Development Of Hydrological Models Research Articles

Sensitivity analysis in the wavelet domain: a comparison study

Sensitivity analysis plays a pivotal role for the development and calibration of hydrological models, since they are often affected by equifinality. Despite a lot of effort has been placed for the development of effective sensitivity analysis methods, hydrological models remain over parametrized. We take advantage of the evidence that hydrological processes can be described as the superposition of effects occurring at different temporal scales (e.g., seasonal precipitation patterns, seasonal and daily snow and glacier melt, seasonal, daily and sub-daily water management operations) to develop a new framework to perform sensitivity analysis. We apply discrete and continuous wavelet transforms to disentangle hydrological signals occurring at different temporal scales and we take advantage of the different information stored at different temporal scales of the wavelet spectrum to perform a scale-dependent sensitivity analysis. This approach aims to increase the number of identifiable model parameters in comparison to standard sensitivity analysis performed in the time domain. As an exemplary problem, we apply the methodology to synthetic data describing surface water-groundwater interaction in rivers affected by hydropeaking (i.e., sudden fluctuations in the river stage due to hydropower production). The method could be applied also to other models displaying the superposition of processes with different intensities at different temporal scales such as ocean tide propagation in aquifers as well as snow and glacier melt models. The results indicate that considering multiple temporal scales allows us to increase the number of parameters that can be identified and hence calibrated with only a little increase in the computational effort.

Impact of land use/land cover changes on evapotranspiration and model accuracy using Google Earth engine and classification and regression tree modeling

This research uses a Classification and Regression Tree (CART) model with Google Earth Engine (GEE) to assess the winter season’s land cover and change detection mapping impact on the evapotranspiration (crop water requirement) parameters. Winter seasons, crucial for agricultural planning, and irrigation water requirement challenges in accurately mapping land cover and detecting changes due to the dynamic nature of farming practices during this period. In this study, Landsat-8 OLI images have been combined to map Land use and Land cover (LULC) and other change detection mapping in Akola Block, Maharashtra, India, during the 2018–2022 winter season. As an discoverer researcher that found detailed information of LULC classes during last 2018 to 2022 winter seasons, the use of the CART model in combination with a cloud-computing GEE demonstrates to be a practical approach for accurate land cover classification and change detection maps to create a pixel-based winter seasons information of study area. The novelty of this study lies in its innovative use of GEE, a powerful platform for remote sensing and geospatial analysis, to create LULC maps with remarkable accuracy. Achieving a 100% training accuracy across the four years under consideration is an exceptional feat, highlighting the reliability and stability of the methodology. Furthermore, the validation accuracy values, ranging from 89 to 94% for the winter seasons of 2018 to 2022, underscore the robustness of this approach. Such consistently high accuracy in mapping LULC over time is a groundbreaking achievement and offers a new dimension to the field of hydrology. For the hydrological community, the implications of this study are profound. Accurate LULC mapping and change detection provide critical data for modeling and analyzing the effects of land use changes on water resources, watershed management, and water quality. The User, Kappa, and Producer accuracy metrics used in this research highlight the model’s performance and its suitability for hydrological applications. These accurate LULC maps can aid in the development of hydrological models, forecasting, and decision-making processes, ultimately contributing to more effective water resource management and environmental conservation. In summary, this study’s innovative use of GEE, its remarkable accuracy in LULC mapping, and its relevance to the hydrological community demonstrate the potential for advanced remote sensing and geospatial tools to significantly improve our understanding of land use changes and their implications for water resources and environmental management.

