Simulation Of Hydrological Processes Research Articles

Groundwater modelling in SWAT+: Considerations for a realistic baseflow simulation

Hydrological models are valuable tools that support the evaluation and management of water resources. Among catchment scale models, SWAT has been widely used, and a revision of this model, SWAT+, was recently released. Groundwater simulation has been pointed out as one of the main weaknesses of these models, and despite the introduction of some improvements in the new version, an accurate simulation of groundwater flow is still a pending task. One of the most common issues is the inability of the model to maintain baseflow during long dry periods. Baseflow is common in aquifer areas and has great relevance for water quality and ecosystems. Although some authors proposed different solutions to its inaccurate simulation (e.g., editing aquifers configuration, coupling with other models, etc.) this work aimed to determine if a realistic groundwater simulation is possible using SWAT+ and determine the reasons why this is not being achieved. For this purpose, a groundwater dominated catchment was modelled, and different calibration approaches were carried out and compared, including strategies such as hard calibration, soft calibration, or sensitivity analysis. A comprehensive analysis of model outputs was crucial to achieve a model with a realistic simulation of groundwater contribution to the streamflow, both in amount and timing. The properly parameterization of some parameters (perco, latq_co, lat_ttime, and especially alpha) was key. Limitations of hard calibration were evidenced, such as the necessity of complementing this strategy with an evaluation of the simulation of hydrological processes. For those parameters governing streamflow components simulation, a detailed description of its functioning is included in this manuscript, together with some guidelines to achieve a statistically accurate model with a realistic groundwater flow simulation. Follow these recommendations will be useful for SWAT users, and will serve to create more reliable models.

Model calibration using hydropedological insights to improve the simulation of internal hydrological processes using SWAT+

AbstractSoils affect the distribution of hydrological processes by partitioning precipitation into different components of the water balance. Therefore, understanding soil‐water dynamics at a catchment scale remains imperative to future water resource management. In this study, the value of hydropedological insights was examined to calibrate a processes‐based model. Soil morphology was used as soft data to assist in the calibration of the Soil Water Assessment Tool (SWAT+) model at five different catchment scales (48, 56, 174, 674, and 2421 km2) in the Sabie River catchment, South Africa. The aim of this study was to calibrate the SWAT+ model to accurately simulate long‐term monthly streamflow predictions as well as to reflect internal soil hydrological processes using a procedure focusing on hydropedology as a calibration tool in a multigauge system. Results indicated that calibration improved streamflow predictions where R2 improved by 2%–8%. Nash‐Sutcliffe Efficiency (NSE) improved from negative correlations to values exceeding 0.5 at four of the five catchment scales compared to the uncalibrated model. Results confirm that soil mapping units can be calibrated individually within SWAT+ to improve the representation of hydrological processes. Particularly, the spatial linkage between hydropedology and hydrological processes, which is captured within the soil map of the catchment, can be adequately reflected within the model simulations after calibration. This research will lead to an improved understanding of hydropedology as soft data to improve hydrological modelling accuracy.

Coupling the remote sensing data-enhanced SWAT model with the bidirectional long short-term memory model to improve daily streamflow simulations

Global climate change has led to an increase in the frequency and scale of extreme weather events worldwide, and there is an urgent need to develop better-performing hydrological models to improve the accuracy of streamflow simulations and to facilitate water resource planning and management. The Soil and Water Assessment Tool (SWAT) has a notable physical foundation and is widely used in hydrological research. However, it uses a simplified vegetation growth model, introducing uncertainty into the simulation results. This study focused on improving the model based on remotely sensed phenological and leaf area index (LAI) data. Phenological data were used to define vegetation dormancy, and the LAI data replaced the corresponding data simulated by the original model. This approach improved the accuracy of the model in describing vegetation dynamics. Then, the enhanced SWAT model was coupled with the bidirectional long short-term memory (BiLSTM) model to validate the simulation of hydrological processes upstream of the Hei River. During model validation, the performance of the enhanced SWAT model in simulating streamflow (R2 = 0.835, NSE = 0.819) was better than that of the original SWAT model (R2 = 0.821, NSE = 0.805). In terms of simulating evapotranspiration, the enhanced SWAT model demonstrated even greater advantages. During the verification period, compared to those of the SWAT model, the R2 and NSE values of the enhanced SWAT model for daily-scale simulations increased from 0.196 and −0.269 to 0.777 and 0.732, respectively. The R2 and NSE values for monthly-scale simulations increased from 0.782 and 0.678 to 0.906 and 0.851, respectively. Simultaneously, the performance levels of two coupling approaches in streamflow prediction were compared, i.e., direct coupling of the original SWAT and BiLSTM models (SWAT-BiLSTM) and coupling of the enhanced SWAT and BiLSTM models (enhanced SWAT-BiLSTM). The results showed that the enhanced SWAT-BiLSTM model always performed better than the SWAT-BiLSTM model during the entire simulation period, especially the enhanced SWAT-BiLSTM model, which could more accurately predict peak streamflow changes. This study demonstrated that coupling an improved physical model with deep learning models could improve the streamflow prediction accuracy.

