Search

Modeling streamflow is essential for understanding flow inundation. Traditionally, this involves hydrologic and numerical models. This research introduces a framework using agent-based modeling (ABM) combined with data-driven modeling (DDM) and Artificial Intelligence (AI). An agent-driven model simulates streamflow and its interactions with river courses and surroundings, considering hydrologic phenomena related to precipitation, water level, and discharge as well as channel and basin characteristics causing increased water levels in the Medio River. A five-year dataset of hourly precipitation, water level, and discharge measurements was used to simulate streamflow. The model's accuracy was evaluated using statistical metrics like correlation coefficient (r), coefficient of determination (R2), root mean squared error (RMSE), and percentage error in peak discharge (Qpk). The ABM's simulated peak discharge (Qpk) was compared with the measured peak discharge across four experimental scenarios. The best simulations occurred in scenario 3, using only rainfall and streamflow data. Data management and visualization facilitated input, output, and analysis. This study's ABM combined with DDM and AI offers a novel approach for simulating streamflow and predicting floods. Future studies could extend this framework to other river basins and incorporate advanced sensor data to enhance the accuracy and responsiveness of flood forecasting. [ABSTRACT FROM AUTHOR]

Remotely sensed evapotranspiration (ETRS) shows promise for enhancing hydrological models, especially in regions lacking in situ streamflow observations. However, model calibration studies showed conflicting results regarding the ability of ETRS products to improve streamflow simulation. Rather than relying on model calibration, here we produce the first randomized experiment that explores the full streamflow–ET skill distribution, and also the first probabilistic assessment of the value of different global ETRS products for streamflow simulation. Using 280,000 randomized SWAT (Soil and Water Assessment Tool) model runs across seven catchments and four ETRS products, we show that the relationship between ET and streamflow skills is complex, and simultaneous improvement in both skills is only possible in a limited range. Parameter sensitivity analysis indicates that the most sensitive parameters can have opposite contributions to ET and streamflow skills, leading to skill trade-offs. Conditional probability assessment reveals that models with good ET skills are likely to produce good streamflow skills, but not vice versa. We suggest that randomized experiments such as ours should be performed before model calibration to determine whether using ETRS is worthwhile, and to help in interpreting the calibration results. [ABSTRACT FROM AUTHOR]

Climate change has exacerbated water stress and water‐related disasters, necessitating more precise streamflow simulations. However, in the majority of global regions, a deficiency of streamflow data constitutes a significant constraint on modeling endeavors. Traditional distributed hydrological models and regionalization approaches have shown suboptimal performance. While current deep learning (DL)‐related models trained on large data sets excel in spatial generalization, the direct applicability of these models in certain regions with unique hydrological processes can be challenging due to the limited representativeness within the training data set. Furthermore, transfer learning DL models pre‐trained on large data sets still necessitate local data for retraining, thereby constraining their applicability. To address these challenges, we present a physics‐informed DL model based on a distributed framework. It involves spatial discretization and the establishment of differentiable hydrological models for discrete sub‐basins, coupled with a differentiable Muskingum method for channel routing. By introducing upstream‐downstream relationships, model errors in sub‐basins propagate through the river network to the watershed outlet, enabling the optimization using limited downstream streamflow data, thereby achieving spatial simulation of ungauged internal sub‐basins. The model, when trained solely on the downstream‐most station, outperforms the distributed hydrological model in streamflow simulation at both the training station and upstream held‐out stations. Additionally, in comparison to transfer learning models, our model requires fewer gauge stations for training, but achieves higher precision in simulating streamflow on spatially held‐out stations, indicating better spatial generalization ability. Consequently, this model offers a novel approach to hydrological simulation in data‐scarce regions, especially those with poor hydrological representativeness. Plain Language Summary: Climate change leads to more water shortages and disasters, requiring better streamflow predictions. Yet, a big hurdle in dealing with this issue is the lack of streamflow data across many parts of the world. Traditional physics‐based distributed hydrological models and current deep learning (DL) models have their limitations, especially for regions with unique hydrological processes and limited observations. To address these challenges, we developed a new tool combining physics‐informed DL and a traditional river routing model based on the distributed framework. The model divides the region into sub‐basins, where a physics‐informed DL rainfall‐runoff model calculates runoff generation, and a physics‐informed DL routing model computes the movement of water within each subunit toward the river. Model errors propagate downstream through the river network, thus requiring only a small amount of downstream data to optimize all sub‐basin models and effectively simulate internal unmonitored sub‐basins. When solely using the downstream‐most discharge stations for training, our model outperforms the traditional physics‐based distributed hydrological model. In addition, our approach requires less training data than transfer learning, while achieving higher spatial generalization accuracy. In summary, our model provides a new way to simulate streamflow in data‐scarce regions with unique processes. Key Points: A distributed physics‐informed deep learning hydrological model was proposed for data‐scarce regionsThe new model outperforms the traditional distributed hydrologic model in simulating streamflow in upstream held‐out stationsOur model requires less data for training but performs better than the transfer learning model in spatial generalization [ABSTRACT FROM AUTHOR]

Streamflow simulation is crucial for flood mitigation, ecological protection, and water resource planning. Process-based hydrological models and machine learning algorithms are the mainstream tools for streamflow simulation. However, their inherent limitations, such as time-consuming and large data requirements, make achieving high-precision simulations challenging. This study developed a hybrid approach to simultaneously improve the accuracy and computational efficiency of streamflow simulation, which integrates Block-wise use of the TOPMODEL (BTOP) model into the eXtreme Gradient Boosting (XGBoost), i.e., BTOP_XGB. In this approach, BTOP generates simulated streamflow using the Latin hypercube sampling algorithm instead of the time-consuming calibration algorithms to reduce computational costs. Then, XGBoost combines BTOP simulated streamflow with multi-source data to reduce simulation errors. In which, serval input variable selection algorithms are employed to choose relevant inputs and remove redundant information for model. The hybrid approach is validated and compared with a standalone model at three hydrological stations in the Jialing River basin, China. The results show that the performance of BTOP_XGB is significantly better than the BTOP and XGBoost models. The NSE of BTOP_XGB at Beibei, Xiaoheba, and Luoduxi stations increases by 54%, 21%, and 83%, respectively. Meanwhile, the computational time of BTOP_XGB is saved by >90% compared to the original calibrated BTOP. BTOP_XGB is less affected by parameter sample sizes and data amounts, demonstrating the robustness of the hybrid model. This study simplifies the complexity of the hydrological model and enhances the stability of machine learning, jointly improving the reliability of streamflow simulation. The hybrid approach provides a potential shortcut for streamflow simulation over basins with large areas or limited observed data.