Comparing the ability of different remotely sensed evapotranspiration products in enhancing hydrological model performance and reducing prediction uncertainty

The mitigation of uncertainties in the identification of natural systems is a fundamental aspect in the development of hydrological models, and represents a major challenge for the improvement of modelling techniques. In particular, the calibration of hydrological models based on streamflow measurements at the outlet of a catchment is exposed to significant sources of uncertainty, such as the impact of landscape features on runoff generation. Remote sensing-based actual evapotranspiration (AET) data can be incorporated with streamflow to improve model accuracy and reduce the uncertainty in hydrological modelling, resulting in a significant enhancement of the model performance. The selection of the right AET dataset for hydrological modelling is a crucial task, in front of the availability of multi-source datasets that differ in methods, parameters, and spatiotemporal resolution. Despite the existence of a few studies proposing the usage of remote sensing-based AET data, there is a lack of systematic comparisons between different products, in terms of performance for hydrological modelling. This paper aims to compare the efficacy of different remote sensing-based AET products in improving the simulation of hydrological responses, both in single and in multi-variable scenarios. In this investigation, the Soil and Water Assessment Tool (SWAT) hydrological model was calibrated with observed streamflow data by experimenting with eight different AET datasets. The findings of our study suggest that the incorporation of remote sensing-based AET data in the calibration process of a hydrological model can significantly enhance the accuracy and reliability of model predictions. Thus, the proposed approach can contribute to improving the effectiveness of hydrological modelling as a quantitative tool for the management of water resources. Another finding of this study is that the calibration of the model based solely on AET yields reasonable simulation results of the streamflow, which is an advantageous and promising feature for ungauged basins.

Development of a cellular automata-based distributed hydrological model for simulating urban surface runoff

The present study developed a distributed urban surface runoff model based on the cellular automata framework (CA-DUSRM) to simulate two-dimensional urban surface runoff processes. CA-DUSRM uses an improved nonlinear reservoir algorithm to simulate the generation of surface runoff outflow at grid cell scale and flow interactions among cells. The CA-DUSRM simulation results (total surface runoff and peak discharge) for an asphalt road sub-catchment during different rainfall events were in good agreement with actual measurements of both surface runoff and rainfall processes. The performance of CA-DUSRM was comprehensively and systematically analyzed under different rainfall–terrain scenarios. The results showed that increasing the temporal grain resulted in uncertain simulation results in relation to the rainfall intensity, slope, and terrain undulation employed; and significant spatial grain effects were mainly caused by large terrain undulation changes. Compared with the storm water management model (SWMM), CA-DUSRM performed better overall and was more sensitive to variation in terrain undulation and fluctuations in rainfall. Based on model structure and simulation mechanism perspectives, CA-DUSRM can present the influence of spatially heterogeneous underlying surface characteristics on the runoff processes and describe the lateral water exchanges, which makes the simulated processes closer to the actual processes when compared with SWMM. CA-DUSRM is potentially suitable for application in urban areas that have complex underlying surfaces.

Enhancing daily streamflow simulation using the coupled SWAT-BiLSTM approach for climate change impact assessment in Hai-River Basin

