoff coefficient.

Our analysis shows that the assumption of a variable loss function yields more realistic results. When the variable losses are considered, the simulated runoff coefficient is closer to the observed values and determines the runoff coefficient for all months, including those characterized by low rainfall. The constant loss φ index exhibits two significant practical limitations: the overestimation of runoff coefficient values, and an inability to estimate runoff coefficient during months with low rainfall. The study emphasizes the need for a variable loss function to provide more realistic results. Our findings suggest that utilizing the variable loss function within the soil moisture algorithm produces more accurate results. Thus, the application for improved forecasting of rainfall and runoff processes is recommended.

 

How to cite: Eslami, Z., Abdollahi, K., and Kirchner, J.: Analyzing the fixed or variable effect of considering hydrological loss functions on the runoff coefficient in continuous modeling, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5982, https://doi.org/10.5194/egusphere-egu24-5982, 2024.

12:25–12:30
Lunch break
Chairperson: Björn Guse
14:00–14:20
|
EGU24-19135
|
ECS
|
solicited
|
Highlight
|
On-site presentation
Keirnan Fowler, Dominic Regan-Beasley, Michael Nixon, and Glen Walker

While there is little opposition to the idea that groundwater can play a central role in rainfall-runoff response, there is little consensus on how this should be modelled. Here, we present modelling supporting a recently advanced hypothesis that links groundwater-surface water interactions with observed shifts in the relationship between rainfall and runoff in south-east Australia.  While many approaches assume a direct and simplistic relationship between groundwater head and baseflow, evidence arising from a multi-year drought in Australia challenges traditional notions. GRACE and bore data during the Millennium Drought (1997-2010) show multi-year declines in groundwater storage, of such severity that we might expect the baseflow to cease, giving a flashier regime.  In reality, the shape of the hydrograph is mostly unchanged, but other changes abound: a year of given rainfall generates less runoff today than it did pre-drought (ie. shift in rainfall-runoff relationship). In other words, during and after the drought we see a hydrograph of similar shape to before, but diminished. While many Australian hydrologists are convinced that groundwater played a key role in this behaviour, it is unclear how these observations can be explained by existing hypotheses or modelling methods for groundwater surface-water interaction, and new paradigms are required.

The hypothesis explored here is that these observations can be explained by leaky bedrock in headwater catchments, which facilitates gradual groundwater export from upslope areas to downslope areas (within the same catchment).  Upslope areas subject to groundwater decline then see groundwater-surface water decoupling and reduced runoff. The hypothesised leakage is slow enough to go unnoticed during wetter periods, but in drier periods recharge may be too low to balance the export, leading to reduced groundwater levels and groundwater surface-water decoupling. When wetter conditions resume, the groundwater deficits may take a while to be replenished, delaying recovery of rainfall-runoff relationships (as observed in Australia).  In downslope areas, the drained water may contribute to streamflow, but may also be lost to evaporation and transpiration, particularly in drier catchments with flatter valley bottoms of alluvium or colluvium. In such catchments, the net effect of these processes is to allow groundwater originating from upslope to supplement evaporative budgets downslope rather than increasing streamflow.

We advance this hypothesis, firstly by presenting evidence of its applicability in south-east Australia; and secondly by building and testing improved numerical models that incorporate a simplified representation of these processes. Modelling results show improved performance when tested across several catchments affected by rainfall-runoff relationship changes, and improved realism such as multi-year declines in simulated groundwater storage, consistent with observations.  These results suggest a promising avenue for further research relevant to a variety of water resource applications including climate change impact assessment.

How to cite: Fowler, K., Regan-Beasley, D., Nixon, M., and Walker, G.: New modelling paradigm linking groundwater, surface water and rainfall-runoff relationship shifts under multi-year drought, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19135, https://doi.org/10.5194/egusphere-egu24-19135, 2024.

14:20–14:30
|
EGU24-9140
|
ECS
|
On-site presentation
Daniele la Cecilia, Accursio Venezia, Davide Maino, and Matteo Camporese

The occurrence of agricultural catchments covered by plastic greenhouses is growing worldwide. Horticultural greenhouse production allows for saving irrigation water at the farm scale, but also alters the natural hydrological cycle. Currently, these alterations are not accounted for in physics-based hydrological models. In this study, we aim to couple the greenhouse climate model KASPRO1, to estimate indoor crop transpiration from outdoor meteorological variables, with the integrated surface-subsurface hydrological model CATchment HYdrology (CATHY2) to simulate the stream discharge as well as the shallow groundwater depth in an agricultural catchment (11 km2) covered by plastic greenhouses in South Italy. The dynamic presence of greenhouses, along with bare soils and vegetated lands, is mapped with the Open field and Protected Agriculture land cover Classifier (OPAC3).

We first compare our simulations against indoor measurements of water use (drip- and sub-irrigation) and soil moisture dynamics at different depths at the plot scale. Next, we run CATHY at the catchment scale and compare the output against measured stream water level.

The aim of our study is to validate the capabilities of KASPRO and CATHY to provide high-fidelity spatially distributed dynamic simulations of evapotranspiration and irrigation fluxes, as well as soil moisture and groundwater flows. Such capabilities are essentials to understand the implications of plastic greenhouse districts on the hydrological cycle and thus making these models useful tools for a more sustainable management of agricultural catchments.

References

1 De Zwart, H.F., 1996. Analyzing Energy-Saving Options in Greenhouse Cultivation Using a Simulation Model. Landbouwuniversiteit, Wageningen.

2 Camporese, M., Paniconi, C., Putti, M., & Orlandini, S. (2010). Surface--subsurface flow modeling with path-based runoff routing, boundary condition-based coupling, and assimilation of multisource observation data. Water Resources Research, 46, W02512.

3 la Cecilia, D., Tom, M., Stamm, C., Odermatt, D., 2023. Pixel-based mapping of open field and protected agriculture using constrained Sentinel-2 data. ISPRS Open Journal of Photogrammetry and Remote Sensing 8. https://doi.org/10.1016/j.ophoto.2023.100033.

 

Acknowledgements: We thank the Consorzio di Bonifica in Destra del Fiume Sele for the continuous support in the MSCA-PF REWATERING project.

Funding: This project has received funding from the European Union’s Horizon Europe research and innovation under the Marie Skłodowska-Curie grant agreement No. 101062255

How to cite: la Cecilia, D., Venezia, A., Maino, D., and Camporese, M.: Towards an understanding of the hydrological processes of greenhouse horticulture districts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9140, https://doi.org/10.5194/egusphere-egu24-9140, 2024.

14:30–14:40
|
EGU24-12534
|
ECS
|
On-site presentation
Qiuyu Zhu, Megan Klaar, Thomas Willis, and Joseph Holden

While natural flood management (NFM) as a resilience flood mitigation strategy is widely used in the UK and Europe, there remains a lack of scientific evidence regarding its effectiveness. The primary uncertainties stem from two aspects: the determination of NFM effectiveness on flood mitigation is limited by the scale of impact assessment; and the combination of multiple NFM interventions implemented within a catchment which may result in flood synchronicity. We argue that the effectiveness of combined scenarios involving multiple NFM interventions within a catchment can vary.  We utilize a hydrological model that simulates both instream and terrestrial interventions at a large catchment scale. To demonstrate how scale and interventions interact to determine flood peaks, we integrated various NFM interventions and land cover changes within the upstream catchment into a model, including afforestation, soil aeration, catchment/floodplain restoration and hedge planting. We modelled existing and planned scenarios using a spatially distributed hydrological model, Spatially Distributed TOPMODEL (SD-TOPMODEL). In comparison to previous versions of TOPMODEL, we have improved the simulation efficiency to allow for the simulation of up to a 200-year return period flood event at a larger catchment scale (~84 km2); and simplified the model parameters which are not related to the effects of NFM interventions and retained three key parameters which are physically significant. Following extensive parameter calibration and validation, the model is stable, providing a reliable fit for flood peaks, with the Nash-Sutcliffe coefficient (NS) between modelled and observed discharge reaching up to 0.905. The modelling results illustrated the effectiveness of NFM interventions in reducing flood peaks at a large catchment scale. Further refinements will involve incorporating additional types of NFM interventions into our next coupled model. 

How to cite: Zhu, Q., Klaar, M., Willis, T., and Holden, J.: Modelling the effectiveness of multiple natural flood management interventions at a large catchment scale, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12534, https://doi.org/10.5194/egusphere-egu24-12534, 2024.

14:40–14:50
|
EGU24-9070
|
Virtual presentation
Dmitri Kavetski, Mark Thyer, David McInerney, Hoshin Gupta, Seth Westra, Holger Maier, Anthony Jakeman, Barry Croke, Daniel Partington, Margaret Shanafield, Craig Simmons, and Christina Tague

The ability of contemporary hydrological models to serve as a basis for credible prediction and decision making is increasingly challenged – especially as hydrological systems are pushed outside the envelope of historical experience. Conceptual models are the most common type of surface water hydrological model used for decision support due to reasonable performance in the absence of change, ease of use and computational speed that facilitate scenario, sensitivity and uncertainty analysis. Hence, conceptual models arguably represent the current "shopfront" of hydrological science as seen by practitioners. However, these models have notable limitations in their ability to resolve internal catchment processes and subsequently capture hydrological change. New thinking is needed to confront the challenges faced by the current generation of conceptual models in dealing with a changing environment. We argue that the next generation of conceptual models should combine the parsimony of conceptual models with our best available scientific understanding. We propose a strategy to develop such models using multiple hydrological lines of evidence. This strategy includes using appropriately selected physically-resolved models as "Virtual Hydrological Laboratories" to test and refine the simpler models' ability to predict future hydrological changes. This approach moves beyond the sole focus on "predictive skill" measured using metrics of historical performance, facilitating the development of the next generation of conceptual models with hydrological fidelity - i.e., that "get the right answers for the right reasons". This quest is more than a scientific curiosity – it is expected by environmental policy makers and broader stakeholders.