Estimation of Permafrost Ground Ice to 10 m Depth on the Qinghai‐Tibet Plateau

ABSTRACTPermafrost ground ice melting could alter hydrological processes in cold regions by releasing water. Currently, there is a lack of gridded data of ground ice from the Qinghai‐Tibet Plateau (QTP). Using 664 borehole sample records, we applied a random forest (RF) method to predict the ground ice content of permafrost between 2 and 10 m depth in three layers (2–3, 3–5, and 5–10 m) at a spatial resolution of 1 km. The RF predictions demonstrated an R2 value exceeding 0.80 for all three layers with a negligible positive overestimation (0.98%–1.85%). The ground ice content of the first layer (2–3 m) can be predicted primarily using climate variables, but the contribution of terrain and soil variables increases as the depth increases. The total water storage of ground ice across the QTP permafrost (2–10 m depth) is approximately 3330.0 km3, with 403.5 km3 in the 2–3 m layer, 857.2 km3 in the 3–5 m layer, and 2069.3 km3 in the 5–10 m layer. This study generated for the first time a gridded dataset of the shallow permafrost ground ice content across the entire QTP which can be used to improve simulations of hydrological processes in the permafrost regions.

Evaluation and Comparison of Reanalysis Data for Runoff Simulation in the Data-Scarce Watersheds of Alpine Regions

Reanalysis datasets provide a reliable reanalysis of climate input data for hydrological models in regions characterized by limited weather station coverage. In this paper, the accuracy of precipitation, the maximum and minimum temperatures of four reanalysis datasets, the China Meteorological Assimilation Driving Datasets for the SWAT model (CMADS), time-expanded climate forecast system reanalysis (CFSR+), the European Centre for Medium-Range Weather Forecast Reanalysis (ERA). and the China Meteorological Forcing Dataset (CMFD), were evaluated by using data from 28 ground-based observations (OBs) in the Source of the Yangtze and Yellow Rivers (SYYR) region and were used as input data for the SWAT model for runoff simulation and performance evaluation, respectively. And, finally, the CMADS was optimized using Integrated Calibrated Multi-Satellite Retrievals for Global Precipitation Measurement (AIMERG) data. The results show that CMFD is the most representative reanalysis data for precipitation characteristics in the SYYR region among the four reanalysis datasets evaluated in this paper, followed by ERA5 and CFSR, while CMADS performs satisfactorily for temperature simulations in this region, but underestimates precipitation. And we contend that the accuracy of runoff simulations is notably contingent upon the precision of daily precipitation within the reanalysis dataset. The runoff simulations in this region do not effectively capture the extreme runoff characteristics of the Yellow River and Yangtze River sources. The refinement of CMADS through the integration of AIMERG satellite precipitation data emerges as a potent strategy for enhancing the precision of runoff simulations. This research can provide a reference for selecting meteorological data products and optimization methods for hydrological process simulation in areas with few meteorological stations.