Accurate hydrological modeling is crucial for water resources management. Process-based models (PBMs) are physically interpretable yet cannot fully utilize large datasets. Deep learning models, especially Long Short-Term Memory (LSTM) networks, have exhibited remarkable simulation accuracy but lack physical interpretability. To integrate the strengths of PBMs and LSTM, this paper develops three hybrid models (HLUDC, HLBDC, and HLBDCT) by coupling LSTM with the HBV model. HLUDC incorporates the output of HBV into LSTM to enhance modeling capability. HLBDC connects the HBV model to an LSTM that estimates the parameters of HBV. HLBDCT further introduces time-varying parameters. Notably, the possible limitation of these models in data-scarce basins is unclear, as all models are trained with available observations. Therefore, we further evaluate the impact of the training data length on model stability. The hybrid models are applied to the daily streamflow simulation in the Jialing River Basin. The results indicate that the hybrid models effectively enhance streamflow simulation compared to benchmark models. HLBDCT performs the best with a 2.44%-13.43% improvement in Nash-Sutcliffe efficiency coefficient over HBV and LSTM, followed by HLBDC, while HLUDC performs the least. HLBDCT also performed well in simulating extreme flow. Both HLBDCT and HLBDC accurately estimate actual evapotranspiration without being trained on it, demonstrating their robust physical coherence. Furthermore, HBV, HLBDC, and HLBDCT exhibit higher stability when trained with different lengths of data compared to HLUDC and LSTM. Overall, this study provides a better understanding of the potential for using hybrid models to enhance hydrological simulation accuracy. [ABSTRACT FROM AUTHOR]

Recent developments in computational techniques and data-driven machine-learning models (MLMs) have shown great potential in capturing the rainfall-runoff relationship. However, whether MLMs outperform the classical physical-based models (PBMs) in streamflow simulation is still controversial. In this study, we chose three representative catchments across the continental United States for a comparative analysis of these two model categories, including two PBMs, i.e., a conceptual hydrological model (EXP-HYDRO) and a classical semidistributed hydrological model (SWAT), and three MLMs, i.e., support vector regression (SVR), backpropagation artificial neural networks (BP-ANN), and a deep learning model, termed long short-term memory (LSTM). Results indicate that the bias of SVR and BP-ANN models is greater than PBMs under the baseline input scenario, while LSTM outperforms other models for the rainfall-runoff relationship. For delayed input scenarios, MLMs have great potential and perform satisfactorily. In addition, MLMs show better performance in the high-flow regime, while PBMs perform better in the low-flow regime, implying that both PBMs and MLMs have their own merits and should be jointly employed holistically to analyze the streamflow. Our comparison analysis demonstrates that the performance of MLMs and PBMs is variable under different seasonal, climatic, and topographic conditions, and conclude that MLMs can better capture the rainfall-runoff relationship than PBMs, when the coefficient of variation (COV) is large. [ABSTRACT FROM AUTHOR]

Study region: Tributaries of the Yellow River, Upper Fen River Basin, China. Study focus: The current research analyzed the streamflow changes resulting from interactions between climate and land use/cover (LULC) under Shared Socioeconomic Pathways (SSPs). The study utilized six General Circulation Models (GCMs) from Climate Model Intercomparison Project-6 (CMIP6) and the Land-Use Harmonization 2 (LUH2) dataset as forcing inputs for the Soil and Water Assessment Tool (SWAT) model. This study aims to quantitatively access the alone and combined effects of climate and LULC changes on future streamflow. New hydrologic insights for the region: The change in future streamflow under the combined effect of climate and LULC exceeded the change observed under the influence of climate or LULC alone in the Fenhe Reservoir Hydrological Station Control Basin (FRCB). The contributions were 72.5 % (SSP126), 73.1 % (SSP245), and 77.2 % (SSP585), respectively. Climate change showed a stronger correlation with future streamflow than LULC changes. Precipitation had the most direct effect on streamflow among the meteorological factors. The expansion of urban land contributes to an increase in streamflow in the source area and southwest tributaries of the FRCB to some extent. The future streamflow will gradually increase in mid-century (2045–2066) under low to high SSP. Adapting effective strategies for the FRCB, can be helpful in water resources planning and management.

The accurate and reliable simulation and prediction of runoff in the Beas River Basin are becoming more and more important due to the increased uncertainty posed by climate change, which is making it difficult to manage water resources efficiently. In order to minimize the effect of flash floods, estimating the accurate peak flow is essential. It can be challenging to comprehend and anticipate peak flow due to natural streamflow variance as well as the streamflow management offered by dams and reservoirs. Which makes it difficult to mimic hydrologic behavior on a daily scale with reliable accuracy. This study evaluated the efficacy of physics-aided machine learning (ML) based regression models for modeling streamflow in combination with process-based hydrological SWAT (soil and water assessment tool). Performance of eight machine learning (ML) models including linear regression (LR), multi-layer perceptron (MLP), light gradient-boosting machine (LGBM), extreme gradient boosting (XGBoost), kernel ridge (KR), elastic net (EN), Bayesian ridge (BR), and gradient boosting (GB) have been analyzed and compared with the calibrated-SWAT (cSWAT) model. The Nash–Sutcliffe efficiency (NSE), root mean square error (RMSE) and coefficient of determination (R2) were used to assess the effectiveness of both models. Results showed that the uncalibrated SWAT in combination with ML regression models (cSWAT-ML) performed well and found comparable to calibrated SWAT (cSWAT), though, few ML regression models in combination with uncalibrated SWAT (uSWAT-MLmodels models) performed superior. cSWAT model performed well with R2 values of 0.73, RMSE value of 276.92 m3/s and NSE value of 0.72. In uSWAT-ML, EN and BR have obtained better results with R2 values of 0.89 and 0.89, NSE values of 0.87 and 0.87, and RMSE values of 158.31 m3/s and 159.48 m3/s. The approached uSWAT-ML models have effectively predicted the peak stream flow rates with models BR and EN have predicted with better results of R2 value of 0.71 each. This study's findings highlight the potential of all the eight ML models as promising techniques for predicting the peak flow discharge values when uncalibrated process-based models are combined. [ABSTRACT FROM AUTHOR]

The established watershed's hydrologic simulation process, involving the calibration and validation of a model through spatiotemporal streamflow observations, has recently been adapted to encompass ungauged watersheds. The research introduces a methodology for simulating ungauged hydrosystems, suggesting the calibration and validation of regional rainfall‐runoff models using large‐scale hydrological models (LSMs) in the absence of in situ streamflow measurements. For doing so, LSM's 30‐years simulated discharges at various basins and subbasins of the case study River Basin District are assessed for their accuracy against the limited available observations. Subsequently, the rectified time series tunes the optimal parametrization of a regional lumped hydrologic model that is triggered by precipitation and temperature data from a partially dense meteorological network. The validated regional model is then applied in selective ungauged neighbouring subbasins employing the regionalization concept used in LSMs, that is in basins with similar physiographic characteristics to those where the regional model was set up. The performance of the research is assessed using statistical metrics. In cases where observations are available, the results show very satisfactory indexes (e.g., coefficient of determination: 0.654 < R2 < 0.949, Kling‐Gupta efficiency: 0.759 < KGE <0.790, Nash–Sutcliffe efficiency: 0.649 < NSE <0.745 for all the cases). In completely ungauged basins and in an order of magnitude index, the comparison of the simulated discharges with those from the LSM also indicates the satisfactory performance of the proposed approach. To sum up, the methodology underscores the alternate use of large and regional scale hydrological modelling coupling as an asset across various basin scales globally, offering reliable water quantity information. [ABSTRACT FROM AUTHOR]