How to cite: Kavetski, D., Thyer, M., McInerney, D., Gupta, H., Westra, S., Maier, H., Jakeman, A., Croke, B., Partington, D., Shanafield, M., Simmons, C., and Tague, C.: Virtual Hydrological Laboratories to develop the next generation of conceptual models and support decision-making under change, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9070, https://doi.org/10.5194/egusphere-egu24-9070, 2024.

14:50–14:55
14:55–15:05
|
EGU24-19765
|
ECS
|
On-site presentation
Alejandro Sánchez-Gómez, Jeffrey Arnold, Katrin Bieger, Nancy Sammons, Silvia Martínez-Pérez, and Eugenio Molina-Navarro

SWAT+ is a completely restructured version of the SWAT model. It includes new capabilities, and the possibility of modelling water resources management is particularly relevant. A new water allocation module allows to allocate water for different purposes inside and outside a basin. Reservoir management can be modelled using decision tables that define which actions occur under different scenarios. Despite the relevance of these novelties, there is a lack of studies demonstrating accurate simulations of reservoir outflows using decision tables in SWAT+, and there are to date no publications regarding the water allocation module.

The Tagus River basin (Spain) is the most populated (11 million inhabitants) basin on the Iberian Peninsula and its water resources management is highly controversial. This basin is a clear example of the importance of including anthropogenic water management in the modelling process, since it is intensively regulated by more than 50 reservoirs, several water transfers, and irrigation. Therefore, the water allocation module (for simulating water transfers and irrigation) and decision tables (for reservoir management and irrigation) were used in a detailed model of the Upper Tagus River Basin (UTRB), where most of the water demands of the basin are located.

Firstly, more than 30 reservoirs were introduced to the model and their management was analyzed using observed data. Different decision table structures were created considering the properties of the reservoirs (purpose, storage, etc.) and then adapted to each of the reservoirs. A satisfactory simulation of reservoir storage and outflow was achieved in most of the cases, demonstrating the reliability of the model and the adequacy of this approach.

There are numerous water transfers in the UTRB, of which the Tagus-Segura water transfer (TSWT) is the most relevant one. Some transfer water from one reservoir to another, while two of them divert water outside the modelled basin. In addition, water is transferred from reservoirs to water treatment plants and subsequently released to selected receiving channels. All these transfers were modelled using the SWAT+ water allocation module and for most of them the modelled volumes matched the observed ones well.

The agricultural water demand was estimated from the River Basin Management Plan. To simulate the irrigation, all demand objects within the UTRB (irrigated agricultural lands, 282 objects) and their respective sources (closest channel to those objects, 118 sources) were identified. An irrigation decision table was developed for the basin, allowing to simulate a demand close to the calculated and to supply enough water to meet more than 80% of this demand.

This works presents a novel approach to simulating water resources management in a highly regulated river basin using SWAT+. Results shows a satisfactory simulation of different management actions (reservoirs, irrigation, water transfers inside and outside the basin, wastewater discharges). Further work on the water allocation module will boost even more the application of SWAT+.

How to cite: Sánchez-Gómez, A., Arnold, J., Bieger, K., Sammons, N., Martínez-Pérez, S., and Molina-Navarro, E.: Application of the new SWAT+ water allocation module in the Tagus River basin, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19765, https://doi.org/10.5194/egusphere-egu24-19765, 2024.

15:05–15:15
|
EGU24-4058
|
ECS
|
On-site presentation
Konstantinos Soulis, Stergia Palli Gravani, and Dionissios Kalivas

One the most difficult challenges in applied hydrology is predicting runoff in ungauged or poorly gauged watersheds. Thus, simple approaches for runoff estimation are especially useful in hydrologic applications. A simple, well established, and widely used technique for predicting the direct runoff depths of rainfall events is the Soil Conservation Service - Curve Number (SCS-CN) method. Due to its straightforward but well-proven approach, readily available and well documented environmental inputs, and incorporation of numerous variables influencing runoff generation into a single CN parameter, it quickly rose to prominence among engineers and practitioners. Tables can be used to identify the CN parameter values corresponding to prevailing soil, land cover and land management conditions. However, it is always better to estimate the CN value using observed rainfall-runoff (P-Q) data when available. Estimating appropriate CN values for additional soil – land cover conditions and additional regions is also critical for extending and updating the method’s documentation given that the SCS-CN approach is extremely sensitive to variations in the CN values.

However, even when the CN value is determined from measured P-Q data, the estimated CN values vary substantially from storm to storm on any given watershed. For this reason, various methods to estimate the CN value characterizing each watershed have been proposed up to know, and many theories on the reasons behind the observed relationships between CN and P for each watershed have been stated. Though, after many years of research, there isn’t still a unique agreed method to estimate the CN values characterizing a watershed or a soil-land cover complex, while the proposed methods lead to different CN values and in many cases neglect spatial variability. Further, an increasing number of modified SCS-CN versions are continuously developed, and new parameters are introduced complicating the situation even more.

Accordingly, this study attempts to collect, categorize, and systematically analyze the huge number of studies on SCS-CN method published in the last 30 years. We selected this period as 30 years ago, in 1993, R.H. Hawkins published his emblematic study on the “Asymptotic determination of runoff curve numbers from data” (J. Irrigat. Drain. Div. ASCE, 119(2): 334–345). In this review study, specific attention is given to the methods focusing on CN value determination from measured P-Q data. The advantages and limitations of the various approaches are investigated, as well as trends and gaps in existing literature. The analysed methods are classified and the main paths are identified. Based on the obtained results, conclusions on the current status are being made, and the more promising approaches are highlighted.  Then, ideas on future research pathways towards the target of a unified CN values determination approach are discussed.

How to cite: Soulis, K., Palli Gravani, S., and Kalivas, D.: SCS-CN parameter determination from observed rainfall runoff data. A critical review., EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4058, https://doi.org/10.5194/egusphere-egu24-4058, 2024.

15:15–15:25
|
EGU24-18302
|
On-site presentation
Clement Roques, Ronan Abhervé, Etienne Marti, Nicolas Cornette, Jean-Raynald de Dreuzy, David Rupp, Alexandre Boisson, Sarah Leray, Philip Brunner, and John Selker

Due to the difficulties of gathering relevant data of groundwater systems and the lack of fundamental physically-based understanding on the processes involved, the representation of groundwater flow heterogeneity in catchment- to regional-scale hydrological models is often overlooked. We often limit the representation of groundwater with simplified homogeneous and shallow aquifers where effective hydraulic properties are derived from global-scale database. This raises questions regarding the validity of such models to quantify the potential impacts of climate change, where subsurface heterogeneity is expected to play a major role in their short- to long- term regulation.

We will present the results of a numerical modelling experiment designed to explore the role of the vertical compartmentalization of hillslopes on groundwater flow and recession discharge. We found that, when hydraulic properties are vertically compartmentalized, streamflow recession behaviour may strongly deviate from what is predicted by groundwater theory that considers the drainage of shallow reservoirs with homogeneous properties. We further identified the hillslope configurations for which the homogeneous theory derived from the Boussinesq solution approximately holds and, conversely, for those for which it does not. By comparing the modelled streamflow recession discharge and the groundwater table dynamics, we identify the critical hydrogeological conditions responsible for the emergence of strong deviations. We further present new solutions to better represent subsurface heterogeneity in catchment-scale models and calibrate hydraulic parameters that properly capture the groundwater and streamflow dynamics.

How to cite: Roques, C., Abhervé, R., Marti, E., Cornette, N., de Dreuzy, J.-R., Rupp, D., Boisson, A., Leray, S., Brunner, P., and Selker, J.: Recession discharge from compartmentalized bedrock hillslopes: hydrogeological processes and solutions for model calibration , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18302, https://doi.org/10.5194/egusphere-egu24-18302, 2024.

15:25–15:35
|
EGU24-7129
|
On-site presentation
Yu-Fen Huang and Yinphan Tsang

Flash floods, characterized by rapid streamflow response to rainfall, pose a significant natural hazard, particularly in small tropical watersheds (< 40 km2). Understanding the role of rainfall event characteristics, including amount, intensity, and spatial structure, is crucial for addressing and predicting flash floods. This study employs the WRF-Hydro® model with high-resolution (250-m, hourly) rainfall data and the Random Balance Designs – Fourier Amplitude Sensitivity Testing (RBD-FAST) method to investigate how rainfall impacts streamflow, specifically peak flow events, in seven watersheds on Oʻahu, USA.