Against the backdrop of accelerated global climate change and urbanization, the frequency and severity of flood disasters have been increasing. In recent years, influenced by climate change, the Hai-River Basin (HRB) has experienced multiple large-scale flood disasters. During the widespread extraordinary flood event from July 28th to August 1st, 2023, eight rivers witnessed their largest floods on record. These events caused significant damage and impact on economic and social development. The development of hydrological models with better performance can help researchers understand the impacts of climate change, provide risk information on different disaster events within watersheds, support decision-makers in formulating adaptive measures, urban planning, and improve flood defense mechanisms to address the ever-changing climate environment. This study examines the potential for enhancing streamflow simulation accuracy in the HRB located in Northeast China by combining the physically-based hydrological model with the data-driven model. Three hybrid models, SWAT-D-BiLSTM, SWAT-C-BiLSTM and SWAT-C-BiLSTM with SinoLC-1, were constructed in this study, in which SWAT was used as a transfer function to simulate the base flow and quick flow generation process based on weather data and spatial features, and BiLSTM was used to directly predict the streamflow according to the base flow and quick flow. In the SWAT-C-BiLSTM model, SWAT parameters with P values less than 0.4 in each hydrological station-controlled watershed were calibrated, while the SWAT-D-BiLSTM model did not undergo calibration. Additionally, this study utilizes both 30 m resolution land use and land cover (LULC) map and the first 1 m resolution LULC map SinoLC-1 as input data for the models to explore the impact on streamflow simulation performance. Among five models, the NSE of SWAT-C-BiLSTM with SinoLC-1 reached 0.93 and the R2 reached 0.95 during the calibration period, and both of them stayed at 0.92 even in the validation period, while the NSE and R2 of the other four models were all below 0.90 in the validation period. The potential impact of climate change on streamflow in the HRB was evaluated by using predicted data from five global climate models from CMIP6 as input for the best-performing SWAT-C-BiLSTM with SinoLC-1. The results indicate that climate change exacerbates the uneven distribution of streamflow in the HRB, particularly during the concentrated heavy rainfall months of July and August. It is projected that the monthly streamflow in these two months will increase by 34% and 49% respectively in the middle of this century. Furthermore, it is expected that the annual streamflow will increase by 5.6% to 9.1% during the mid-century and by 6.7% to 9.3% by the end of the century. Both average streamflow and peak streamflow are likely to significantly increase, raising concerns about more frequent urban flooding in the capital economic region within the HRB.

Mapping global non-floodplain wetlands.

Non-floodplain wetlands - those located outside the floodplains - have emerged as integral components to watershed resilience, contributing hydrologic and biogeochemical functions affecting watershed-scale flooding extent, drought magnitude, and water-quality maintenance. However, the absence of a global dataset of non-floodplain wetlands limits their necessary incorporation into water quality and quantity management decisions and affects wetland-focused wildlife habitat conservation outcomes. We addressed this critical need by developing a publicly available "Global NFW" (Non-Floodplain Wetland) dataset, comprised of a global river-floodplain map at 90 m resolution coupled with a global ensemble wetland map incorporating multiple wetland-focused data layers. The floodplain, wetland, and non-floodplain wetland spatial data developed here were successfully validated within 21 large and heterogenous basins across the conterminous United States. We identified nearly 33 million potential non-floodplain wetlands with an estimated global extent of over 16×106 km2. Non-floodplain wetland pixels comprised 53% of globally identified wetland pixels, meaning the majority of the globe's wetlands likely occur external to river floodplains and coastal habitats. The identified global NFWs were typically small (median 0.039 km2), with a global median size ranging from 0.018-0.138 km2. This novel geospatial Global NFW static dataset advances wetland conservation and resource-management goals while providing a foundation for global non-floodplain wetland functional assessments, facilitating non-floodplain wetland inclusion in hydrological, biogeochemical, and biological model development. The data are freely available through the United States Environmental Protection Agency's Environmental Dataset Gateway (https://gaftp.epa.gov/EPADataCommons/ORD/Global_NonFloodplain_Wetlands/, last access: 24 May 2023) and through https://doi.org/10.23719/1528331 (Lane et al., 2023a).

Application of Boruta algorithms as a robust methodology for performance evaluation of CMIP6 general circulation models for hydro-climatic studies

Regional climate models are essential for climate change projections and hydrologic modelling studies, especially in watersheds that are overly sensitive to changes in climate. Accurate hydrologic model development is a daunting task in data-sparse regions where climate change’s impact on hydrologic and water quality processes is necessary for a well-informed policy decision on adaptation and hazard mitigation strategies. Novel approaches have been evolving that evaluated GCMs with the objective of improved parameterization to limit uncertainty and improve hydrologic model development. However, conclusions drawn should be purpose-driven based on intended usage. This study provides an overview of the state-of-the-art Boruta random forest as a robust methodology in the performance evaluation of GCMs models for hydroclimatic study. Highlights from the assessment indicate that (1) there is consistency in replicating the three observed climate variables of daily precipitation, maximum and minimum temperature respectively, (2) better temporal correlation (R2 = 0.95) in annual precipitation with a mean bias of 0.638mm/year, when compared to symmetrical uncertainty (SU) (R2 = 0.82), and all models ensembles (AME) (R2 = 0.88) with associated biases of 68.19mm/year and 10.57mm/year, respectively. Evaluation of the multi-year climate extreme indices, trends and magnitude reveal that there is a fair representation of basin-scale observed climate extreme events. However, the Boruta random forest approach exhibited a better statistical trend and magnitude of the extreme event in the basin. The findings of the study revealed enhanced GCM dataset evaluation and present a simple and efficient methodology to examine the limitations associated with the selected GCM ensemble for impact study in hydrology.