Advancing isotope‐enabled hydrological modelling for ungauged calibration of data‐scarce humid tropical catchments

AbstractRealistic projections of the future climate and how this translates to water availability is crucial for sustainable water resource management. However, data availability constrains the capacity to simulate streamflow and corresponding hydrological processes. Developing more robust hydrological models and methods that can circumvent the need for large amounts of hydro‐climatic data is crucial to support water‐related decisions, particularly in developing countries. In this study, we use natural isotope tracers in addition to hydro‐climate data within a newly developed version of the spatially‐distributed J2000iso as an isotope‐enabled rainfall‐runoff model simulating both water and stable isotope (δ2H) fluxes. We pilot the model for the humid tropical San Carlos catchment (2500 km2) in northeastern Costa Rica, which has limited time series, but spatially distributed data. The added benefit of simulating stable isotopes was assessed by comparing different amounts of observation data using three model calibration strategies (i) three streamflow gauges, (ii) three gauges with stream isotopes and (iii) isotopes only. The J2000iso achieved a streamflow Kling–Gupta efficiency (KGE) of 0.55–0.70 across all the models and gauges, but differences in hydrological process simulations emerged when including stable water isotopes in the rainfall‐runoff calibration. Hydrological process simulation varied between the standard J2000 rainfall‐runoff model with a high simulated surface runoff proportion of 37% as opposed to the isotope version with 84%–89% simulated baseflow or interflow. The model solutions that used only isotope data for calibration exhibited differences in simulated interflow, baseflow and model performance but captured bulk water balances with a reasonable match between the simulated and observed hydrographs. We conclude that J2000iso has shown the potential to support water balance modelling for ungauged catchments using stable isotope, satellite and global reanalysis data sets.

Quantifying the Impact of Human Activities on Hydrological Drought and Drought Propagation in China Using the PCR‐GLOBWB v2.0 Model

AbstractThe economic and human losses caused by drought are increasing, driven by climate change, human activities, and increased exposure of livelihood activities in water‐dependent sectors. Mitigation of these impacts for socio‐ecological securit is necessary to gain a better understanding of how human activities contribute to the propagation of drought as water management further develops. The previous studies investigated the impact of human activities on a macro level, but they overlooked the specific effects caused by human water management measures. In addition, most studies focus on the propagation time (PT, the number of months from meteorological drought propagation to hydrological drought), while other drought propagation characteristics, such as duration, magnitude, and recovery time, are not yet sufficiently understood. To tackle these issues, the PCR‐GLOBWB v2.0 hydrological model simulated hydrological processes in China under natural and human‐influenced scenarios. The study assessed how human activities impact hydrological drought and its propagation. Result shows that human activities have exacerbated hydrological drought in northern China, while it is mitigated in the south. The propagation rate (PR, proportion of meteorological drought propagation to hydrological drought) ranges from 45% to 75%, and the PT is 6–23 months. The PR does not differ substantially between the north and south, while the PT is longer in the north. The PR decreases by 1%–60% due to human activities, and the PT decreases (1–13 months) in the north and increases (1–10 months) in the south. Human activities display significant variations in how they influence the propagation process of drought across different basins. The primary factors driving the spatial pattern of drought disparities are regional variations in irrigation methods and the storage capacity of reservoirs.

A New Benchmark on Machine Learning Methodologies for Hydrological Processes Modelling: A Comprehensive Review for Limitations and Future Research Directions

The best practice of watershed management is through the understanding of the hydrological processes. As a matter of fact, hydrological processes are highly associated with stochastic, non-linear, and non-stationary phenomena. Hydrological processes simulation and modeling are challenging issues in the domains of hydrology, climate and environment. Hence, the development of machine learning (ML) models for solving those complex hydrological problems took essential place over the past couple decades. It can be observed, hydrological data availability has increased remarkably, and thus computational resources has led to a resurgence in ML models’ development. It has been witnessed huge efforts on the hydrological processes modeling using the facility of ML models and several review researches have been conducted. Literature studies approved the capacity of ML models in the field of hydrology over the classical “traditional models” based on their forecastability, flexibility, precision, generalization, and modeling execution convergence speed. However, although several potential merits were observed in ML model’s development, several limitations are allied such as the interpretability of those black-box models, the practicality of the ML models in watershed management, and difficulty to explain the physical hydrological processes. In this survey, an exhibition for all the published review articles on the development of ML models for hydrological processes and recognize all the research gaps and potential research direction. The ultimate aim of the current survey is to establish a new milestone for the interested hydrology, environment and climate researchers on the applications of ML models.