Analyzing storm events from 2015 to 2020, we examined peak flow responses to corresponding rainfall event characteristics and estimated their contributions to model efficiency. In addition, (1) random redistribution of rainfall and (2) spatial shifting of rainfall were experimented with to assess the sensitivity of peak flow to rainfall event characteristics. Not only the rainfall amount and intensity but heavy rainfall areas (>= 25 mm) within an event also exerted a significant impact on peak flow, while other spatial features contributed varying degrees of influence. Notably, spatially shifting rainfall for at least 250-m in any direction highly affected event peak streamflow, emphasizing the importance of rainfall amount, intensity, heavy rainfall areas, total rainfall areas, and connectivity among rainfall areas.

Given the significance of rainfall's spatial heterogeneity, these findings underscore the benefits of incorporating rainfall spatial characteristics in probabilistic flood forecasting and the mitigation of flood risks. This research contributes valuable insights for enhancing flood prediction strategies in small tropical watersheds, providing a basis for informed decision-making and risk management.

How to cite: Huang, Y.-F. and Tsang, Y.: Sensitivity analysis of streamflow responses to varied rainfall spatial patterns in small tropical watersheds, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7129, https://doi.org/10.5194/egusphere-egu24-7129, 2024.

15:35–15:45
Coffee break
Chairperson: Simon Stisen
16:15–16:35
|
EGU24-13626
|
ECS
|
solicited
|
On-site presentation
Luiz Bacelar, Hongli Liu, Guoquiang Tang, Noemi Vergopolan, Naoki Mizukami, Andy Wood, and Nathaniel Chaney

Efficiently modeling water and energy fluxes across spatial scales has historically involved grouping landscapes based on hydrological similarities. The HydroBlocks (HB) modeling framework, using unsupervised machine learning of high-resolution environmental datasets, emerges as a robust tool for representing heterogeneity in Land Surface Models (LSMs). This framework effectively discretizes complex gridded LSMs, such as Noah-MP, into spatially unstructured Hydrological Response Units (HRUs), facilitating the modeling of hydrological processes at hyper-resolution (10-100 m) with computational efficiency suitable for continental and global simulations. However, extending process-based hydrological models to such scales does not inherently ensure heightened simulation accuracy. For operational purposes, especially in flood warning systems, calibrating new LSMs remains imperative. Therefore, this study proposes a spatial parameter sensitivity methodology based on the pyVISCOUS algorithm, with the potential to facilitate HRU-level parameter calibration and enhance the application of hyper-resolution resolving LSMs for real-time streamflow prediction.

Our investigation delves into the relationship between spatial parameter sensitivity and model discretization across the Contiguous United States (CONUS), mainly focusing on surface and subsurface runoff states. Two clustering architectures were used to generate HB HRUs for an ensemble of simulations varying Noah-MP LSM parameters. The simplified HB configuration clusters HRUs based on terrain and hillslope variations, while the formal HB incorporates finer-scale land heterogeneity from high-resolution land cover and soil properties maps. Results reveal that saturated hydraulic conductivity was considered the most sensitive parameter for runoff production independent of the HRU grid configuration. The infiltration controlling parameter REFDK was ranked as the second most important in first-order sensitivity and had a higher spatial impact (% of HRUs) over the experiment with a higher level of clustering small-scale heterogeneity. Lower sensitivities were found in HRUs classified as urban areas, while soil properties parameters demonstrate reduced sensitivity near streams, where the floodplain remains closer to saturation. We intend to demonstrate that excluding the least sensitive HRU groups within a defined parameter range from calibration could potentially minimize computational costs while preserving physically realistic spatial patterns of LSM fluxes and states at field-scale resolutions, mitigating artifacts introduced by conventional methods (e.g. constant parameter multiplier over subbasins).

How to cite: Bacelar, L., Liu, H., Tang, G., Vergopolan, N., Mizukami, N., Wood, A., and Chaney, N.: How can spatial parameter sensitivity analysis enhance streamflow calibration routines in hyper-resolution hydrological models?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13626, https://doi.org/10.5194/egusphere-egu24-13626, 2024.

16:35–16:45
|
EGU24-12287
|
ECS
|
Virtual presentation
Chiru Naik Devavat and Dhanya Chandrika Thulaseedharan

Land surface processes exert a significant impact on local, regional, and global climate through intricate physical exchanges, including energy, water cycle dynamics, vegetation response, soil moisture variations, and heat fluxes between land and atmosphere. A comprehensive understanding of these processes necessitates the analysis of land surface states (e.g., soil moisture, temperature) and fluxes (e.g., evapotranspiration, runoff) over an extended period for various research fields such as hydrological process modeling, weather and climate forecast, drought/flood monitoring, and water resource conservation. However, the accuracy of analyses is hindered by the sparse and uneven distribution of in-situ measurements. To overcome this limitation, satellite-based data and land surface models are employed. While satellites provide continuous global data, they only capture surface-level conditions and have limited daily spatial coverage. Daily, multi-depth soil profile information is essential for understanding land condition dynamics and their impact on the water cycle and agriculture. The Community Land Model (CLM), specifically its latest version CLM5, stands as a pivotal tool for simulating biophysical and biogeochemical processes, including interactions with the atmosphere. Nevertheless, its efficacy in accurately simulating water and energy cycles over India, where local land surface changes are particularly pertinent due to sparse in-situ data remains to be evaluated. To address this gap, our study employs CLM5 to simulate the land surface process at a 0.1° resolution from 1980 to 2020 over India. The evaluation process is comprehensive, involving comparisons with diverse land surface datasets, encompassing in-situ, remotely sensed, and reanalysis measurements. For soil moisture, CLM5 demonstrates good agreement with in-situ data (correlation: 0.66 to 0.67) but exhibits wet biases when compared to in-situ and GLEAM. In the case of evapotranspiration and runoff, CLM5SP closely matches the patterns observed in GLEAM and GRUN datasets (correlation: 0.89 to 0.95 for evapotranspiration and 0.77 to 0.96 for runoff). However, it is noteworthy that CLM5SP tends to overestimate both evapotranspiration and runoff when compared to the reference datasets. The anticipated outcome of this study provides valuable insights into the capabilities of CLM5 simulations over India, offering applications and references for enhancing the model's characterization of water and energy fluxes in the future.

How to cite: Devavat, C. N. and Chandrika Thulaseedharan, D.: How well does CLM5 simulate water and energy cycles over India? - A performance evaluation , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12287, https://doi.org/10.5194/egusphere-egu24-12287, 2024.

16:45–16:55
|
EGU24-8347
|
ECS
|
On-site presentation
Awad M. Ali, Ruben O. Imhoff, and Albrecht H. Weerts

Recent advances in the application of machine learning techniques to estimate soil hydraulic properties using soil datasets have shown promising results. PedoTransfer Functions (PTFs) can facilitate the mapping of the complex relationship between soil properties and soil hydraulic properties, e.g., lateral hydraulic conductivity—a necessity for estimating lateral subsurface flow in spatially distributed hydrological models like wflow_sbm. The vertical-to-horizontal saturated hydraulic conductivity ratio (fKh0) is crucial for model calibration, but an established PTF is currently lacking. Our objective is to investigate the potential of ML algorithms in estimating PTFs for fKh0 prediction. First, optimized fKh0 across Great Britain (GB) resulting from a sensitivity analysis of the wflow_sbm model (Weerts et al., 2024) were used to train two ML algorithms; Random Forest (RF) and Boosted Regression Trees (BRT), using seven soil parameters from SoilGrids v1.0. Both algorithms effectively predicted fKh0 of 92 subbasins (i.e., test set of 25%) with high performance as compared against the optimized values, and RF slightly outperformed BRT. As a next step, we compared wflow_sbm simulated discharge results using uncalibrated fKh0 (default value of 100) and predicted values. The predictions notably improved wflow_sbm predictive accuracy by rising the median KGE from 0.55 (using uncalibrated fKh0) to 0.75 (using predicted fKh0). Following, we generated two globally distributed fKh0 maps, allowing us to further investigate the transferability of the ML-based PTFs. Therefore, we tested the predicted fKh0 across 559 gauge stations within the Loire basin in France. The utilization of either RF or BRT improved performance in around 75% of these subbasins with a KGE that was, on average, 0.06 higher. Furthermore, fKh0 prediction uncertainty and the impact of model spatial resolution were further analyzed. In conclusion, our study demonstrates the potential of ML methods to find relationships between soil properties and (model) soil hydraulic properties, which assists in parameter estimates for distributed hydrological models in gauged and ungauged basins.

How to cite: Ali, A. M., Imhoff, R. O., and Weerts, A. H.: Machine learning for predicting spatially variable lateral hydraulic conductivity: a step towards efficient hydrological model calibration and global applicability, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8347, https://doi.org/10.5194/egusphere-egu24-8347, 2024.

16:55–17:05
|
EGU24-13129
|
On-site presentation
Stephan Thober, Juliane Mai, Cinzia Mazzetti, Gianpaolo Balsamo, Christel Prudhomme, Robert Schweppe, Matthias Kelbling, Sebastian Müller, and Luis Samaniego

Accurately and efficiently estimating parameters for spatially distributed environmental models is impossible without proper regularization of the parameter space. The Multiscale Parameter Regionalization (MPR, Samaniego et al. 2010) makes use of high-resolution physiographic data (i.e., physiographic data such as soil maps and land cover information) to translate local land surface properties into model parameters. MPR consists of two steps: first, the high-resolution model parameters are derived from physiographic data via transfer functions at the native resolution. Second, the model parameters are upscaled to the target resolution the environmental model is applied on. MPR has already been successfully applied to the mesoscale hydrologic model (mHM, Samaniego et al. 2010, Kumar et al. 2013). An agnostic, stand-alone version implementation of MPR (Schweppe et al., 2022) allows applying this technique to any land-surface model or hydrological model.