Study Region: Sore and Masha river catchments in Baro Akobo river basin: Ethiopia. Study Focus: This research addresses the challenges associated with conventional data-driven streamflow modelling, which often exhibits inconsistent performance across different variability states. To bridge this gap, we explore the efficiency of model ensembles, a popular hydrological approach that harnesses the strengths of multiple models while preserving the fundamental characteristics of the data. Specifically, we compare three modified super ensemble meta-learners with eight single and hybrid machine learning base learners. The study also incorporates a semi-distributed HBV-light conceptual hydrological model and utilizes a decade of remote sensing and ground hydro-meteorological daily time series data. Different remote sensing-based vegetation index data products are employed to simulate single-step daily streamflow. The Recursive Feature Elimination (RFE) algorithm is applied to extract influential input parameters. New Hydrological Insights for the Region: Our findings consistently demonstrate that the three super ensemble learners outperform both the eight base models and the HBV-light model. The top-ranked Extra Tree Regression Super Ensemble (ETRSE) model exhibits superior performance, surpassing the HBV-light model by 24% according to the R2 performance measure. This study highlights the positive impact of selecting influential input parameters on the overall performance of machine learning models, providing valuable insights for hydrological modelling in the region.

Modeling streamflow is essential for understanding flow inundation. Traditionally, this involves hydrologic and numerical models. This research introduces a framework using agent-based modeling (ABM) combined with data-driven modeling (DDM) and Artificial Intelligence (AI). An agent-driven model simulates streamflow and its interactions with river courses and surroundings, considering hydrologic phenomena related to precipitation, water level, and discharge as well as channel and basin characteristics causing increased water levels in the Medio River. A five-year dataset of hourly precipitation, water level, and discharge measurements was used to simulate streamflow. The model’s accuracy was evaluated using statistical metrics like correlation coefficient (r), coefficient of determination (R2), root mean squared error (RMSE), and percentage error in peak discharge (Qpk). The ABM’s simulated peak discharge (Qpk) was compared with the measured peak discharge across four experimental scenarios. The best simulations occurred in scenario 3, using only rainfall and streamflow data. Data management and visualization facilitated input, output, and analysis. This study’s ABM combined with DDM and AI offers a novel approach for simulating streamflow and predicting floods. Future studies could extend this framework to other river basins and incorporate advanced sensor data to enhance the accuracy and responsiveness of flood forecasting.

Study region: Bahe River basin on the Tibetan Plateau, China. Study focus: The thawing of glaciers and snow in cold regions is critical to the impact of runoff. This study proposes a distributed degree-day Xin'anjiang model (DD-XAJ) considering ice and snow thawing, which added the degree-day module describing snow and ice thawing to the previously distributed Xin'anjiang model. A sensitivity analysis of the main parameters of the improved model was conducted to discuss the effect of snow and glacier melt on the runoff process. The DD-XAJ model was also applied to the Bahe River basin to analyze its runoff change process and identify the composition of runoff components in the basin. New hydrological insights: The DD-XAJ model was successfully applied to the Bahe River basin, which is located on the Tibetan Plateau, with good daily runoff simulation results. The mean values of the Nash efficiency coefficients were 0.828 and 0.859 in the calibration and validation periods, respectively, and the mean values of the correlation coefficients were 0.917 and 0.941, respectively. The runoff composition of the Bahe River basin is mainly rainfall runoff (76.3 %), snowmelt runoff (16.7 %), and glacier melt runoff (7 %), which indicates that rainfall runoff plays a dominant role in the Bahe River basin. Snowmelt runoff is influenced by rainfall events, while glacier melt runoff is mainly influenced by the seasons. In addition, the contribution of runoff components in different elevation zones differed significantly. The contribution of elevation 4000–5500 m to the total runoff was the largest. This study provides a new understanding of runoff processes and runoff components in the cryosphere basin.

The effects of climate change and of land use/cover change (LUCC) on streamflow as demonstrated by hydrological models are pressing issues on the frontiers of global environmental change research. The Nandu River Basin (NRB) as the largest of three river basins on the tropical Hainan Island, China, is subjected to an analysis of streamflow response to climate and to land-use change. It is based on the Soil and Water Assessment Tool (SWAT) coupled with climate change signals extracted from the global climate model data in the Coupled Model Intercomparison Project Phase 6 (CMIP6) and with land-use change scenarios modeled by Cellular Automata (CA)—Markov. The results are summarized as follows: (1) Climate change contributed more to streamflow change than land-use change in the NRB, with contributions of 97.57% and 2.43%, respectively. Precipitation and temperature were the most important climate variables, contributing 92.66% and 4.91% to streamflow change. (2) In the tropical island basin from 1990 to 2015, LUCC regulated the hydrological processes in the NRB and affected hydrological processes by increasing evapotranspiration and decreasing surface runoff and subsurface flow, which resulted in decreasing streamflow. (3) Under the climate change and land-use change scenarios of the near-term period (2021–2040), the annual streamflow decreased as during the reference period (1995–2014); particularly, it decreased most (−6.16%) on the SSP126 path. These results present a case study for understanding the hydrological cycle of tropical island basins and to provide a theoretical basis for water resources management and regional sustainable development of tropical islands. [ABSTRACT FROM AUTHOR]

Prediction in ungauged basins is a challenge and hot issue. Parameters regionalization is a useful methodology estimating hydrological model parameters in ungauged basins and has a critical effect on streamflow simulation. With kernel density estimation and Monte Carlo stochastic simulation methods, a framework was constructed to assess the uncertainty of simulated streamflow caused by parameters' error estimated by regionalization methodology. The Xin'anjiang model was applied for streamflow simulation in 42 small and medium-sized catchments with observed hydrologic stations located in the Guangxi Province. As each catchment being supposed an ungauged basin, the parameters of the Xin'anjiang model were calculated by regionalization methodologies including regression-based, similarity-based, and machine learning-based methodology. The performance of flood simulation using regression-based methodology was better than that of the similarity-based methodology. Using optimized machine learning-based regionalization methodology, the flood simulation accuracy was improved by 7%—15%. Compared with calibrated values, there were pronounced errors of model parameters estimated by parameters regionalization methodologies. The errors of sensitive parameters were lower than non-sensitive ones. The results indicated that there were significant uncertainties of randomly modeled floods by Monte Carlo methodology. The relative errors of simulated flood volumes and peak discharges were 10%—30% and 10%—40%, respectively. The results could provide a new technique for streamflow probability modeling and uncertainty assessment in ungauged basins. And this would be useful for flood forecasting and disaster prevention in small and medium-sized rivers. [ABSTRACT FROM AUTHOR]