Multi-Scale Periodic Variations in Soil Moisture in the Desert Steppe Environment of Inner Mongolia, China

Uncovering the complex periodic variations in soil moisture can provide a significant reference for climate prediction and hydrological process simulation. We used wavelet analysis to quantify and identify the multi-scale periodic variations of soil moisture in the desert steppe of Inner Mongolia from 2009 to 2019 and analyzed the differences between the multi-scale periodic changes in soil moisture at the bottom (BS) and upper slope (US). The results show that the interannual variability in soil moisture at the BS and US has a significant upward trend. Moreover, the amount and volatility decrease with the increase in soil depth in the range of 0–30 cm, and the soil moisture at the BS is 26.4% higher than the US. The soil moisture has periodic changes of 0.5a, 1a, 2a, 3a and 3.5a in the desert steppe environment of Inner Mongolia. The periodic structure and intensity of different slope positions are greatly different. Soil moisture at the US has more complex multi-scale periodic changes, and the periodic oscillations of 3.5a, 3a, and 1a are dominant. The periodic oscillations of 0.5a and 1a are dominant at the BS. At the BS, the intensity of periodic oscillation of 1a after January 2015 has weakened. The weakening of soil moisture by temperature, rainfall and soil temperature caused the change in the multiple time-scale periodic oscillation of soil moisture.

LandBench 1.0: A benchmark dataset and evaluation metrics for data-driven land surface variables prediction

The advancements in deep learning methods have presented new opportunities and challenges for predicting land surface variables (LSVs) due to their similarity with computer sciences tasks. However, few researchers focus on the benchmark datasets for LSVs predictions that hampers fair comparisons of different data-driven deep learning models. Hence, we propose a LSVs benchmark dataset and prediction toolbox to boost research in data-driven LSVs modeling and improve the consistency of data-driven deep learning models for LSVs. LSVs benchmark dataset contains a large number of hydrology-related variables, such as global soil moisture, runoff, etc., which can verify the simulation of hydrological processes. Various global data from European Centre for Medium-Range Weather Forecasts reanalysis 5 (ERA5), ERA5-land, global gridded soil information (SoilGrid), soil moisture storage capacity (SMSC), and moderate-resolution imaging spectroradiometer (MODIS) datasets have been pre-processed into daily data at 0.5-, 1-, 2-, and 4-degree resolutions to facilitate their use in data-driven models. Simple statistical metrics, i.e., the root mean squared error and correlation coefficient, are chosen to evaluate the performance of different deep learning (DL) models, including convolutional neural network, long short-term memory and convolution long short-term memory models, with lead times of 1 and 5 days. A processed-based model serves as a physic baseline, soil moisture and surface sensible heat fluxes are taken as the target variables. The developed benchmark dataset and evaluation metrics for predicting LSVs using data-driven approaches, named as the LandBench toolbox, were implemented using Pytorch. This toolbox facilitates the reimplementation of existing methods, the development of novel predictive models, and the utilization of unified evaluation metrics. Additionally, the toolbox incorporates address mapping technology to enable high-resolution global predictions with constrained computing resources. We hope LandBench will not only serves as a standardized framework, fostering equitable model comparisons, but also provides indispensable data and a robust scientific foundation essential for advancing climate change research, disaster management, and sustainable development initiatives.

Improved thermal conductivity parameterization of SHAW model in permafrost regions on the Qinghai-Tibet Plateau