A Robust Strategy to Account for Data Sampling Variability in the Development of Hydrological Models

AbstractIt is typical to use a single portion of the available data to calibrate hydrological models, and the remainder for model evaluation. To minimize model‐bias, this partitioning must be performed so as to ensure distributional representativeness and mutual consistency. However, failure to account for data sampling variability (DSV) in the underlying Data Generating Process can weaken the model's generalization performance. While “K‐fold cross‐validation” can mitigate this problem, it is computationally inefficient since the calibration/evaluation operations must be repeated numerous times. This paper develops a general strategy for stochastic evolutionary parameter optimization (SEPO) that explicitly accounts for DSV when calibrating a model using any population‐based evolutionary optimization algorithm (EOA), such as Shuffled Complex Evolution (SCE). Inspired in part by the machine‐learning strategy of stochastic gradient descent (SGD), we use various representative random sub‐samples to drive the EOA toward the distribution of the model parameters. Unlike in SGD, derivative information is not required and hence SEPO can be applied to any hydrological model where such information is not readily available. To demonstrate the effectiveness of the proposed strategy, we implement it within the well‐known SCE, to calibrate the GR4J conceptual rainfall‐runoff model to 163 hydro‐climatically diverse catchments. Using only a single optimization run, our Stochastic SCE method converges to population‐based estimates of model parameter distributions (and corresponding simulation uncertainties), without compromising model performance during either calibration or evaluation. Further, it effectively reduces the need to perform independent evaluation tests of model performance under conditions that are represented by the available data.

Sediment load estimation using a novel regionalization sediment-response similarity method for ungauged catchments

The development and application of reliable hydrological models are challenging in ungauged or poorly gauged catchments. Although several regionalization methods have been developed for predicting hydrological variables in ungauged catchments, a large uncertainty usually exists in the prediction because previous methods typically depended on several static catchment attributes or long-term mean climatic descriptors, thereby neglecting the spatiotemporal variability of the climate. In this study, we propose a new regionalization method called sediment-response similarity (SRS) using the soil and water assessment tool model and self-organizing maps to estimate sediment load in ungauged catchments, considering spatiotemporal variability and the sediment load relation to rainfall characteristics of individual catchments. The performance of SRS was evaluated by comparing it against the conventional regionalization method (i.e., physical similarity) and calibration model results, in four gauged catchments of the lower Mekong River basin. The SRS method showed improved performance over the conventional approach in estimating sediment loads, with an average Nash-Sutcliffe efficiency (NSE) and coefficient of determination (R2) of 0.75 and 0.76, respectively, which were close to the respective values of NSE and R2 = 0.77, of the calibration model. Additionally, the SRS approach showed an error reduction of up to 7 %, compared with that of the physical similarity method, indicating the potential of the proposed regionalization method in estimating catchment-wide sediment load in ungauged catchments. The findings indicate that the SRS method ideally determines the similarity of sediment loads between gauged and ungauged catchments. The SRS approach effectively tackled the main challenge of selecting catchment attributes in the conventional regionalization method and minimizing the uncertainty of sediment prediction in ungauged catchments. This novel SRS method is a promising method for estimating, not only sediment load, but also other hydrological variables and rainfall-driven phenomena in the ungauged catchments globally.