This study sought to project future changes in hydroclimatic variables and to establish how climate change affects flood inundation extent in the Pungwe River Basin. Climate ensembles of 10 Regional Climate Models from the CORDEX project were selected. The historical rainfall and temperature time series and the downscaled climate data were input into the HBV model to generate streamflow for the 2022–2099 period. Flood extents for 50-, 100-and 1,000-year return periods were predicted using the HEC-RAS hydraulic model. By 2070, annual rainfall at all nine studied meteorological stations is predicted to reduce by a maximum of 61%. Temperature is expected to increase up to 1.5% over the same period. By the 2070s, simulations from HBV revealed that the peak flows for the Pungwe River Basin will increase by up to 100% and decrease by approximately 57% as projected by the model ensemble. The analyses also show that by 2070 climate change may cause a minimum of 2,784.4 km2 and a maximum of 8,235.6 km2 of flood extension. These results are essential for decision-making on flood hazard mapping and early warning systems, prompting a pathway for sustainable development. HIGHLIGHTS Satellite rainfall estimates, RCMs, hydrological, and hydraulic models for flood inundation extent mapping.; By 2070, the future projections show increasing temperature and decreasing rainfall.; Climate change will, however, increase peak flows and inundation extent by the 2070s.; Significantly affected places in the Pungwe River Basin include cropland, built-up areas, roads, schools, hospitals, and business infrastructure.;

Insufficient precipitation observations hinder the bias-correction of Global Climate Model (GCM) precipitation outputs in ungauged and remote areas. As a result, the reliability of future precipitation and water resource projections is restricted for these areas. Open-access quantitative precipitation estimation (QPE) products offer a potential solution to this challenge. This study assesses the effectiveness of three widely used, long-term QPEs, including ERA5, PERSIANN-CDR, and CHIRPS, in bias-correcting precipitation outputs from the CMIP6 GCMs. The evaluation involves the reproduction of precipitation distribution, streamflow simulation utility based on a hydrological model, and the accuracy of extreme indices associated with rainstorm/flood/drought events. This study selects the Beijiang basin located in the subtropical monsoon area of South China as the case study area. The results demonstrate that bias-correction using QPEs improves the performance of GCM precipitation outputs in reproducing precipitation/streamflow distribution and extreme indices, with a few exceptions. PCDR generally exhibits the most effective bias-correction utility, consistently delivering reasonable performance across various cases, making it a suitable alternative to gauge data for bias-correction in ungauged areas. However, GCM outputs corrected by ERA5 tend to overestimate overall precipitation and streamflow (by up to about 25% to 30%), while the correction results of CHIRPS significantly overestimate certain extreme indices (by up to about 50% to 100%). Based on the revealed performance of QPEs in correcting GCM outputs, this study provides references for selecting QPEs in GCM-based water resource projections in remote and ungauged areas. [ABSTRACT FROM AUTHOR]

Global warming has led to a high incidence of extreme hydrometeorological events. Predicting the spatial and temporal response of water resources to future climate change is necessary for the construction of regional water facilities. In this study, Global Circulation Models (GCMs) and the Soil and Water Assessment Tool (SWAT) were combined with the Delta downscaling method in the Dongyu River Basin (DRB), China. Three different GCMs under two Shared Socioeconomic Pathways (SSPs) 2–4.5 and 3–7.0 scenarios were considered. Relative to the baseline (2008–2018), streamflow under two scenarios was assessed from 2025 to 2064. The results show that the mean annual precipitation would rise by 12.3% and 12.8% under SSP2-4.5 and SSP3-7.0 scenarios, respectively. The mean monthly maximum temperature and minimum temperature would increase in summer but decrease in winter under two scenarios. Future climate change was estimated to increase the mean annual streamflow by 9.00% under SSP2-4.5 scenario and by 6.92% under SSP3-7.0 scenario. The wet season was forecast to arrive earlier and the frequency of extreme events would likely increase in the future. A high positive correlation was projected between the annual precipitation and the annual streamflow, with an average correlation coefficient of 0.76 (p < 0.05). The dominant climate variables would be different on the monthly scale. Streamflow in the dry season would be more susceptible to temperature changes. The findings of this study show how climate change affects streamflow in the DRB and give adaptive management a scientific foundation. [ABSTRACT FROM AUTHOR]

Streamflow modelling is one of the most important elements for the management of water resources and flood control in the context of future climate change. With the advancement of numerical weather prediction and modern detection technologies, more and more high-resolution hydro-meteorological data can be obtained, while traditional physical hydrological models cannot make full use of them. In this study, a hybrid deep learning approach is proposed for the simulation of daily streamflow in two mountainous catchments of the Daqing River Basin, northern China. Two-dimensional high-resolution (1 km) output data from a WRF model were used as the model input, a convolutional neural network (CNN) model was used to extract the physical and meteorological characteristics of the catchment at a certain time, and the long short-term memory (LSTM) model was applied to simulate the streamflow using the time-series data extracted by the CNN model. To reduce model input noise and avoid overfitting, the Gamma test method was adopted and the correlations between the input variables were checked to select the optimal combination of input variables. The performance of the CNN-LSTM models was acceptable without using the Gamma test (i.e., with all WRF input variables included), with NSE and RMSE values of 0.9298 and 9.0047 m3/s, respectively, in the Fuping catchment, and 0.8330 and 1.1806 m3/s, respectively, in the Zijingguan catchment. However, it was found that the performance of the model could be significantly improved by the use of the Gamma test. Using the best combination of input variables selected by the Gamma test, the NSE of the Fuping catchment increased to 0.9618 and the RMSE decreased to 6.6366 m3/s, and the NSE of the Zijingguan catchment increased to 0.9515 and the RMSE decreased to 0.6366 m3/s. These results demonstrate the feasibility of the CNN-LSTM approach for flood streamflow simulation using WRF-downscaled high-resolution data. By using this approach to assess the potential impacts of climate change on streamflow with the abundant high-resolution meteorological data generated by different climate scenarios, water managers can develop more effective strategies for managing water resources and reducing the risks associated with droughts and floods. [ABSTRACT FROM AUTHOR]

In this study, the hydrological system of the Weather Research and Forecasting model (WRF-Hydro) is applied to simulate the streamflow at the Kaidu River Basin, which is vital to the ecological system in the lower reaches of the Tarim River in Northwest China. The offline WRF-Hydro model is coupled with the Noah multi-parameterization land surface model (Noah-MP) and is forced by the China Meteorological Forcing Dataset (CMFD), with the grid spacing of the hydrological routing modules being 250 m. A 3-year period (1983–1985) is used for calibration and a 17-year period (1986–2002) for the evaluation. Several key parameters of WRF-Hydro and four Noah-MP parameterization options are calibrated, and the performance of WRF-Hydro with the optimized model setting is evaluated using the daily streamflow observations. The results indicate that WRF-Hydro can reproduce the observed streamflow reasonably, with underestimation of the streamflow peaks. The simulated streamflow is sensitive to the parameters of bexp, dksat, smcmax, REFKDT, slope, OVROUGHRTAC and mann in the Kaidu River Basin. At the same time, the parameterization options of Noah-MP also have a large influence on the streamflow simulation. The WRF-Hydro model with optimized model settings can achieve correlation coefficient (CC) and Nash efficiency coefficient (NSE) statistical scores of 0.78 and 0.61, respectively, for the calibration period. Meanwhile, for the evaluation period, the scores are 0.7 and 0.50, respectively. This study indicates the importance of applying the physical-based WRF-Hydro model over Northwest China and provides a reference for the nearby regions. [ABSTRACT FROM AUTHOR]