In this study, we apply MPR to optimize parameters for the land-surface model ECLand (Boussetta et al. 2021) of the ECMWF Integrated Forecasting System. ECLand is calibrated at multiple locations simultaneously to provide an improved representation of river discharge at a global scale. We demonstrate the flexibility of the MPR approach by optimizing different transfer functions including the default one used in ECLand. In particular, we will discuss how specific choices in the calibration setting (i.e., chosen model parameters and ranges, basin locations, transfer function) affect the obtained ECLand model performance.

 

References:

Samaniego L., Kumar, R., and Attinger, S.: “Multiscale parameter regionalization of a grid-based hydrologic model at the mesoscale”, Water Resour. Res., 46, 2010.

Kumar, R., Samaniego, L., and Attinger, S.: “Implications of distributed hydrologic model parameterization on water fluxes at multiple scales and locations”, Water Resources Res, 2013

Schweppe, R., Thober, S., Müller, S., Kelbling, M., Kumar, R., Attinger, S., and Samaniego, L.: MPR 1.0: a stand-alone multiscale parameter regionalization tool for improved parameter estimation of land surface models, Geosci. Model Dev., 15, 859–882, https://doi.org/10.5194/gmd-15-859-2022, 2022

Boussetta S, Balsamo G, Arduini G, Dutra E, McNorton J, Choulga M, Agustí-Panareda A, Beljaars A, Wedi N, Munõz-Sabater J, de Rosnay P, Sandu I, Hadade I, Carver G, Mazzetti C, Prudhomme C, Yamazaki D, Zsoter E. ECLand: The ECMWF Land Surface Modelling System. Atmosphere. 2021; 12(6):723. https://doi.org/10.3390/atmos12060723

How to cite: Thober, S., Mai, J., Mazzetti, C., Balsamo, G., Prudhomme, C., Schweppe, R., Kelbling, M., Müller, S., and Samaniego, L.: Multi-basin calibration of the ECMWF land-surface model ECLand, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13129, https://doi.org/10.5194/egusphere-egu24-13129, 2024.

17:05–17:10
17:10–17:20
|
EGU24-2600
|
ECS
|
Virtual presentation
Chen Yang, Laura Condon, and Reed Maxwell

ParFlow CONCN model is the first integrated groundwater-surface water modeling platform over the mainland of China with a resolution of 30-arcsec and a depth of 492 m. With the flexibility of reconstruction and prediction of groundwater and surface water states and fluxes, it is undoubtedly an efficient tool for scientific understanding on water cycle and decision making on water resources and thus to tackle China’s water crisis in the changing world. Nonetheless, the CONCN model may have broader significances to the hydrologic community. Model evaluation and comparison is a common practice in the geoscientific modeling communities, such as those of land surface and earth system models. Due to the challenges in aquifer parameterization and the expensive computing requirement, high-resolution, large-scale, 3D groundwater modeling or integrated hydrologic modeling with 3D groundwater component is still under development though becoming more active in the past decade. Several national-scale such integrated hydrologic models, for example, ParFlow models over CONUS, west Africa, and Germany, have been built at 1 km resolution or higher with satisfying performances. However, the wide extension of modeling in this category to other places worldwide or to global scale is limitedly explored, preventing the evaluation of modeling workflows at different places and comparison with models using other parameterization schemes. Here, we demonstrate the construction and the first-phase evaluation of CONCN model by leveraging global datasets. Global permeability (GLHYMPS 1.0) and hydrography (MERIT Hydro) products were helpful to build the model while global water table depth (Fan et al., Science, 2013 and Zeng et al., JAMES, 2018) and streamflow (GRADES-HYDRDL and CNRD v1.0) products were adopted to preliminarily evaluate the simulation results. In this data-poor modeling area, both the construction and evaluation of the CONCN model are impossible about five years earlier as most of these global datasets did not exist. Therefore, the CONCN model can be one of the pioneers to evaluate and then to improve the current workflow of the existing models and address the challenges in new modeling areas with hydrogeology, hydrography, and climatology unseen in existing models. We also expect our dilemma caused by lacking observations as many other modelers in China can push the data-sharing to constrain hydrologic models and to motivate the collaboration such as model intercomparison in the Chinese hydrologic modeling community, which are well developed in the global community.

How to cite: Yang, C., Condon, L., and Maxwell, R.: Building and evaluating the high-resolution, integrated groundwater-surface water ParFlow modeling platform of continental China (CONCN): leveraging global datasets in a data-poor region   , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2600, https://doi.org/10.5194/egusphere-egu24-2600, 2024.

17:20–17:30
|
EGU24-3343
|
On-site presentation
Huang Lina and Zhang Shupeng

To represent the physical processes at hillslope scales for hyper-resolution land surface modeling, we propose a hierarchical, catchment-based spatial tessellation method. The land surface is divided into a hierarchical structure: catchments, height bands along hillslopes within a catchment, and land cover patches within a height band. This catchment-based structure explicitly represents hillslope drainage networks and can be applied at various resolutions determined by a pre-defined maximum height band size. The proposed tessellation method is superior to the conventional grid-based structure in representing land surface heterogeneity, resulting in a higher aggregation skill through the height band representation. The spatial variations in air temperature, leaf area index, saturated soil hydraulic conductivity, and soil porosity are generally lower within a height band than those in a conventional rectangular grid, reflecting the nature of topographic control on climate, vegetation, and soil distribution. The improvement in aggregation skill depends on resolutions and terrain slope angle, more pronounced at 1/6° model resolution and over steeper terrains. Finally, we demonstrate that our proposed catchment-based structure performs better than the grid-based structure through modeling tests over the Columbia River basin at resolutions of 1/2°, 1/6°, and 1/20° and a global test at 1/2° using the ILAMB model evaluation metrics.

How to cite: Lina, H. and Shupeng, Z.: A Catchment-Based Hierarchical Spatial Tessellation Approach to a Better Representation of Land Heterogeneity for Hyper-Resolution Land Surface Modeling , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3343, https://doi.org/10.5194/egusphere-egu24-3343, 2024.

17:30–17:40
|
EGU24-12485
|
On-site presentation
Juliane Mai and Nandita B. Basu

Conducting highly standardized model intercomparison studies of hydrologic models across large scales is beneficial in various aspects such as improving model accuracy and robustness, informing decision making, addressing uncertainties, enhancing educational and outreach opportunities and facilitating model benchmarking among others. However, looking beyond streamflow for hydrologic models is required to ensure that models simulate the right results for the right reasons. Continental scale analyses provide further insights into which systematic limitations a model has.

In this study, seven models (GR4J-CemaNeige, HMETS, Blended-v1, Blended-v2, HBV-EC, HYPR, SAC-SMA) have been setup for more than 2500 watersheds across Canada and the US using the RAVEN modeling framework. The models are setup using a standardized set of meteorologic and geophysical datasets to inform the model regarding forcings, soil, landcover, and terrain. All models are calibrated with respect to daily streamflow (2001-2015) and are subsequently validated on an independent time period (1986-2000). Calibration was performed using 10 independent trials of the Dynamically Dimensioned Search algorithm each using a budget of 2000 model evaluations and Kling-Gupta Efficiency (KGE) as the objective function. Additional variables such as actual evapotranspiration (AET), surface soil moisture (SSM), and snow water equivalent (SWE) for the calibrated model setups were recorded and compared against independent gridded reference datasets (AET and SSM from GLEAM, SWE from ERA5-Land). 

The results (surprisingly) show that all tested models perform equally well for streamflow prediction (range of median KGE values across all sites during calibration period is [0.83, 0.87] and validation period is [0.46, 0.54]). 

Differences between models are most apparent for the auxiliary variables analyzed, i.e. AET, SSM, and SWE. The most interesting differences between the models lie in their abilities to predict AET, with median KGE being the highest for SAC-SMA (0.71), followed by GR4J-CemaNeige (0.65), while the lowest values were observed for HMETS (0.37) and HBC-EC (0.17). Indeed SAC-SMA showed highest performances across 51% of locations while the second-best model is GR4J-CemaNeige with best performance at 13% of locations. 

The SSM, evaluated using the Pearson correlation (r) coefficient, was predicted relatively well by all models (r ranging between 0.62 and 0.72); however, while most models had poorer predictions in the Rocky mountains and at higher latitudes, the SAC-SMA was definitely a better predictor of the temporal dynamics in SSM in these regions.

While the median performance for SWE prediction was relatively low across all models (median KGE between 0.23 and 0.40), poorer predictions mostly occurred in regions with low annual SWE, and predictions improved with increasing annual snow amounts. 

The study reveals novel insights regarding the consistent ability of a suite of models to predict streamflow, while clear ranking of models was apparent based on their ability to simulate spatially distributed variables like AET. Such differences likely arise due to model equifinality highlighting the value of model evaluation against multiple spatially distributed and lumped metrics, generating the correct streamflow for the right reasons.

How to cite: Mai, J. and Basu, N. B.: Beyond streamflow predictions: A continental scale hydrologic model intercomparison experiment, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12485, https://doi.org/10.5194/egusphere-egu24-12485, 2024.