Soil thermal conductivity is critical for determining the hydrological and energy transfer processes in permafrost regions and simulating hydrological and thermal dynamic processes with land surface models. In this study, improved thermal conductivity parameterizations of the Simultaneous Heat and Water model (SHAW) were assessed using observation data from permafrost regions on the Qinghai-Tibet Plateau (QTP). The results indicated that the SHAW could precisely describe soil hydrological and thermal processes at the Tanggula (TGL) site. In general, the performances of Johansen's, Luo's, and the ORCHIDEE-MICT model's soil organic matter (SOM) parameterizations were superior to the original De Vries's parameterization. The SOM parameterization performed the best. The root mean square error (RMSE) of the active layer's soil temperature was 0.42 °C lower than that of the original parameterization at the Xidatan (XDT) site and 0.2 °C lower at the TGL site. In particular, the RMSE of the soil temperature at 3 m depth at TGL was 0.8 °C lower than that of the original parameterization. The average Nash-Sutcliffe Efficiency (NSE) for simulating the active layer soil temperature using the SOM parameterization was 0.915 at TGL site and 0.806 at XDT site. While De Vries parameterization yielded average NSE values of 0.771 and 0.627 for soil temperature simulations at the two sites, respectively. The results indicate that the soil organic carbon cannot be ignored in the simulation of hydrological and thermal processes in permafrost areas with a high soil organic carbon content. Notably, the simulated soil temperatures during the freezing period and the soil liquid water content during the thawing period improved significantly after modifying the thermal conductivity parameterization of SHAW. Therefore, the revised SHAW more accurately reflects the water and heat transfer processes of permafrost and can be used to analyze land-atmosphere interactions.

Hillslope groundwater and river channel precipitation recharge to rivers in a continuous permafrost catchment of northeastern Tibetan Plateau

Climate-change driven degradation of permafrost and changes in precipitation have resulted in significant changes to hydrological processes in permafrost areas. Previous studies on hillslope-stream connectivity and associated runoff-recharge to rivers have mainly focused on the threshold conditions and processes. In contrast, there has been limited study on the capacity of the permafrost active layer to recharge rivers and the relationships between river channel precipitation and river runoff, needed to predict flood events. This study aimed to characterize river runoff generation processes in the Yakou Catchment, northeastern Tibetan Plateau. Continuous monitoring of meteorological variables (precipitation and air temperature) and hillslope hydrological elements (thaw depths, supra-permafrost groundwater, and the thickness of the saturated zone) was conducted between June–August 2021–2022. The results showed using the thickness of the saturated zone (TSZ) to determine wet and dry conditions yielded significantly higher low flow (average of 0.153 m3 s−1) and lower low flow (average of 0.049 m3 s−1) with average TSZ depths of 0.40 m and 0.12 m under wet and dry conditions, respectively. However, no significant difference was noted in quick flow. Precipitation during typical rainfall events determined the generation of quick flow, with low flow constituting the main component of river runoff. The application of a partial least squares path model showed that TSZ on the permafrost determined the generation of river low flow which mainly originated from hillslope lateral subsurface flow. Conversely, river channel precipitation determined the generation of quick flow, which can contribute up to 80 % of the peak runoff during extreme rainfall events. Specifically, this study enhances the understanding of the connectivity between hillslopes and rivers and the storage-discharge relationship in permafrost catchments. This study provides a new theoretical reference for simulations of hydrological processes in the permafrost region.

Simulation of runoff process based on the 3-D river network

Many rivers originating from mountainous areas cannot be adequately measured in the field to obtain sufficient basic data due to the challenging terrain, which poses difficulties for conducting relevant research such as hydrodynamic and sediment transport simulations, fluvial material fluxes. This study utilized a combination of 71 river cross sections extracted from multisource remote sensing data and 34 cross sections measured by hydrological stations between 2010 and 2018 to generate generalized cross sections for all river orders in the middle reaches of the Yellow River. By integrating the generalized cross sections, river surfaces, and traditional DEM river network, a comprehensive 3-D river network has been established to provide a more accurate representation of the morphological changes along the river channel. To tackle the challenge of simulating runoff in data-deficient areas, the Digital Yellow River Integrated Model (DYRIM), a potent distributed hydrological model, was enhanced by incorporating geometric parameters derived from a 3-D river network into almost all input parameters that can be obtained from multi-source remote sensing. This enhancement enabled the model to accurately simulate runoff in representative data-deficient rivers of the Loess Plateau. The Nash Efficiency coefficients for the simulated runoff exhibited improvement of 23% compared to previous studies, surpassing 0.8 for the period between 2010 and 2018. Additionally, the simulated errors for flow velocity and water depth were 7% and 9% respectively. This approach can provide essential data for the simulation of hydrological processes and improving hydrological models, as well as technical methodologies for related research such as sediment transport simulations and river carbon emissions.