Study region: The Dongting Lake Basin is a typical regional study in humid southern China with data-scarce areas. Study focus: This study improved the streamflow simulation by coupling hydrological and hydrodynamic models and considering water yields in data-scarce areas. We constructed a soil and water assessment tool (SWAT) hydrological model of the Dongting Lake Basin to simulate the streamflow in the data-scarce areas, which was further coupled into the MIKE21 hydrodynamic model as additional boundary conditions. New hydrological insights: The results showed that the relative error of streamflow simulation was reduced from 24.64 % to 10.50 % in the coupled hydrological-hydraulic model over the singular hydrodynamic model, which also indirectly verified the results of streamflow simulation in the data-scarce area. Based on the coupled model, the annual water yields in the data-scarce areas were estimated to be 38.95 × 109 m3, representing 15.13 % of the yearly water yields in the basin. The water yields in the data-scarce areas showed a seasonal variation, which was concentrated from April to July. The monthly water balance error of the Dongting Lake Basin was significantly reduced (57.42 %) when considering the water yields in data-scarce areas. The model-coupling approach in this study can be applied to other data-scarce areas to improve streamflow simulation.

Streamflow simulation in a snow dominated basin is complex due to the presence of a high number of interrelated hydrological processes. This complexity is affected by the delayed responses of the catchment to snow accumulation and snow melting processes. In this study, long short-term memory (LSTM) and artificial neural network (ANN) models were utilized for rainfall–runoff simulation in a snow dominated basin, the Carson River basin in the United States (US). The input structure of the models was determined using the simulated annealing algorithm with a naïve Bayes model from a high dimensional feature space to represent the long-term impacts of historical events (i.e. the hysteresis effect) on current observations. Further, to represent the different responses of the catchment in the model structure, a base flow separation method was included in the simulation framework. The obtained performance indices, root mean square error, percentage bias, Nash–Sutcliffe and Kling–Gupta efficiencies are 0.331 m3 s−1, 13.00%, 0.848, and 0.852 for the ANN model and 0.235 m3 s−1, − 0.80%, 0.923, and 0.934 for the LSTM model, respectively. The proposed methodology was found to be promising for improving the streamflow simulation capability of LSTM and ANN models by only considering precipitation, temperature, and potential evapotranspiration as input variables. Analysing the flow duration curves indicated that the LSTM model is more efficient in representing different flow dynamics within the basin due to embedded cell states. Further, the uncertainty and reliability analyses were conducted by using expanded uncertainty ( U 95 ), reliability, and resilience indices. The obtained U 95 , reliability and resilience indices are 1.78–1.72 m3 s−1, 31.28–66.67% and 11.58–38.27% for the ANN and LSTM models, respectively, showed that the LSTM model produced less uncertainty and is more reliable. However, while lacking a memory component, the proposed methodology significantly contributes to the simulation capability of the ANN model in rainfall–runoff modelling. The results of this study indicated that the proposed methodology could enhance the learning capabilities of machine learning models in rainfall–runoff simulation. [ABSTRACT FROM AUTHOR]

Climate change impact studies that evaluated the biases of climate models' simulations showed the presence of large systematic errors in their outputs. However, many studies continue to arbitrarily select bias correction methods for error reduction. This work evaluated the implications of bias correction methods on the projections of climate change impact on streamflow of the Gidabo sub-basin, Ethiopia. Climate outputs from four global climate model and regional climate model (GCM–RCM) combinations for the representative concentration pathway (RCP4.5) scenario were used. Five bias correction methods were used to reduce the systematic errors of the simulated rainfall data. The future changes in rainfall pattern, evapotranspiration, and streamflow were analyzed by using their relative percentage difference between the projected and the baseline period. The distribution mapping method provided better results in mean and extreme rainfall cases. This is also reflected in streamflow projections, as the daily interquartile range value indicates the lowest variability of the projected streamflow. The wet season streamflow will likely decrease in the future, whereas the short rainy season streamflow will increase. Our findings show that climate models and bias correction methods considerably limit the magnitude of future projections of streamflow. However, similar research should be conducted in other catchments to extend the conclusions of this study. HIGHLIGHTS COordinated Regional Climate Downscaling Experiment (CORDEX)-Africa domain climate model data were used for impact analysis.; Projected climate change impact on streamflow is affected both by climate models and bias correction methods.; The distribution mapping method provided better results in correcting the errors of mean and extreme rainfall projection.; Wet season streamflow of the study area will likely decrease in the future.;

Streamflow simulation is crucial for flood mitigation, ecological protection, and water resource planning. Process-based hydrological models and machine learning algorithms are the mainstream tools for streamflow simulation. However, their inherent limitations, such as time-consuming and large data requirements, make achieving high-precision simulations challenging. This study developed a hybrid approach to simultaneously improve the accuracy and computational efficiency of streamflow simulation, which integrates Block-wise use of the TOPMODEL (BTOP) model into the eXtreme Gradient Boosting (XGBoost), i.e., BTOP_XGB. In this approach, BTOP generates simulated streamflow using the Latin hypercube sampling algorithm instead of the time-consuming calibration algorithms to reduce computational costs. Then, XGBoost combines BTOP simulated streamflow with multi-source data to reduce simulation errors. In which, serval input variable selection algorithms are employed to choose relevant inputs and remove redundant information for model. The hybrid approach is validated and compared with a standalone model at three hydrological stations in the Jialing River basin, China. The results show that the performance of BTOP_XGB is significantly better than the BTOP and XGBoost models. The NSE of BTOP_XGB at Beibei, Xiaoheba, and Luoduxi stations increases by 54%, 21%, and 83%, respectively. Meanwhile, the computational time of BTOP_XGB is saved by >90% compared to the original calibrated BTOP. BTOP_XGB is less affected by parameter sample sizes and data amounts, demonstrating the robustness of the hybrid model. This study simplifies the complexity of the hydrological model and enhances the stability of machine learning, jointly improving the reliability of streamflow simulation. The hybrid approach provides a potential shortcut for streamflow simulation over basins with large areas or limited observed data. • A hybrid approach is proposed by combining the hydrological model (BTOP) with machine learning (XGBoost) and multi-source data. • The hybrid approach significantly improves the accuracy of streamflow simulation. • The Latin hypercube sampling algorithm reduce the computing time of original BTOP by >90%. • Input variable selection algorithms can reduce data redundancy and choose more relevant information for the hybrid model. • The hybrid approach is less affected by the data volume and parameter sample sizes, providing a potential shortcut to simulate streamflow in data-scarce basins. [ABSTRACT FROM AUTHOR]