17:40–17:50
|
EGU24-9082
|
On-site presentation
Jonas Olsson, Yiheng Du, Kristina Isberg, Johan Strömqvist, and Yeshewatesfa Hundecha

The calibration and validation of hydrological models often involve a suite of established statistical metrics, which may not always match the needs of local stakeholders, thereby constraining the evaluative scope, particularly in the context of global climate services. This study introduces an alternative, complementary evaluation approach by formulating two types of application-based evaluation metrics (Du et al., 2024), representing model performance in terms of (i) temporal matching of the extreme quantiles and (ii) reproduction of the maximized split-sample difference in flow signatures. The introduced metrics are compared to conventional statistical metrics, at seven case study areas across the world, with three model settings representing different datasets and calibrations, generated from the global hydrological model World-Wide HYPE (WWH; Arheimer et al., 2020). The different performances found using application-based and conventional metrics, respectively, reveal their ability to uncover the models' capability in various aspects. Ultimately, the comprehensive analysis of conventional and application-based metrics allows us to delineate two scenarios for model application: generally applicable models, and conditionally applicable models. For example, in some areas the WWH model, when applied with global dataset and local calibration, is well capable of producing predictions for the timing of extreme quantiles and the relative difference in flow signatures, even though it may not excel according to conventional evaluation metrics. Consequently, this model can be classified as conditionally applicable, suitable for areas where local data is scarce, yet providing reliable information that can aid decision-makers in developing strategies for water resources management.

Arheimer, B., Pimentel, R., Isberg, K., Crochemore, L., Andersson, J.C.M., Hasan, A.,  Pineda, L. (2020). Global catchment modelling using World-Wide HYPE (WWH), open data, and stepwise parameter estimation. Hydrology and Earth System Sciences, 24, 535-559.

Du, Y., Olsson, J., Isberg, K., Strömqvist, J., Hundecha., Y., Silva, B.C., Rafee, S.A.A., Fragoso Jr., C.R., Beldring, S., Hansen, A., Uvo, C.B., Sörensen, J. (2024). Application-based evaluation of multi-basin hydrological models. Journal of Hydrology, under revision.

How to cite: Olsson, J., Du, Y., Isberg, K., Strömqvist, J., and Hundecha, Y.: Assessment of a global hydrological service by application-based metrics, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9082, https://doi.org/10.5194/egusphere-egu24-9082, 2024.

17:50–18:00

Posters on site: Fri, 19 Apr, 10:45–12:30 | Hall A

Display time: Fri, 19 Apr, 08:30–Fri, 19 Apr, 12:30
A.22
|
EGU24-9217
|
ECS
|
Diana Spieler and Niels Schütze

Discussions calling for more rigorous evaluation practices for hydrologic models have recently increased. In addition to the widely used aggregated objective functions, hydrologic signatures are becoming common evaluation metrics for testing the adequacy of hydrologic models for specific application purposes.

This work calibrates 7533 conceptual model structures using KGE as an objective function. These structures are evaluated based on their accuracy (KGE performance) and their adequacy. We defined adequacy as showing less than a +/- 50% percentage bias on inter- and intra-annual flow representation as well as on ten selected signatures. These signatures represent five aspects of the hydrological regime (magnitude, frequency, duration, rate of change, and timing). The large number of model structures, calibrated to the streamflow of 12 hydro-climatically differing MOPEX catchments, allows general insight into how well common conceptual model structures can represent observed hydrological behavior evaluated by signatures.

Results show that a large number of model structures perform accurately (high KGE performance) but almost none of these may be considered adequate (poor signature performance). In nine catchments not a single model can be considered adequate. In the remaining three catchments, only between 1 (0.1%) and 49 (0.7%) of all tested model structures are adequate according to all testing requirements. While inter-annual mean flow representation is typically represented well, the number of models able to represent intra-annual mean flow and/or individual signatures rapidly decreases.

This study presents overwhelming evidence that traditional single-objective function-based calibration is unlikely to return model structures that adequately represent complete hydrologic regimes. We therefore recommend that any model intercomparison or evaluation study needs to be constrained with additional data and/or evaluated by more meaningful metrics than traditional objective functions alone.

How to cite: Spieler, D. and Schütze, N.: Testing the Adequacy of 7533 KGE Calibrated Conceptual Model Structures, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9217, https://doi.org/10.5194/egusphere-egu24-9217, 2024.

A.23
|
EGU24-12492
|
Peter Wagener, Diana Spieler, and Niels Schütze

Hydrologists are generally aware that the choice of calibration metric will affect how their model reproduces catchment runoff. However, scientific literature mainly provides a theoretical or case study specific discussion of the topic and no general guidelines. This study thus aims to develop a broader picture by evaluating the influence of 8 different objective functions on the representation of 15 hydrologic signatures for 45 lumped conceptual models in 10 climatically diverse catchments.

The 10 selected catchments are a subset of the CARAVAN dataset (Kratzert et al. [2023]) chosen by using a k-means clustering algorithm based on climate indices (Willmott and Feddema [1992]). The 45 models are taken from the MARRMoT toolbox (Knoben et al. [2019], Trotter et al. [2022]) and only models performing over a specified benchmark are used for the analysis. The signatures that will be analysed represent different processes and aspects of the hydrological regime and the following 8 calibration metrics are investigated: KGE, NSE, log KGE, log NSE, NP-KGE (Pool et al. [2018], Split KGE (Fowler et al. [2018], SHE (Kiraz et al. [2023], DE (Schwemmle et al. [2021]).

Preliminary results show that the ability to reproduce specific signatures is clearly influenced by the chosen metric and therefore this choice should always be based on the specific goal of the prospective modelling study. Each metric has specific strengths and weaknesses that may be used to make a decision. However, the results vary based on climate conditions, the applied model structure and the investigated signature. It is therefore difficult to disentangle all interdependencies and develop more general guidelines with the limited catchment set used in this study. We speculate that very dominant processes shaping the general runoff generation in a catchment (such as snow melt) reduces the impact of the choice of calibration metric, and that more complex models typically are more consistent in process representation.

References:
Fowler et al. (2018): doi: 10.1029/2017WR022466
Gupta et al. (2009): doi: 10.1016/j.jhydrol.2009.08.003
Kiraz et al. (2023): doi: 10.1029/2023WR035321
Knoben et al. (2019): doi: 10.5194/gmd-12-2463-2019
Kratzert et al. (2023): doi: 10.1038/s41597-023-01975-w
Nash and Sutcliffe (1970): doi: 10.1016/0022-1694(70)90255-6
Pool et al. (2018): doi: 10.1080/02626667.2018.1552002
Schwemmle et al. (2021): doi: 10.5194/hess-25-2187-2021
Trotter et al. (2022): doi: 10.5194/gmd-15-6359-2022
Willmott and Feddema (1992): doi: 10.1111/j.0033-0124.1992.00084.x

Disclaimer: The first author conducted the presented research at TUD, now a PhD student at UofC

How to cite: Wagener, P., Spieler, D., and Schütze, N.: Metrics that Matter: Calibration Choices and Their Impact on Signature Representation in Conceptual Hydrological Models, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12492, https://doi.org/10.5194/egusphere-egu24-12492, 2024.

A.24
|
EGU24-8809
|
ECS
Svenja Hoffmeister and Erwin Zehe

We explore the potential of the empirical mode decomposition (EMD), a signal-processing method, to evaluate hydrological model simulations. Usually, hydrological models are assessed in the time domain observing and comparing residuals on a point-to-point basis. An additional model evaluation in the frequency domain might provide useful and complementary insights about the model’s capability to reproduce dynamic system behaviour. EMD separates a signal (e.g. a soil moisture time series) into fast and slow oscillations based on a sifting process, in which subtracting the signals moving average from itself reveals the highest frequency oscillation. This allows for instance to analyse phase shifts of different signature modes (e.g. daily fluctuations) in different depths and by that to make assumptions on soil hydraulic properties such as the conductivity. Naturally, a model will always miss high-frequency components of the “real” signal as measurement devices used as model input already act as a filter of such. However, the ability to capture the lower frequency remains interesting as they include relevant hydrological processes. Advantages of EMD over traditional methods like Fourier or wavelet transform are that no prior assumptions are needed and that it works well for nonlinear or non-stationary signals.

We test the EMD method on soil moisture and matric potential time series of observations and a process-based hydrological model extracted for the same site and compare the phase shifts and spectral components. We want to test whether metrics such as the RMSE of frequency spectra help to further compare and elucidate different signals. First results underpin the potential of including EMD as a tool to quantify models from a different perspective. We observe difference in observation and model frequencies of soil water time series and can related certain intrinsic modes to hydrological processes.

How to cite: Hoffmeister, S. and Zehe, E.: The potential of empirical mode decomposition to evaluate hydrological model simulations, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8809, https://doi.org/10.5194/egusphere-egu24-8809, 2024.