Connecting hydrological modelling and forecasting from global to local scales: Perspectives from an international joint virtual workshop

AbstractThe unprecedented progress in ensemble hydro‐meteorological modelling and forecasting on a range of temporal and spatial scales, raises a variety of new challenges which formed the theme of the Joint Virtual Workshop, ‘Connecting global to local hydrological modelling and forecasting: challenges and scientific advances’. Held from 29 June to 1 July 2021, this workshop was co‐organised by the European Centre for Medium‐Range Weather Forecasts (ECMWF), the Copernicus Emergency Management (CEMS) and Climate Change (C3S) Services, the Hydrological Ensemble Prediction EXperiment (HEPEX), and the Global Flood Partnership (GFP). This article aims to summarise the state‐of‐the‐art presented at the workshop and provide an early career perspective. Recent advances in hydrological modelling and forecasting, reflections on the use of forecasts for decision‐making across scales, and means to minimise new barriers to communication in the virtual format are also discussed. Thematic foci of the workshop included hydrological model development and skill assessment, uncertainty communication, forecasts for early action, co‐production of services and incorporation of local knowledge, Earth observation, and data assimilation. Connecting hydrological services to societal needs and local decision‐making through effective communication, capacity‐building and co‐production was identified as critical. Multidisciplinary collaborations emerged as crucial to effectively bring newly developed tools to practice.

Application of Soil Moisture Data Assimilation in Flood Forecasting of Xun River in Hanjiang River Basin

Accurate projection of floods is of great significance to safeguard economic and social development as well as people’s life and property. The development of hydrological models can improve the level of flood projection, however, the numerous uncertainties in the models limit the projection accuracy. By adding observations to correct the operation of prediction models, the accuracy can be improved to some extent. In this paper, taking the Xun River, of the Hanjiang River Basin in China, as the research object, combined with the soil moisture satellite data obtained by the soil moisture active and passive satellite (SMAP), the ensemble Kalman filter (EnKF) algorithm was used to assimilate the upper soil water content (WU) of the Xinanjiang model. In addition, based on the simultaneous assimilation of state variables and parameters, two improved assimilation schemes were proposed here, namely, the augmented ensemble Kalman filter (AEnKF) scheme and the dual ensemble Kalman filter (DEnKF) scheme. The results showed that compared with the WU assimilation scheme, the simultaneous assimilation of parameters and WU improved the prediction ability of the Xinanjiang model to a greater extent. The two improved schemes had similar effects on flood prediction accuracy, and improved the overall Nash–Sutcliffe efficiency coefficient (NSE) from 0.725 for non-assimilated, and 0.758 for assimilated WU, to 0.781. Among them, AEnKF and DEnKF schemes, respectively, improved the NSE by 10.1% and 11% at maximum. This study demonstrated that the application of data assimilation for the Xun River effectively improved the flood forecast accuracy of the Xinanjiang model, which will provide a reference basis and technical support for future flood prevention and mitigation and flood projection in this basin.

Evaluation of surface mass-balance records using geodetic data and physically-based modelling, Place and Peyto glaciers, western Canada

AbstractReliable, long-term records of glacier mass change are invaluable to the glaciological and climate-change communities and used to assess the importance of glacier wastage on streamflow. Here we evaluate the in-situ observations of glacier mass change for Place (1982–2020) and Peyto glaciers (1983–2020) in western Canada. We use geodetic mass balance to calibrate a physically-based mass-balance model coupled with an ice dynamics routine. We find large discrepancies between the glaciological and geodetic records for the periods 1987–1993 (Place) and 2001–2006 (Peyto). Over the period of observations, the exclusion of ice dynamics in the model increased simulated cumulative mass change by ~10.6 (24%) and 7.1 (21%) m w.e. for Place and Peyto glacier, respectively. Cumulative mass loss using geodetic, modelled and glaciological approaches are respectively − 30.5 ± 4.5, − 32.0 ± 3.6, − 29.7 ± 3.6 m w.e. for Peyto Glacier (1982–2017) and − 45.9 ± 5.2, − 43.1 ± 3.1, − 38.4 ± 5.1 m w.e. for Place Glacier (1981–2019). Based on discrepancies noted in the mass-balance records for certain decades (e.g. 1990s), we caution the community if these data are to be used for hydrological model development.