Copyright of China Rural Water & Hydropower is the property of China Rural Water & Hydropower Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In this study, five hydrological models, including the soil and water assessment tool (SWAT), identification of unit hydrograph and component flows from rainfall, evapotranspiration, and streamflow (IHACRES), Hydrologiska Byråns Vattenbalansavdelning (HBV), Australian water balance model (AWBM), and Soil Moisture Accounting (SMA), were used to simulate the flow of the Hablehroud River, north-central Iran. All the models were validated based on the root mean square error (RMSE), coefficient of determination (R2), Nash-Sutcliffe model efficiency coefficient (NS), and Kling-Gupta efficiency (KGE). It was found that SWAT, IHACRES, and HBV had satisfactory results in the calibration phase. However, only the SWAT model had good performance in the validation phase and outperformed the other models. It was also observed that peak flows were generally underestimated by the models. The sensitivity analysis results of the model parameters were also evaluated. A hybrid model was developed using gene expression programming (GEP). According to the error measures, the ensemble model had the best performance in both calibration (NS = 0.79) and validation (NS = 0.56). The GEP combination method can combine model outputs from less accurate individual models and produce a superior river flow estimate. HIGHLIGHTS The semi-distributed entirely conceptual model (SWAT) had relatively better performance than the lumped semi-conceptual models (IHACRES, HBV light, AWBM, and SMA).; The GEP was used to construct a hybrid model by ensembling the five calibrated hydrological models to improve the results.; The ensemble models and SWAT yielded high flow and low flow calculations close to the observed data in both the calibration and validation phases.;

The impacts of changing climate and land use/cover on streamflow in the Kibungo sub-catchment were evaluated using the Soil and Water Assessment Tool (SWAT). Rainfall, minimum, and maximum temperature data for six stations from the ensemble mean of the RCMs (RCA4, RACMO22T, CCLM4-8-17) were used under RCP 4.5 and RCP 8.5. The homogeneity and trend test were used to detect the change point and identify the pattern in the annual time series, respectively. Land Change Modeler (LCM) was used to predict land use maps of 2040 and 2070 from historical maps. Streamflow was simulated for 2021–2040 and 2041–2070 based on the homogeneity test results. The model calibration (2009–2016) and validation (2009–2016) for streamflow were successful. The homogeneity test detects a change point in 2040. A significant decrease in annual rainfall by 22.9 mm/yr (RCP 4.5) and 57 mm/yr (RCP 8.5) for 2021–2040 and an insignificant decrease was obtained during 2041–2070 under both emission scenarios. The annual temperature increased insignificantly by 0.004 °C/yr under RCP 4.5 while a significant increase of 0.21 °C/yr under RCP 8.5 was observed for 2021–2040. For 2041–2070, a significant increase of 0.016 °C/yr (RCP 4.5) and 0.045 °C/yr (RCP 8.5) was observed. The change in land use/cover resulted in increasing the build-up area (84%), agricultural fields (55.6%), and a decrease in the forest area (10.5%) during 2021–2040. During 2041–2070, the build-up area increased by 32.1%, the agricultural field by 36%, and the forest area decreased by 11%. Streamflow decreased significantly by 65.4 m3/yr (RCP 4.5) and 195.9 m3/yr (RCP 8.5) from 2021 to 2040. An insignificant decrease of 13.7 m3/s (RCP 4.5) and 7.63 m3/s (RCP 8.5) was observed during 2041–2070. This study provides an insight to the managers, planners, and policymakers on environmental protection/conservation for sustainable utilization of water resources at the Kibungo sub-catchment.

Study region: Varahi River originating from the Western Ghats of India. Study focus: We developed a hybrid model that integrates process-based hydrological model (PHM) and data-driven (DD) techniques to generate streamflow simulations precisely. The hybrid modeling framework is practical as it respects hydrological processes through the PHM while considering the advantage of the DD model's ability to simulate the complex relationship between residuals and input variables. Further, we have proposed an optimization-based nudging scheme for post-processing the hybrid model simulated streamflow to overcome the limitations in PHM and DD. New hydrological insights for the region: We formulated two approaches for simulating streamflow ensembles using DD and PHM models. In approach− 1, DD models are initially used to ensemble meteorological variables and then use the ensembles in a PHM to simulate streamflows. In approach− 2, PHM is forced with different sets of meteorological variables to simulate multiple streamflow sets and then use DD models to ensemble the PHM-derived streamflows. Random forest exhibited better performance for ensembling precipitation, temperature, and streamflow datasets compared to the other five DD algorithms in the study. Streamflows generated using approach− 2 showed reliable estimates when compared against observed streamflow values. However, post-processing the hybrid streamflows using an optimization-based nudging scheme outperformed the streamflows generated in approach− 1 and approach− 2 with better model fit statistics (R2 and NSE of 0.69 and 0.66). The output from the nudging scheme was further utilized for streamflow predictions under the combined impact of land use/cover (LULC) and climate change (CC) under the Representative Concentration Pathway 4.5 scenario. It depicted a decrease in monthly and seasonal stream flows with − 22.65 %, − 31.77 %, − 11.81 % for winter, summer, and monsoon seasons, respectively. These results suggest that water availability will decline, and water scarcity will increase in the study region. These variations in streamflow might negatively impact agriculture and natural ecosystems and even lead to water restrictions in the region.

Insufficient precipitation observations hinder the bias-correction of Global Climate Model (GCM) precipitation outputs in ungauged and remote areas. As a result, the reliability of future precipitation and water resource projections is restricted for these areas. Open-access quantitative precipitation estimation (QPE) products offer a potential solution to this challenge. This study assesses the effectiveness of three widely used, long-term QPEs, including ERA5, PERSIANN-CDR, and CHIRPS, in bias-correcting precipitation outputs from the CMIP6 GCMs. The evaluation involves the reproduction of precipitation distribution, streamflow simulation utility based on a hydrological model, and the accuracy of extreme indices associated with rainstorm/flood/drought events. This study selects the Beijiang basin located in the subtropical monsoon area of South China as the case study area. The results demonstrate that bias-correction using QPEs improves the performance of GCM precipitation outputs in reproducing precipitation/streamflow distribution and extreme indices, with a few exceptions. PCDR generally exhibits the most effective bias-correction utility, consistently delivering reasonable performance across various cases, making it a suitable alternative to gauge data for bias-correction in ungauged areas. However, GCM outputs corrected by ERA5 tend to overestimate overall precipitation and streamflow (by up to about 25% to 30%), while the correction results of CHIRPS significantly overestimate certain extreme indices (by up to about 50% to 100%). Based on the revealed performance of QPEs in correcting GCM outputs, this study provides references for selecting QPEs in GCM-based water resource projections in remote and ungauged areas.

Study region: The Terauchi catchment, southwestern Japan Study focus: This paper evaluates two gauge-merged precipitation datasets derived from weather radars (R/A) and satellites (GsMAP_G) based on their capability to improve streamflow simulation using the Soil and Water Assessment Tool (SWAT) model. A third dataset includes measurements from local rain gauges used for producing the R/A database but not the GsMAP_G product was prepared for comparison reasons. The R/A and GsMAP_G data were first compared to gauge observations. The performance and prediction uncertainty of the SWAT model forced by the evaluated datasets were subsequently quantified and compared. New hydrological insights for the region: The R/A dataset overestimated the precipitation, while the GsMAP_G underestimated it. After calibration, the R/A performed best (NSE = 88–91%), followed by the Gauge (NSE = 84%) and GsMAP_G (NSE = 54%) scenarios. The R/A product improved the overall simulation performance by 6.50% and 62.40% in terms of NSE and absolute percent bias compared to the Gauge model. The performance of the evaluated datasets varied depending on streamflow occurrence exceedance probability (OEP). The R/A dataset improved the simulation of extremely high (OEP < 1%) and low (OEP > 60%) streamflow events as it resulted in the lowest simulation biases and errors. The current investigation suggests the use of the R/A product for improving the simulation of daily streamflow, including hydro-climatic extremes.