Simulation of Hydrological Processes in the Jing River Basin Based on the WEP Model

Inappropriate vegetation reconstruction in the Loess Plateau region has led to a significant increase in regional evapotranspiration and water consumption, further aggravating the shortage of soil water resources in the Loess Plateau region. The Jing River basin is a typical area for vegetation reconstruction in the Loess Plateau region. A thorough understanding of changes in hydrological processes in the Jing River basin is of significant scientific importance for efficient utilization of soil water resources and sustainable vegetation restoration in the region. In this study, the physically based Water and Energy Transfer Processes (WEP) distributed hydrological model was used to simulate key hydrological processes in the Jing River Basin during different periods before and after the implementation of cropland conversion to forest and grassland from 1980 to 2019. The results showed that after the implementation of cropland conversion to forest and grassland from 2000 to 2019, the average runoff volume in the Jing River Basin decreased by 20.91%. The most significant decrease in average runoff occurred in the central and northern parts of the basin, with a maximum reduction of 48.6%. The decrease in runoff in flood season is more obvious. The peak discharge decreased by 24.91%, and the most significant decrease occurred in the northern and central parts of the basin, ranging from 10.3% to 50.2%. The spatial distribution pattern of average soil moisture in the 0–0.8 m soil layer showed more moisture in the south and less in the north, with the minimum value occurring in certain areas in the eastern part of the basin. Overall, the implementation of cropland conversion to forest and grassland led to a certain degree of decrease in soil moisture in the basin. After the implementation of cropland conversion to forest and grassland, reference evapotranspiration fluctuated only in specific areas of the basin with no significant overall change.

Modeling fate and transport of pesticides from dryland agriculture using SWAT model

The aim of this study was to understand pesticide fate and transport from dryland agriculture in a major drinking water basin using SWAT and to identify critical source areas in the basin. Hydrological calibration results indicated satisfactory simulation of hydrologic processes within the catchment. Long term average observed sediment values (0.16 ton/ha) were compared with the annual average simulated SWAT outputs (0.22 ton/ha). Generally, the simulated concentrations were higher than the observed values, but the distribution pattern and trends were similar among the months. Average concentrations in water were 0.036 μg/L and 0.006 μg/L for fenpropimorph and chlorpyrifos, respectively. Transfer rates of pesticides from landscape to rivers showed that 0.36% of fenpropimorph and 0.19% of the applied amount of chlorpyrifos were exported to the river. Higher amount of fenpropimorph transport from land to the reach was attributed to its lower Koc (soil adsorption coefficient) value compared to chlorpyrifos. Higher amounts from HRUs were observed in the application month (April) and following month (May) for fenpropimorph, while the months after September showed higher amounts for chlorpyrifos. The specific HRUs (Hydrological Response Units) located in sub-basins 3, 5, 9 and 11 presented highest dissolved pesticide amounts, while HRUs in sub-basins 4 and 11 exhibited highest concentrations for adsorbed pesticides. Best management practices (BMPs) were recommended in critical subbasins to protect the watershed. Despite the limitations, the results demonstrate the potential contributions of modeling in terms of assessing pesticide loadings, critical zones and application timing.

Towards the development of an isotope‐enabled rainfall‐runoff model: Improving the ability to capture hydrological and anthropogenic change

AbstractGlobal hydrological alterations are being driven by climate change. However, while modelling tools have been instrumental in identifying these changes, their calibration is often dependent solely on streamflow data, which limits the ability to simulate important hydrological processes with the required level of certainty. Constraining hydrological process simulations is required to identify robust model parameters, reducing model uncertainty and providing a baseline model for climate‐related risk assessments. In this study, stable isotope measurements of rainwater, groundwater and stream water (δ2H and δ18O), together with an end member mixing analysis (EMMA) (as a post‐processing step) were integrated with the Jena adaptative modelling system (JAMS)/J2000 rainfall‐runoff model, named J2000iso. The J2000iso models of δ2H and δ18O were able to achieve a streamflow Nash–Sutcliffe efficiency (NSE) of 0.75 and 0.72 and an isotope mixing Kling–Gupta efficiency (KGE) of 0.51 and 0.60, respectively. Compared with the base version, the J2000iso had a much smaller variability with 56% less ensemble variance in the NSE, 13% more simulated interflow which resulted in a better soil‐moisture simulation at an observation point, and lower simulated flow component uncertainty. The J2000iso δ2H model was more robust than the base version during a subsequent validation in terms of KGE and Bias with 0.60 and 0.002 compared with 0.51 and −0.14 for the base version. The EMMA provided a means to improve the hydrograph separation ability of the JAMS/J2000. As many catchments around the world are still ungauged and are affected by anthropogenic activities, the development and enhancement of isotope‐enabled modelling provides a means to improve the simulation of key hydrological processes.