A.25
|
EGU24-14494
Caihong Hu, Chengshuai Liu, Chaojie Niu, and Qiying Yu

Abstract
Study region: Typical basin in humid areas in the Huaihe River 
Study focus: Accurate flood forecasting is essential for making timely decisions regarding flood control and disaster reduction. The theory of watershed runoff generation and convergence serves as a crucial foundation for flood forecasting, while the calculation of runoff is necessary to simulate flood discharge. Identifying watershed runoff generation mechanisms has been a challenging task, particularly under complex underlying surface conditions. To improve the accuracy of flood simulation, this study examines the underlying surface information in the watershed, such as particle composition and content, soil bulk density, geological slope, land use, and other spatial attributes, aiming to analyze the mechanisms of runoff generation. In the study of sub-watersheds, various combinations of runoff generation mechanisms are identified to determine the patterns of runoff. Subsequently, a semi-distributed hydrological model is developed, which incorporates both saturation-excess and infiltration-excess runoff, utilizing the information obtained from the underlying surface. The model is validated using rainfall-runoff data from 14 events at the Xiagushan watershed. 
New hydrological insights for the region: The analysis of the fundamental physical conditions of the underlying surface of the watershed revealed that 69.70% of the area is prone to saturation-excess runoff, with an additional 30.30% of the area being susceptible to infiltration-excess runoff. The model considers the spatial distribution of runoff patterns by incorporating complex underlying surface information and demonstrates high accuracy in simulating flood events (NSE= 0.87, Epeak = 12.08%, Wpeak = 13.16%, Tpeak = 0.14 hours, R2 = 0.90). The model is straightforward, practical, and exhibits promising potential in terms of timeliness and applicability, thus lending itself well to further application in other watersheds, contributing to the scientific foundation of flood warning and forecasting efforts.

How to cite: Hu, C., Liu, C., Niu, C., and Yu, Q.: Construction of a semi-distributed hydrological model considering the combination of saturation-excess and infiltration-excess runoff space under complex substratum, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14494, https://doi.org/10.5194/egusphere-egu24-14494, 2024.

A.26
|
EGU24-10306
Wenzhao Liu, Zhaotao Mu, and Changwu Cheng

Land-surface evapotranspiration (ET) is a major component of the hydrologic cycle. It is a very attractive approach to estimate land surface ET by means of complementary relationship (CR). After 60 years of continuous exploration, the CR has developed from linear relationships to the present nonlinear ones. There are usually four boundary conditions (BCs) for the nonlinear CR, among which the first-order one in completely wet environments (dy/dxx=1) has been a debatable issue, including both the difference in values of dy/dxx=1, and the divergence in definitions of the independent variable x. It has always been a problem how to consider the advection effect in CR. The effect degree of advection from outside the region varies in the ET process at different spatial scales. In this paper, x denotes the ratio of equilibrium ET (ETe) to apparent potential ET (ETpa), y denotes the ratio of ET to ETpa, and x=1 is set as the benchmark with ETe as the lower limit of ETpa. According to the characteristics of ET processes at different spatial scales, we extend the value range of dy/dxx=1, and take dy/dxx=1=k (k≥0) to establish the generalized BC. The generalized CR model for ET is then proposed by using an exponential function, expressed as y=EXP[k/d(1-1/x^d)] (denoted by GCR-EXP; d>0), where k and d are model parameters. k is equal to 2 in the absence of advection, which is the most complementary case. When k < 2, warm advection plays a role, and the value of k gradually decreases as the advection influence increases. Brutsaert (2015) considered the effect of minimal advection, and used the potential ET (Priestley and Taylor,1972) as ET’s constraint to determine the first-order BC in completely wet environments for the polynomial model of CR, which is a case that fits quite well with a large number of observed data. When k = 0, the CR is no longer valid, and the ET is always equal to ETpa, which reflects the ET of a small wet surface. When 0≤x≤xmin, y is equal or approximately equal to 0. xmin and Priestley-Taylor coefficient α can be determined by the values of x at y close to 0 and to 1 in GCR-EXP model, respectively. For instance, the value of x at y=0.001 can be taken as the value of xmin. k reflects advection effects and the corresponding degrees of CR. Moreover, the GCR models, which satisfy the four BCs including dy/dxx=1=k, can be also expressed as a power-exponential function form and other ones besides the proposed exponential one (Supported by Project 41971049 of NSFC).

How to cite: Liu, W., Mu, Z., and Cheng, C.: Advection Effect and Boundary Condition in the Formulation of Generalized Complementary Relationship for Evapotranspiration, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10306, https://doi.org/10.5194/egusphere-egu24-10306, 2024.

A.27
|
EGU24-5237
Zi-Jun Hsu, Hong-Ru Lin, and Jet-Chau Wen

Past studies of hydrogeological parameters of aquifers have not using drawdown data from multiple sets of sequential pumping tests (SPT) at the same site to characterize the interaction of hydrogeological parameters (such as transmittance, T and storage coefficient, S) distribution field. Therefore, the purpose of this study was to use the same site (well site at the northeast corner of Yunlin University of Science and Technology, Douliu City, Yunlin County) collected for many years (20xx,20xx..year, five groups in total) of SPT drawdown data. First, the interaction between the drawdown water levels from the same observation well and five sets of pumping tests was analyzed. Afterwards, this study used the numerical method of (hydraulic tomography, HT) to analyze the leakage data of five groups of SPTs., reverse calculation the distribution of T and S of five groups of SPT and a spatial comparison was performed, comparing the interaction between 5 sets of T and S distribution fields, discuss in different time and space background, interaction of distribution field of local hydrogeological parameters.

How to cite: Hsu, Z.-J., Lin, H.-R., and Wen, J.-C.: Exploring the Reciprocity Behavior Distributions in Space for Hydrogeological Parameters, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5237, https://doi.org/10.5194/egusphere-egu24-5237, 2024.

A.28
|
EGU24-16210
Yannick Marcon and Emmanuel Dubois

[9:56 AM] Emmanuel Dubois

The project introduces the new R package, rechaRge, dedicated to open-source groundwater recharge (GWR) models. The goal is to facilitate the simulation of GWR estimates for researchers, professionals, and stakeholders, for both hydrogeologists and non-hydrogeologists, by providing all the tools for state-of-art modelling and the available GWR models in a single R package. The package includes functions for data preparation (utility functions), automatic calibration, sensitivity analysis, and uncertainty analysis, all integrated directly in the R environment. A first open-source GWR model, the HydroBudget model, is also incorporated in the package. The model’s excellent performance allowed for the simulation of large datasets of spatially distributed and transient GWR in several projects in Canada, ranging from small watershed scale (few km2) to regional scale (thousands of km2). Sensitivity analysis, calibration, and uncertainty for the models were greatly facilitated by the utility functions of the package. At the region scale, GWR was simulated within a global change context with a spatial resolution of a 500 m x 500 m and a monthly time step for up to 150 years and 24 scenarios. Moreover, the rechaRge package is a collaborative effort and developers of open-source GWR modelling codes are warmly invited to publish their models in this package and take advantage of the existing functions.

How to cite: Marcon, Y. and Dubois, E.: rechaRge – a package for integrated groundwater recharge modelling in R, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16210, https://doi.org/10.5194/egusphere-egu24-16210, 2024.

A.29
|
EGU24-16116
|
ECS
|
Tamara Leins, Nikolai Späth, Christian Reinhardt-Imjela, and Andreas Hartmann

Subsurface stormflow (SSF) is an important runoff-generation process, especially in humid, mountainous regions. It can play a major role in flood generation and contaminant transport at the catchment scale. However, as it is a subsurface and heterogeneous process, its monitoring can be very challenging. In turn, the identification of SSF parameters in hydrological models is a difficult task and is often affected by equifinality.  Our study uses the HBV-light model, a conceptual model at the catchment scale, to simulate SSF (and catchment discharge) at a catchment in the Ore Mountains in Saxony, Germany. To see whether it is possible to improve the identifiability of SSF parameters by looking at different flow conditions separately, we divide discharge data according to flow duration curve (FDC) percentiles. We then calibrate the conceptual model several times, each time using only discharge data within one percentile of the FDC. Using a Monte Carlo based calibration, we select the same number of behavioural parameter sets for every FDC percentile based on the Kling-Gupta-Efficiency as an objective function. With a regional sensitivity analysis as well as a GLUE uncertainty estimation, we analyse and compare the parameter sets and discharge simulations of the different percentile calibrations. In this way, we analyse whether there is more information content on SSF hidden in a specific part of the FDC and thus, SSF parameters and processes become better quantifiable.

How to cite: Leins, T., Späth, N., Reinhardt-Imjela, C., and Hartmann, A.: Can we parameterise Subsurface Stormflow in a conceptual simulation model using flow duration curve percentiles for calibration?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16116, https://doi.org/10.5194/egusphere-egu24-16116, 2024.

A.30
|
EGU24-528
|
ECS
Magali Ponds, Markus Hrachowitz, Marie-Claire ten Veldhuis, Gerrit Schoups, Harry Zekollari, Sarah Hanus, and Roland Kaitna

Hydrological models play a vital role in evaluating future changes in streamflow. Despite the strong awareness of non-stationarity in hydrological system characteristics, model parameters are typically assumed to be stationary and derived through calibration on past conditions. Integrating the dynamics of system change in these models remains challenging due to uncertainties surrounding changes in future climate and ecosystems.
Nevertheless, studies show that ecosystems evolve in response to prevailing climate conditions. There is increasing evidence that vegetation adjusts its root zone storage capacity – considered a critical parameter in hydrological models – to prevailing hydroclimatic conditions. This adaptation of the root zone to moisture deficits is central to the water balance method. When combined with long-term water budget estimates from the Budyko framework, the water balance method offers a promising approach to describe future climate-vegetation interactions within process-based hydrological models

Our study provides an exploratory analysis of the role of non-stationary hydrological model parameters for six catchments in the Austrian Alps. More specifically, we investigate future changes in the root zone storage and their consequent impact on modeled streamflow. Using the water balance method, we derive climate-based parameter estimates of the root zone storage capacity under historic and projected future climate conditions. These climate-based estimates are then implemented in our hydrological model to assess their consequent impact on modeled past and future streamflow.
Our findings show that climate-based parameter estimations significantly narrow the parameter ranges linked to root zone storage capacity. This stands in contrast to the broader ranges obtained solely through calibration. Moreover, using projections from 14 climate models, our findings indicate a substantial increase in the root zone storage capacity parameters across all catchments in the future, ranging from +10% to +100%. Despite these alterations, the model performance remains relatively consistent when evaluating past streamflow, independent of using calibrated or climate-based estimations for the root zone storage capacity parameter. Additionally, no significant differences are found when modeling future streamflow when including future climate-induced adaptation of the root zone storage capacity in the hydrological model. Variations in annual mean, maximum, and minimum flows remain within a 5% range, with slight increases found for monthly streamflow and runoff coefficients.