• HRU precipitation data were obtained based on WRF precipitation data and area weighting methods. • We improved the SWAT model with HRU-based precipitation, and better simulated runoff in mountain watershed. • The modified SWAT model showed significant improvement in accurately estimating runoff during periods of high water compared to the original SWAT model. Precipitation is crucial in simulating hydrological processes using the Soil and Water Assessment Tool (SWAT), which directly impacts runoff and the transport of nutrient loads. Achieving precipitation data at high spatial resolution is essential for accurately forecasting hydrological simulations within watersheds. The primary uncertainty in SWAT arises from the homogenization of precipitation within Hydrological Response Units (HRUs) in the same subbasin, a result of the simplistic method employed by the precipitation subroutine for reading and distributing precipitation data. In this study, we propose the utilization of the area-weighted method to calculated HRU-specific precipitation data using the Weather Research and Forecast (WRF) model for the first time. Additionally, we enhance the precipitation allocation mechanism within the SWAT model subroutine, enabling the derivation of precipitation in the basin from each HRU rather than relying solely on average rainfall. The performance of the modified SWAT model was evaluated in a representative mesoscale mountainous basin covering 5172 km2 within the Yangtze River Basin, presenting substantial improvements compared to the original SWAT model. Specifically, the use of HRU-specific precipitation data effectively captures the spatial variability of rainfall, resulting in a significant enhancement in the accuracy of runoff simulation. This enhancement is evidenced by an increase of 0.25 in the coefficient of determination (R2) and 0.27 in the Nash–Sutcliffe Efficiency (NSE). Moreover, it was observed that the original SWAT model notably underestimated monthly runoff during wet periods, with a percent bias (PBIAS) ranging from −19.20 % to −5.28 %. This discrepancy is largely attributed to the inaccurate representation of rainfall heterogeneity. Our findings emphasize the importance of accurately of rainfall in the precipitation module to improve the SWAT model's assessment of water resources in mountainous basins. [ABSTRACT FROM AUTHOR]

Streamflow modelling is one of the most important elements for the management of water resources and flood control in the context of future climate change. With the advancement of numerical weather prediction and modern detection technologies, more and more high-resolution hydro-meteorological data can be obtained, while traditional physical hydrological models cannot make full use of them. In this study, a hybrid deep learning approach is proposed for the simulation of daily streamflow in two mountainous catchments of the Daqing River Basin, northern China. Two-dimensional high-resolution (1 km) output data from a WRF model were used as the model input, a convolutional neural network (CNN) model was used to extract the physical and meteorological characteristics of the catchment at a certain time, and the long short-term memory (LSTM) model was applied to simulate the streamflow using the time-series data extracted by the CNN model. To reduce model input noise and avoid overfitting, the Gamma test method was adopted and the correlations between the input variables were checked to select the optimal combination of input variables. The performance of the CNN-LSTM models was acceptable without using the Gamma test (i.e., with all WRF input variables included), with NSE and RMSE values of 0.9298 and 9.0047 m3/s, respectively, in the Fuping catchment, and 0.8330 and 1.1806 m3/s, respectively, in the Zijingguan catchment. However, it was found that the performance of the model could be significantly improved by the use of the Gamma test. Using the best combination of input variables selected by the Gamma test, the NSE of the Fuping catchment increased to 0.9618 and the RMSE decreased to 6.6366 m3/s, and the NSE of the Zijingguan catchment increased to 0.9515 and the RMSE decreased to 0.6366 m3/s. These results demonstrate the feasibility of the CNN-LSTM approach for flood streamflow simulation using WRF-downscaled high-resolution data. By using this approach to assess the potential impacts of climate change on streamflow with the abundant high-resolution meteorological data generated by different climate scenarios, water managers can develop more effective strategies for managing water resources and reducing the risks associated with droughts and floods.

Precipitation is the most important climate variable in hydrological practices, so accurate estimation of its intensity and volume is very crucial for hydrological applications. Remote sensing precipitation estimations have recently been widely employed in water resources management due to the lack of observed precipitation measurements in remote areas. However, remote sensing precipitation estimations are not free from systematic errors. This study aims to bias-correct the Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) satellite precipitation estimations using the Gaussian-Copula approach and illustrates how it improves the simulated flow characteristics in the Shah Bahram basin in Kohgiluyeh and Boyer-Ahmad Province, southwestern Iran. The Nash-Sutcliff Efficiency (NSE) calculated between the original CHIRPS precipitation estimation and observation in the Shah Bahram basin equals − 0.14; however, when bias-corrected CHIRPS data was compared to observation, the NSE increased to 0.23, suggesting about 158% improvement in the CHIRPS precipitation estimation when bias-corrected with the Gaussian-Copula approach. Next, the bias-corrected precipitation time series were utilized as the hydrologic modeling system inputs to simulate flow specifications such as discharge and peak value. Then, the simulation of the flow parameters was carried out with both original and bias-corrected CHIRPS satellite precipitation estimations and the ground-based precipitation. Though the NSE statistic of the simulation for the testing period has not changed significantly, the Pbias statistic has considerably improved. The result of the study indicates the good performance of the proposed bias correction approach in reducing the CHIRPS satellite estimations errors, concluding that it is a suitable approach for bias correction of the other satellite precipitation estimations in areas that suffer from the lack of ground-based observations necessary for flood forecasting and other hydrological practices. [ABSTRACT FROM AUTHOR]

The global scale wetland loss or degeneration triggers the assessment of how their function provisions are likely to change under different management scenarios. However, how and to what extent river damming can modify the hydrological function of wetlands remain largely unknown. In this study, we apply a distributed hydrological modeling platform for a larger river basin in Northeast China with a paired modeling scenario: (1) modeling with no dam present (i.e., under natural conditions), and (2) modeling with dam present (i.e., under disturbed conditions). The overarching goal of the study is to quantify the effect of damming on wetland hydrological processes. The modeling study demonstrates that river damming can alter the wetland effect on daily flow by significantly reducing flow under low flow conditions but slightly increasing flow under high flow conditions. Damming can impact the wetland function in alleviating floods with a 7% decrease when compared to the natural conditions without a dam. Consequently, the supporting effect of wetlands on baseflow is weakened substantially by damming regulation. These results indicate that river damming can impair flow regulation functions of downstream wetlands, and therefore, new flow regulation and optimization strategies for achieving complementary hydrological functions of wetlands and reservoirs are important to maximize basin resilience to hydrological extremes under climate change. [ABSTRACT FROM AUTHOR]