GEE can prominently reduce uncertainties from input data and parameters of the remote sensing-driven distributed hydrological model

The coupling of multisource remote sensing data and the lack of measured runoff introduce input data and model parameters uncertainties to the remote sensing-driven distributed hydrological model (RS-DHM). The PB satellite remote sensing datasets of the Google Earth Engine (GEE) are widely used in RS-DHM and remote sensing runoff inversion research, but whether GEE can reduce the two abovementioned uncertainties is still unknown. To answer this question, twelve remote sensing data sources provided by GEE were used in this study to drive a typical RS-DHM called the remote sensing-driven distributed time-variant gain model (RS-DTVGM) and the remote sensing runoff inversion technology called remote sensing hydrological station (RSHS), and the contribution of GEE to the improving hydrological model uncertainties was quantitatively analyzed from 2001 to 2020. The results showed that (1) the GEE-based improved data preparation not only effectively reduced the uncertainty in the input data with better spatial-temporal continuity and a 6.20 % reduction in the total area occupied by invalid grids, but also enhanced the operational efficiency by reducing the image number, memory size and data processing time of the satellite remote sensing data by 83.63 %, 99.53 %, and 98.73 %, respectively; (2) the GEE-based RSHS technology provided sufficient data support for parameter adjustment and accuracy validation of the RS-DTVGM, which effectively reduced the uncertainty in the model parameters and increased the Nash efficiency coefficient (NSE) in the calibration and validation period from 0.67 to 0.87 and 0.75, respectively; and (3) the calibrated RS-DTVGM was more reliable and robust, and its runoff and evapotranspiration were consistent with the actual statistical data. In the future, GEE and RSHS technology should be widely adopted to drive the RS-DHM to more quickly and easily provide reliable hydrological processes simulation results for integrated water resource management, therefore achieving win–win results in terms of efficiency and accuracy.

Reliability of satellite-derived precipitation data in driving hydrological simulations: A case study of the upper Huaihe River basin, China

• SPD are better at capturing temporal rather than spatial variations in precipitation. • SPD-driven models are typically better at characterizing temporal variations in HFS. • SPD-driven models exhibited large uncertainty in reproducing daily HFS. • Effective reproduction of HFS has more stringent requirements on the quality of SPD. • Model calibration based on streamflow may hinder the best reproduction of HFS. The vigorous development of satellite precipitation observations has gained widespread attention in the field of hydrology. Many studies have explored the reliability of hydrological models driven by satellite precipitation data (SPD) to reproduce streamflow. However, the reliability of SPD in the simulation of other hydrological processes remains unclear, even though a full understanding of the dynamic variations in hydrological processes is essential for the scientific management of water resources. In this study, we investigated the effect of SPD uncertainty on the distributed hydrological model in reproducing hydrological simulations. The investigation was conducted in the upper Huaihe River basin, which has a high-density ground precipitation observation network. The uncertainty inherent in 8 types of SPD in characterizing the temporal and spatial variations in precipitation was comprehensively examined. The Variable Infiltration Capacity hydrological model (VIC) was then applied to further explore the reliability of the simulated hydrological fluxes and states (HFS) driven by the selected SPD over multiple temporal-spatial scales. The results showed that: (1) the selected SPD tend to have greater uncertainty in recognizing spatial variations in precipitation than in identifying temporal variations; the VIC models driven by the selected SPD generally performed better in characterizing temporal variations in HFS than in reproducing their spatial variations. In addition, the corrected SPD (CSPD) are more suitable for HFS simulations than the uncorrected SPD (USPD); (2) the effective reproduction of the spatio-temporal variations in HFS has more stringent requirements on the quality of SPD than the satisfactory reproduction of streamflow. For the gridded scale simulations, the selected SPD-driven models exhibited large uncertainties in reproducing spatio-temporal variations in daily HFS. In contrast, although these models significantly improved the simulation performance of monthly HFS, they were still relatively deficient in reproducing their spatial variations. In terms of the basin-scale simulations, the CSPD-driven models tend to reliably reproduce monthly HFS; and (3) independent calibration can allow the selected SPD-driven models to reproduce streamflow more effectively, but it may not result in the best reproduction of HFS. This study provides guidance for strengthening the application of SPD in hydrological modeling and provides feedback for further improving the quality of SPD.