Improved data splitting methods for data-driven hydrological model development based on a large number of catchment samples

• New methods are developed to optimally allocate calibration and evaluation data. • 754 different catchments are used to enable model performance comparison. • A guideline is developed to select data-splitting methods for hydrological models. • Results are relevant to other model types such as CRR models. Data-driven hydrological models are widely used for many practical purposes. However, the reliability of such models depend heavily on the strategy used to partition available observations into model calibration and evaluation subsets. Unfortunately, available data splitting methods are poor at ensuring consistency of statistical properties between different subsets, resulting in considerable bias and/or inconsistency in evaluation performance as well as poor generalization ability. To address this problem, we propose and test two new data splitting methods applied to hydrological models that do not consider time-dependent structure. The SOMPLEX approach uses a self-organizing map to analytically cluster the data based on its distributional properties, after which a portion of each cluster is allocated to the calibration and evaluation subsets using the previously developed DUPLEX method. In the MDUPLEX approach, rather than clustering the data, the DUPLEX allocation strategy is modified to better maintain statistical similarity of the data subsets. When tested using a data-driven rainfall-runoff modelling study applied to 754 catchments, the new methods were significantly better at splitting the data into subsets with similar statistical properties. However, performance of different methods was found to depend strongly on the skewness of the streamflow data, based on which we present practical recommendations regarding which method to use in different circumstances. Since the task of partitioning the data into mutually consistent statistically subsets is generic, these concepts are broadly applicable to many hydrological models where time-dependency of hydro-climatic data can be ignored, ranging from physics-based to data-driven, including machine learning.

Global hydroclimatic drivers of terrestrial water storage changes in different climates

The Gravity Recovery and Climate Experiment (GRACE) mission provides an unprecedented way to assess terrestrial water storage changes (TWSC) worldwide. However, hydroclimatic drivers of GRACE-derived TWSC in different climates have not been systematically examined, which hinders its applications in water resources management and hydrological model development. In this study, we derived the monthly TWSC using the GRACE dataset, and analyzed its partial correlations with precipitation (P), evapotranspiration (ET), runoff (R), and temperature (T) from 2003 to 2019 under the Köppen-Geiger world climate classification system. Relative contributions of the four hydroclimatic elements (P, ET, R, and T) to TWSC were quantified using the hierarchical partitioning method. The results indicate that P mainly controls TWSC in tropical climates, hot semi-arid climate, and temperate and continental climates with dry winter. ET is the primary driver of TWSC in mid- and high-latitude regions that feature temperate and continental climates with no dry season, subarctic climates, and polar climates. T mainly influences TWSC in cold arid climates, temperate and continental climates with dry summer or no dry season, and ice cap climate. However, TWSC in many arid climates near the Tropic of Cancer and the Tropic of Capricorn are not well explained by all four hydroclimatic variables according to their non-significant partial correlations. Large trends in annual TWSC are found in South America, western Canada, eastern contiguous United States, and Africa, which are likely due to marked changes in P based on their similar spatial pattern. In particular, decreasing TWSC are found in most continental climates, which manifests a deterioration of water resources availability in these regions. Overall, our results clarify the major hydroclimatic drivers of TWSC in different climates at a global scale, which may help improve TWSC modeling and prediction through a judicious selection of explanatory hydroclimatic variables in different climates.