In summary, our research shows that although climate-induced changes in root zone storage capacity occur, they do not notably affect future streamflow projections in the Alpine catchments under study. This suggests that incorporating a dynamic representation of the root zone storage capacity parameter may not be crucial for modeling streamflow in humid and energy-limited catchments. However, our observations indicate relatively larger changes in root zone storage capacity within the less humid catchments studied, corresponding to higher variations in modeled future streamflow. This points to a potential higher significance of dynamically representing root zone characteristics in arid regions and underscores the necessity for further research in these areas.

How to cite: Ponds, M., Hrachowitz, M., ten Veldhuis, M.-C., Schoups, G., Zekollari, H., Hanus, S., and Kaitna, R.: The effects of future climate-induced adaptation of the root-zone storage capacity on modeled streamflow dynamics, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-528, https://doi.org/10.5194/egusphere-egu24-528, 2024.

A.31
|
EGU24-14583
|
ECS
Advancing Hydrological Modeling for Detailed Water Budget Assessment in Indian River Basins: A High-Resolution Approach
(withdrawn after no-show)
Anuj Prakash Kushwaha and Vimal Mishra
A.32
|
EGU24-6789
Dongryeol Ryu, Sri Priyanka Kommula, Bharat Lohani, and Stephan Winter

Spatially distributed runoff information is one of the most critical inputs to determining suitable locations of rainwater harvesting (RWH) structures. The majority of hydrological assessments for siting RWH structures rely on empirical formula, such as the Soil Conservation Service – Curve Number method that combines soil type, land covers, land use practices, surface conditions, and antecedent moisture conditions with a weak basis on hydrological processes. In addition, runoff generation by topography is considered separately through the computation of flow accumulation.  As a result, the current practice of determining suitable RWH locations is done using arbitrary scores rather than the actual spatiotemporal estimate of runoff.

The present study employs a topography-based hydrological model, TOPMODEL, to explicitly generate runoff for an experimental catchment of 1800 ha located in Haryana, India. The catchment has been subdivided into 102 sub-catchments where sub-catchment-scale runoff was calculated using daily forcing data of 40 years (1980 - 2020) with other static inputs such as soil and topography data.  For topography input, a 1-m resolution digital elevation model (DEM) collected by a Light Detection and Ranging (LiDAR) was used. The input variables of the model were calibrated using ground-based discharge values.

The daily sub-catchment-scale runoff from TOPMODEL was aggregated to monthly, seasonal, and annual time scales to produce more detailed picture of water availability for harvest over wet and dry seasons. Finally, the runoff was converted to grid-based values using the flow accumulation scheme widely used on GIS tools. The final grid-based map at 1-m resolution contains the runoff information across the entire catchment at monthly, seasonal and annual time scales. The improved spatio-temporal representation of runoff using TOPMODEL in combination with flow accumulation scheme offers enhanced assistance to designing RWH structures tailored by the actual water volume available at candidate locations and its seasonal and interannual variability.

How to cite: Ryu, D., Kommula, S. P., Lohani, B., and Winter, S.: High-resolution Spatial Mapping of Runoff Prediction for Micro-scale Surface Rainwater Harvesting, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6789, https://doi.org/10.5194/egusphere-egu24-6789, 2024.

A.33
|
EGU24-15445
|
ECS
Meseret Menberu, Juho Jakkila, Noora Veijalainen, Kristin Böttcher, Stefan Fronzek, Vesa Kolhinen, Paula Havu, Nasim Fazel, Miia Kumpumäki, Ari Koistinen, and Markus Huttunen

This study introduces the WSFS-P model, an evolution of the well-established national WSFS (Watershed Simulation and Forecasting System) hydrological model. This new model represents a significant shift, moving from a conceptual WSFS hydrological model framework to a more physically based, and process-oriented approach (WSFS-P). WSFS-P is a two-layer semi-distributed hydrological model developed at the Finnish Environment Institute (Syke) in order to offer more detailed physical representations in hydrological forecasting and research. This hydrological model incorporates a number of sub-models that cover a wide range of hydrologic processes, including precipitation, snow dynamics, evapotranspiration, lake evaporation, soil moisture, groundwater, river routing, and ice thickness. The model utilizes meteorological inputs such as precipitation, temperature, relative humidity, air pressure, net radiation, cloudiness, and wind speed to deliver a comprehensive and detailed simulation of the hydrological cycle. The WSFS-P aims to enhance the accuracy and effectiveness of hydrological forecasting and research in Finland by leveraging spatially distributed data, such as Corine land use, altitude, and Finnish soil database. This model covers the entire Finnish mainland and transboundary catchments but excludes islands and smaller coastal catchments. This study assesses the WSFS-P model in 58 different catchments in Finland that were selected to cover diverse hydrological characteristics, reliable data, and minimal influence from lake regulation. The selected catchments feature a variety of catchment sizes and topographical and land-use patterns, including forests and agricultural areas, and have varying soil types and distinct climatic conditions. Several catchments are characterized by numerous lakes typical to Finland. Additionally, the study provided a comprehensive examination of five specific catchments to highlight the model’s effectiveness. The preliminary results demonstrate the model’s capabilities in predicting water availability, contributing to efficient water resource management and enhanced flood and drought prediction in Finland. This study aims not only to introduce the WSFS-P model but also to validate its operational readiness for diverse hydrological conditions.

How to cite: Menberu, M., Jakkila, J., Veijalainen, N., Böttcher, K., Fronzek, S., Kolhinen, V., Havu, P., Fazel, N., Kumpumäki, M., Koistinen, A., and Huttunen, M.: Introducing WSFS-P, Process-based Version of the Watershed Simulation and Forecasting System (WSFS) in Finland, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15445, https://doi.org/10.5194/egusphere-egu24-15445, 2024.

A.34
|
EGU24-12273
|
ECS
Mohammad Ghoreishi, Shervan Gharari, Mohamed Elshamy, and Karl-Erich Lindenschmidt

Ice jams present a significant flood risk in communities located along northern rivers, especially during the breakup of ice cover, resulting in increased backwater levels and flooding beyond riverbanks. The accurate simulation of ice formation relies on precise streamflow data, a vital input for hydraulic models. This study aims to evaluate how different streamflow products influence ice formation, focusing on simulating ice-jam flooding of the Athabasca River at Fort McMurray, Canada, with the broader goal of assessing the suitability of global datasets for predicting such events at a local scale. In our investigation, we integrate MizuRoute, a river network routing tool, and RIVICE, a one-dimensional, hydrodynamic, and river-ice hydraulic model. By employing various large-scale runoff from different models and datasets, such as MESH, ERA5, and VIC among others, our goal is to comprehensively understand how each product impacts the formation of ice jams and the subsequent flooding events. The incorporation of these runoff products is particularly relevant to investigate utilizing global datasets for predicting ice-jam flooding at a local scale.

How to cite: Ghoreishi, M., Gharari, S., Elshamy, M., and Lindenschmidt, K.-E.: Exploring the Impact of Various Streamflow Products on Ice-Jam Formation , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12273, https://doi.org/10.5194/egusphere-egu24-12273, 2024.

A.35
|
EGU24-9701
Thomas de Fournas, Nathalie Folton, François Colleoni, and Killian Pujol--Nicolas

This study delves into the exploration of low-flow rivers, a crucial subject within the global river network. These rivers, characterized by reduced flows over significant periods, play an essential role in various ecosystems. They constitute a substantial portion of the global river network, spanning diverse regions, including arid, semi-arid, temperate, humid tropical, boreal, and alpine areas. The flow variations observed in these watercourses are influenced by multiple factors, including climate change and increased water withdrawals associated with human activities. Ungauged basins, where reliable flow data is not readily available, present a significant hurdle in hydrological modelling. The absence of direct measurements makes it difficult to understand and predict the flow dynamics of rivers and streams, particularly in regions with low flow watercourses. To overcome this challenge, the study leverages the SMASH platform (Spatially-distributed Modelling and ASsimilation for Hydrology), a versatile multi-model framework capable of handling the complexities associated with ungauged territories.

The model implemented within the SMASH platform draws inspiration from the GR model family, a collection of global and semi-distributed models developed over the past years at INRAE. SMASH is a flexible, spatially distributed hydrological modelling platform capable of operating at high spatial and temporal resolution in both gauged and ungauged catchments. It is designed to simulate flow hydrographs across all grid cells in the computational domain.