The impact of evaporation fractionation on the inverse estimation of soil hydraulic and isotope transport parameters

• Stable isotope data are used to estimate water flow and solute transport parameters; • Considering evaporation fractionation has significant impacts on parameter estimation; • Selecting suitable fractionation scenarios can effectively improve model performance; • The impact of evaporation fractionation is more important for arid conditions. Choosing a suitable process-oriented eco-hydrological model is essential for obtaining reliable simulations of hydrological processes. Determining soil hydraulic and solute transport parameters is another fundamental prerequisite. Research discussing the impact of considering evaporation fractionation on parameter estimation and practical applications of isotope transport models is limited. In this study, we analyzed parameter estimation results for two datasets for humid and arid conditions using the isotope transport model in HYDRUS-1D, in which we either did or did not consider fractionation. The global sensitivity analysis using the Morris and Sobol’ methods and the parameter estimation using the Particle Swarm Optimization algorithm highlight the significant impact of considering evaporation fractionation on inverse modeling. The Kling-Gupta efficiency (KGE) index for isotope data can increase by 0.09 and 1.49 for the humid and arid datasets, respectively, when selecting suitable fractionation scenarios. Differences in estimated parameters propagate into the results of two practical applications of stable isotope tracing: i ) the assessment of root water uptake (RWU) and drainage travel times (i.e., the time elapsed between water entering the soil profile as precipitation and leaving it as transpiration or drainage) in the lysimeter (humid conditions) and ii ) evaporation estimation in a controlled experimental soil column (arid conditions). The peak displacement method with optimized longitudinal dispersivity provides much lower travel times than those obtained using the particle tracking algorithm in HYDRUS-1D. Considering evaporation fractionation using the Craig-Gordon (CG) and Gonfiantini models is likely to result in estimates of older water ages for RWU than the no fractionation scenario. The isotope mass balance method that uses the isotopic composition profile simulated by HYDRUS-1D while considering fractionation using the CG and Gonfiantini models, or the measured evaporation isotope flux, provides comparable results in evaporation estimation as the HYDRUS-1D water mass balance method and direct laboratory measurements. In contrast, the no fractionation scenario reasonably estimates evaporation only when using the HYDRUS-1D water mass balance method. The direct use of simulated isotopic compositions in the no fractionation scenario may result in large biases in practical applications in the arid zone where evaporation fractionation is more extensive than in humid areas.

Prediction of Runoff in Watersheds Located within Data-Scarce Regions

The interest in the use of mathematical models for the simulation of hydrological processes has largely increased especially in the prediction of runoff. It is the subject of extreme research among engineers and hydrologists. This study attempts to develop a simple conceptual model that reflects the features of the arid environment where the availability of hydrological data is scarce. The model simulates an hourly streamflow hydrograph and the peak flow rate for any given storm. Hourly rainfall, potential evapotranspiration, and streamflow record are the significant input prerequisites for this model. The proposed model applied two (2) different hydrologic routing techniques: the time area curve method (wetted area of the catchment) and the Muskingum method (catchment main channel). The model was calibrated and analyzed based on the data collected from arid catchment in the center of Jordan. The model performance was evaluated via goodness of fit. The simulation of the proposed model fits both (a) observed and simulated streamflow and (b) observed and simulated peak flow rate. The model has the potential to be used for peak discharges’ prediction during a storm period. The modeling approach described in this study has to be tested in additional catchments with appropriate data length in order to attain reliable model parameters.