A modified hydrologic model for examining the capability of global gridded PET products in improving hydrological simulation accuracy of surface runoff, streamflow and baseflow

• SWAT model was modified to examine the capability of PET products in improving hydrological simulation accuracy. • Distributed input and higher spatial resolution of PET products are more conducive to hydrological modeling. • The input of PET products has a great impact on streamflow and baseflow, followed by surface runoff. • The accuracy of PET products should be improved from both improving PET products and considering water budget closure. Terrestrial evapotranspiration (ET), which links water and energy exchanges between atmosphere, land surface and underground water, significantly affects hydrological modelling. However, accurate ET estimation is still a great challenge for hydrological models. Currently, most hydrological models obtain ET mainly from their built-in potential evapotranspiration (PET) modules due to lack of observations. Note that the ET obtained in this way cannot be verified by observations and inevitably brings large model uncertainty. Fortunately, several PET products have been produced based on satellite remote sensing and/or reanalysis climate model output with global coverage and high spatiotemporal resolution, providing a great opportunity for tunning hydrological models. However, the accuracy of global gridded PET products varies in different regions. It is still unknown whether PET products can improve the accuracy of hydrological simulation, and how different resolutions and input types (lumped/distributed) of PET products affect the performance of hydrological models. The question about how hydrological models will be further developed in combination with PET products is also not well answered, given the complexity in interaction between ET and other water cycle variables. In this study, a modified Soil and Water Assessment Tool (SWAT) was chosen to examine the capability of PET products in improving hydrological simulation in terms of surface runoff, streamflow and baseflow, together with the Shaying River Basin in China as a case study. Two Moderate Resolution Imaging Spectroradiometer (MODIS) PET products were employed to form various PET scenarios. Results indicate that PET products are significantly different from the PET simulated by the SWAT built-in PET module in terms of spatial distribution and magnitudes. According to statistical metrics, the use of PET products to replace the SWAT built-in PET module can improve the accuracy of hydrological simulation compared with the original SWAT using its built-in PET module. The distributed input and higher spatial resolution of PET products are more conducive to improve the accuracy of hydrological simulation. Our research shows that both improving the accuracy of PET products and balancing water budget closure are critical for accurate hydrological modeling. These conclusions obtained based on MODIS products were further verified against a hourly PET data product (hPET). This study provides insights into understanding the use of PET products in improving the accuracy of hydrological simulation. The study findings can provide a beneficial reference for the further development of hydrological models in combination with PET products.

Evaluation of daily gridded meteorological datasets for hydrological modeling in data-sparse basins of the largest lake in Southeast Asia

Study regionTonle Sap Lake (TSL) Basin in Cambodia. Study focusThe development and application of hydrological models for data-sparse basins are hindered by the limited hydro-meteorological data. Although gridded meteorological products are alternatively considered in many studies, the validation of the products with limited point observations overlooks the original spatiotemporal characteristics, thus leading to a selection of datasets with high uncertainty. Here, we evaluated seven gridded meteorological datasets of rainfall and air temperature covering the data-sparse Tonle Sap Lake Basin by employing the statistical approach based on the bilinear-interpolation method and hydrological approach using the SWAT model, which ensures the reliable estimates of streamflow and evapotranspiration. New hydrological insights for the regionThe results of the statistical approach indicate that APHRODITE, ERA5, TRMM and IMERG-based precipitation and CPC and SA-OBS-based air temperature performed comparably well (R ≥ 0.75) with the gauged data. However, ERA5-based streamflow performed relatively poor, while SWAT driven by APHRODITE underestimated satellite-based evapotranspiration, indicating the underestimation of basin-wide precipitation by APHRODITE. Although TRMM and IMERG provide more reliable estimation of streamflow and evapotranspiration, slightly better performance and a higher spatial resolution of IMERG dataset suggest that IMERG precipitation is superior for basin-wide hydrological modeling and impact assessment. These findings showed that statistical comparisons with gauge-data and hydrological evaluation of streamflow are not enough to justify the reliability of each gridded dataset.