Additionally, it incorporates functionalities for parameter sensitivity analysis and methods for both uniform and spatially distributed parameter calibration with different objective functions.

The principal aim of this study is to test the performance of various hydrological model structures, inspired by the GR model on the SMASH platform in low flow context. The evaluation centers on calibration and validation processes, employing uniform calibration techniques and regionalization approaches over a comprehensive dataset spanning 40 years at a daily time step. This extensive evaluation aims to elucidate the efficacy of these models in reproducing the low flows, seasonnality and bilan of watercourses over a set of basins (100) covering France with differents hydrometeorologic catchments. Furthermore, the study introduces a novel dimension by leveraging an artificial neural network (ANN) to process catchment descriptors specific to France. The ANN facilitates the exploration of regionalization by establishing a meaningful correspondence between select catchment descriptors and model parameters.

The study will then conclude with a comprehensive comparison of all simulations, highlighting the best hydrological model structure and regionalisation.

How to cite: de Fournas, T., Folton, N., Colleoni, F., and Pujol--Nicolas, K.: Using a regionalized distributed hydrological modelling approach for prediction low flow on ungauged French territory based on the training of artificial neural network , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9701, https://doi.org/10.5194/egusphere-egu24-9701, 2024.

A.36
|
EGU24-15260
|
Joël Gailhard, Antonin Belin Mergy, Matthieu Le Lay, and Alexandre Devers

The adaptation to climate change of thermal power plants necessitates the identification and characterization of high impact hazards. Extremely low river flow is one of these situations. The estimation methods traditionally used today still rely on extreme value theory (i.e., statistical adjustment on few observations and/or simulations), but these methods suffer from numerous limitations. Recent developments now make it possible to consider another approach, based on hydro-climatic simulations: extreme low flow quantiles are estimated by coupling a climate generator and a hydrological model. A first proof of concept was recently tested on a single basin and showed significant potential (Parey et al. 2022). Areas for improvement were also identified, both on the climate generator and the hydrological model.

The purpose of this work was then (i) to extend this first proof of concept to a larger number of basins and (ii) to quantify the sensitivity of the simulation chain (i.e., extreme low flow quantiles estimation) to the parameters of the hydrological model (in our case, the MORDOR-SD daily lumped rainfall-runoff model, Garavaglia et al. 2017).

A dataset of 33 catchments, each of them being associated with at least one piezometer, was selected to investigate whether the MORDOR-SD model could be constrained by piezometric time series to improve low flow simulations. By performing calibrations using only streamflow information we first confirmed that a particular state of the model was well correlated with piezometry in most studied catchments (the level of the so-called « deep » store, dedicated to the baseflow component).

A multi-objective calibration approach was then implemented, optimizing both (i) flow simulation with 4 criterions focusing on different streamflow signatures and (ii) eventually one supplementary criterion base on the affine correspondence between the deep storage level of the model and piezometry (i.e., calibration with or without piezometric information).

The results led us to propose a classification of the 33 basins based on two indices. The first index characterizes the importance of the baseflow in the streamflow (BFI = baseflow index). The second index characterizes the a priori representativity of the piezometric time series during low flows (Cor QMNA/ZMNA, index also used in Andreassian & Pelletier 2023).

For 14 out of the 33 basins (BFI > 0.7), piezometric information was almost neutral and did not lead to a significant improvement: the streamflow information was sufficient to constrain the low flow simulations. For 11 out of the 33 basins (Cor QMNA/ZMNA < 0.6 and BFI < 0.7), piezometric information was misleading and degraded the results: we assume that the piezometric information was not sufficiently relevant. Ultimately, only 8 out of the 33 basins (Cor QMNA/ZMNA > 0.6 and BFI < 0.7) emerged as interesting case studies. For these 8 watersheds, the piezometric information appears relevant to be included in the calibration process to derive a physics-based extrapolation of extremely low flow quantiles.

How to cite: Gailhard, J., Belin Mergy, A., Le Lay, M., and Devers, A.: Extreme low flow estimation: added value of piezometry to constrain the asymptotic behavior of a lumped rainfall-runoff model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15260, https://doi.org/10.5194/egusphere-egu24-15260, 2024.

A.37
|
EGU24-3356
|
ECS
Mathilde Puche, Magali Troin, Dennis Fox, and Paul Royer-Gaspard

Assessing the benefits of increasing the discretization level of semi-distributed hydrological models is of great importance for hydrological applications. The impact of spatial discretization on model performance is investigated with the use of the Soil and Water Assessment Tool (SWAT) model when applied on a Mediterranean watershed (Argens, France). This study aims to explore how the spatial discretization (number of sub-basins and of hydrological response units (HRUs)) affects the model’s performance at simulating daily streamflows, and if the choice of soil and land use input datasets modifies model accuracy. Low and moderate resolution soil (5 km and 250 m) and land use (400 m and 100 m) maps are considered. Four SWAT input sets are created, each corresponding to a different combination of land use and soil datasets. Each input set is used to build 17 configurations with an increasing number of sub-basins (4, 12, and 18) and HRUs (from 4 to 320). The 68 models (4 input sets x 17 configurations) are evaluated on the 2001-2021 period using the Kling-Gupta efficiency (KGE) metric. Results indicate no influence of the number of sub-basins on SWAT performance. However, increasing the number of HRUs leads to a significant performance decrease (from 0.13 to 0.26 of KGE loss), regardless of the number of sub-basins and input datasets. The SWAT model is found to be more sensitive to soil dataset than to land use dataset. Despite significant differences in hydrological soil groups between the two soil maps, no clear impact on the derived hydrological properties is observed, such as the curve number. The observed decline in SWAT performance with an increasing number of HRUs is attributed to the calibration process rather than the soil and land use input datasets. This study suggests that, when the calibration of the semi-distributed SWAT model is not performed at the finer spatial HRU level, an increase in the spatial discretization does not lead to an improvement of the overall model accuracy.  Therefore, minimizing the number of HRUs during the watershed subdivision is recommended for getting optimal simulations of streamflow while dealing with the computational efficiency of SWAT.

How to cite: Puche, M., Troin, M., Fox, D., and Royer-Gaspard, P.: Effects of watershed subdivision based on soil and land use inputs on SWAT performance in a coastal Mediterranean catchment, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3356, https://doi.org/10.5194/egusphere-egu24-3356, 2024.

A.38
|
EGU24-19742
|
ECS
José Manuel Rodríguez-Castellanos, Alejandro Sánchez-Gómez, Silvia Martínez-Pérez, and Eugenio Molina-Navarro

The Tagus River is the longest in the Iberian Peninsula and its basin is highly managed: it is the most populated and subject to the Tagus-Segura water transfer, which sends an annual average of 330 hm3 to southeast Spain. Its starting point is a hyper-reservoirs' system (Entrepeñas-Buendía-Bolarque) located in the basin´s headwaters sector. The total inflow to this reservoir system has decreased by 50% in the last decades, mostly as a consequence of the already noticeable impacts of climate change, and this situation will be further aggravated in the future. Thus, both gaining knowledge about the hydrological behaviour of the Tagus River headwaters and developing reliable tools to predict inflows to this reservoirs' system are highly relevant tasks to aid for a sustainable water resources management in coming years.

In this work, we set up a highly detailed catchment model with SWAT+ in the Tagus River headwaters. Before calibration, two additional tasks were addressed: 1) HRUs were grouped into three classes with varying lithology and permeability (carbonate, detrital of high and medium permeability, detrital of low permeability), and 2) two hydrological indices, the runoff rate and the baseflow index, were obtained for eight subbasins that have streamflow records. Three sets of parameters were designed, one for each HRU geological class, and then a complex calibration procedure was addressed. A soft calibration, narrowing parameters´ ranges to reproduce the hydrological indices realistically, was followed by a multi-site hard calibration of the streamflow in eight subbasins. During hard calibration, the streamflow simulation performance and the accuracy of the model reproducing the runoff coefficient and the groundwater contribution were considered. Afterwards, a best simulation was chosen and tested with an initial validation of the daily streamflow produced in each reservoir watershed, obtaining both statistically and graphically satisfactory results. After some final modifications in the model, a second and final validation on the monthly inflows into the hyper-reservoirs system was done, successfully reproducing the observed records, with NSE, R2 and PBIAS values of 0.86, 0.88, 2.5% in Entrepeñas and 0.89, 0.91 and -8.5% in Buendía.

The SWAT+ calibration approach followed in this work is novel because it takes into account the heterogeneity in the hydrogeological properties of a catchment to parameterize it, optimizing the parameters values at a geological class level and evaluating the model performance at subbasin level, which implies a higher spatial calibration resolution that the usual one in SWAT studies. As a result, the model not only simulates accurately subbasins with uniform geological properties,  but the entire Tagus headwaters,  including those subbasins  with  mixed  geology, thus resulting  in a model that accurately simulates reservoir inflows. By performing a multi-site calibration on smaller subbasins, results are more accurate, and the model represents more realistically local hydrological conditions. For that reason, this methodology helps also to understand how specific environmental conditions might affect all hydrological model process, thus also helping in water management decision making.

How to cite: Rodríguez-Castellanos, J. M., Sánchez-Gómez, A., Martínez-Pérez, S., and Molina-Navarro, E.: Predicting reservoir inflows with an advanced SWAT+ model calibration in the Tagus River headwaters (Spain), EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19742, https://doi.org/10.5194/egusphere-egu24-19742, 2024.