Hydrological Models

This paper discusses the highly topical issue of forming and using hydrological models. Their significance in water resource management and achieving sustainable development goals is underlined, particularly in the context of climate change. Some shortcomings in hydrological process modelling are pointed out, ones that can significantly affect the probability of achieving reliable results. Given the important role water plays, it is necessary to understand the complexity of producing hydrological models. Indeed, hydrological models themselves play an important role, and they must as such meet several important assumptions. Above all, they must consistently follow the basic principles of hydrology. We find that contemporary modelling technology (computer use, rapid development of numerical methods, use of state-of-the-art monitoring techniques) has strongly surpassed the level of our knowledge regarding physical hydrological processes and their interactions with the living and non-living environments that we aim to model. The complexity of the procedures and the problem of transferring information from one dimension to another in space and/or time have been highlighted. The need for further development of hydrological models, involving strict verification of the results provided by these models, has been underlined.

Integrated Hydrologic-Hydrodynamic Inundation Modeling in a Groundwater Dependent Tropical Floodplain

The rapid development of free and open-access hydrological models and coupling framework tools continues to present more opportunities for coupled model development for improved assessment of floodplain hydrology. In this study, we set up an Upper Zambezi hydrological model and a fully spatially hydrological-hydrodynamic coupled model for the Barotse Floodplain using GLOFRIM (GLObally applicable computational FRamework for Integrated hydrological–hydrodynamic Modelling). The hydrological and hydrodynamic models used are WFLOW and LISFLOOD-FP, respectively. The simulated flows generated by the wflow model for the upstream gauge stations before the Barotse Floodplain were quite similar and closely matched the observed flow as indicated by the evaluation statistics; Chavuma, nse = 0.738; kge = 0.738; pbias = 2.561 and RSR = 0.511; Watopa, nse = 0.684; kge = 0.816; pbias = 10.577 and RSR = 0.557; and Lukulu, nse = 0.736; kge = 0.795; pbias = 10.437 and RSR = 0.509. However, even though the wflow hydrological model was able to simulate the upstream hydrology very well, the results at the floodplain outlet gauge stations did not quite match the observed monthly flows at Senanga gauge station as indicated by the evaluation statistics: nse = 0.132; kge = 0.509; pbias = 37.740 and RSR = 0.9233. This is mainly because the representation of both floodplain channel hydrodynamics and vertical hydrological processes is necessary to correctly capture floodplain dynamics. Thus, the need for an approach that saves as a basis for developing fully spatially distributed coupled hydrodynamic and hydraulic models’ assessments for groundwater dependent tropical floodplains such as the Barotse floodplain, in closing the gap between hydrology and hydrodynamics in floodplain assessments. A fully coupled model has the potential to be used in implementing adaptive wetland management strategies for water resources allocation, environmental flow (eflows), flood control, land use and climate change impact assessments. Doi: 10.28991/HEF-2022-03-02-09 Full Text: PDF

Integrated hydrologic model development and postprocessing for GSFLOW using pyGSFLOW

Integrated hydrologic model development and postprocessing for GSFLOW using pyGSFLOW

Uncertainty estimation with deep learning for rainfall–runoff modeling

Abstract. Deep learning is becoming an increasingly important way to produce accurate hydrological predictions across a wide range of spatial and temporal scales. Uncertainty estimations are critical for actionable hydrological prediction, and while standardized community benchmarks are becoming an increasingly important part of hydrological model development and research, similar tools for benchmarking uncertainty estimation are lacking. This contribution demonstrates that accurate uncertainty predictions can be obtained with deep learning. We establish an uncertainty estimation benchmarking procedure and present four deep learning baselines. Three baselines are based on mixture density networks, and one is based on Monte Carlo dropout. The results indicate that these approaches constitute strong baselines, especially the former ones. Additionally, we provide a post hoc model analysis to put forward some qualitative understanding of the resulting models. The analysis extends the notion of performance and shows that the model learns nuanced behaviors to account for different situations.