The WRF model is nowadays the most widely applied mesoscale numerical weather prediction model. Its land-surface hydrological modeling module, WRF-Hydro, which is designed to facilitate the land-surface modeling being coupled with WRF, draws more and more attentions from both the meteorological and the hydrological community. In this study, four sensitive and principle parameters of WRF-Hydro are tested in semi-humid and semi-arid areas of northern China. These parameters include the runoff infiltration parameter (REFKDT), the surface retention depth (RETDEPRT) controlled by a scaling parameter named RETDEPRTFAC, the channel Manning roughness parameter (MannN), and the overland flow roughness parameter (OVROUGHRT) controlled by the scaling parameter OVROUGHRTFAC. WRF-Hydro is designed with a 100-m horizontal grid spacing in two catchments of northern China. The performance of WRF-Hydro with different parameterisation combination schemes is tested for simulating a typical 24-h storm events with uniform rainfall evenness in space and time. The Nash-Sutcliffe efficiency and the root mean squared error of the simulated streamflow, together with the cumulative amount of the simulated rainfall is chosen as the evaluation statistics. It is found that REFKDT and MannN are the most sensitive parameter among the four parameters, and the case is especially evident with unsaturated soil conditions. In order to obtain the most reasonable value range, REFKDT and MannN are further verified by another three 24-h storm events with different spatial and temporal evenness. The range of REFKDT from 2.0 to 3.0, and the MannN scale factor from 1.5 to 1.8 is found to give the best results. The findings of this study can be used as references for calibration of the WRF-Hydro modeling system in semi-humid and semi-arid regions with similar rainfall-runoff response characteristics. The methodologies to design and test the combination schemes of parameterisations can also be regarded as a reference for evaluation of the WRF/WRF-Hydro coupled system for land-surface process modeling. [ABSTRACT FROM AUTHOR]

Simulating streamflow under both high-and low-flows is required for versatile eco-hydrological modeling. Typical streamflow simulators require hydrological data such as river geometry and observed river discharge/water level as upstream/downstream boundary conditions. However, these are not always available in data-sparse regions. Furthermore, because of the potential inaccuracy of digital elevation model (DEM) data around water surfaces, this data has not generally been utilized in streamflow simulations. Therefore, this study explores the potential applicability of DEM data to extract river cross-sections, focusing on the upstream Hiikawa River, Japan. A 1-D streamflow simulation was performed using river cross-sections extracted from a 5 m LiDAR DEM and the observed dam discharge from 2018 to 2020 as the upstream boundary condition. The simulated water depths with Manning's roughness coefficients of 0.03 to 0.05 m-1/3 s reproduce the observation results with Nash-Sutcliffe coefficients of 0.91-0.97 for the whole period and 0.60-0.97 for a flood event. The accurate results for both low and high flows were considered to reflect the reasonable representations of the river cross-section. Finally, the velocity-based suitability index for Ayu (P. altivelis) was evaluated. We demonstrate applicability and several possible limitations of DEM data for eco-hydrological modeling of data-scarce rivers. [ABSTRACT FROM AUTHOR]

Streamflow modeling is considered as an essential component for water resources planning and management. There are numerous challenges related to streamflow prediction that are facing water resources engineers. These challenges due to the complex processes associated with several natural variables such as non-stationarity, non-linearity, and randomness. In this study, a new model is proposed to predict long-term streamflow. Several lags that cover several years are abstracted using the potential of Extreme Gradient Boosting (XGB) then after the selected inputs variables are imposed into the predictive model (i.e., Extreme Learning Machine (ELM)). The proposed model is compared with the stand-alone schema in which the optimum lags of the variables are supplied into the XGB and ELM models. Hydrological variables including rainfall, temperature and evapotranspiration are used to build the model and predict the streamflow at Goksu-Himmeti basin in Turkey. The results showed that XGB model performed an excellent result in which can be used for predicting the streamflow pattern. Also, it is clear from the attained results that the accuracy of the streamflow prediction using XGB technique could be improved when the high number of lags was used. However, the implementation of the XGB is tree-based technique in which several issues could be raised such as overfitting problem. The proposed schema XGBELM in which XGB approach is selected the correlated inputs and ranking them according to their importance; then after, the selected inputs are supplied into the ELM model for the prediction process. The XGBELM model outperformed the stand-alone schema of both XGB and ELM models and the high-lagged schema of the XGB. It is important to indicate that the XGBELM model found to improve the prediction ability with minimum variables number.

Modeling the effect of spatial variability of precipitation and basin characteristics on streamflow requires the use of distributed or semi-distributed hydrologic models. This paper addresses a DMIP 2 study that focuses on the advantages of using a semi-distributed modeling structure. We first present a revised semi-distributed structure of the NWS SACramento Soil Moisture Accounting (SAC-SMA) model that separates the routing of fast and slow response runoff components, and thus explicitly accounts for the differences between the two components. We then test four different calibration strategies that take advantage of the strengths of existing optimization algorithms (SCE-UA) and schemes (MACS). These strategies include: (1) lumped parameters and basin averaged precipitation, (2) semi-lumped parameters and distributed precipitation forcing, (3) semi-distributed parameters and distributed precipitation forcing and (4) lumped parameters and basin averaged precipitation, modified using a priori parameters of the SAC-SMA model. Finally, we explore the value of using discharge observations at interior points in model calibration by assessing gains/losses in hydrograph simulations at the basin outlet. Our investigation focuses on two key DMIP 2 science questions. Specifically, we investigate (a) the ability of the semi-distributed model structure to improve stream flow simulations at the basin outlet and (b) to provide reasonably good simulations at interior points.The semi-distributed model is calibrated for the Illinois River Basin at Siloam Springs, Arkansas using streamflow observations at the basin outlet only. The results indicate that lumped to distributed calibration strategies (1 and 4) both improve simulation at the outlet and provide meaningful streamflow predictions at interior points. In addition, the results of the complementary study, which uses interior points during the model calibration, suggest that model performance at the outlet can be further improved by using a semi-distributed structure calibrated at both interior points and the outlet, even when only a few years of historical record are available. © 2009 Elsevier B.V.

Modeling the effect of spatial variability of precipitation and basin characteristics on streamflow requires the use of distributed or semi-distributed hydrologic models. This paper addresses a DMIP 2 study that focuses on the advantages of using a semi-distributed modeling structure. We first present a revised semi-distributed structure of the NWS SACramento Soil Moisture Accounting (SAC-SMA) model that separates the routing of fast and slow response runoff components, and thus explicitly accounts for the differences between the two components. We then test four different calibration strategies that take advantage of the strengths of existing optimization algorithms (SCE-UA) and schemes (MACS). These strategies include: (1) lumped parameters and basin averaged precipitation, (2) semi-lumped parameters and distributed precipitation forcing, (3) semi-distributed parameters and distributed precipitation forcing and (4) lumped parameters and basin averaged precipitation, modified using a priori parameters of the SAC-SMA model. Finally, we explore the value of using discharge observations at interior points in model calibration by assessing gains/losses in hydrograph simulations at the basin outlet. Our investigation focuses on two key DMIP 2 science questions. Specifically, we investigate (a) the ability of the semi-distributed model structure to improve stream flow simulations at the basin outlet and (b) to provide reasonably good simulations at interior points.The semi-distributed model is calibrated for the Illinois River Basin at Siloam Springs, Arkansas using streamflow observations at the basin outlet only. The results indicate that lumped to distributed calibration strategies (1 and 4) both improve simulation at the outlet and provide meaningful streamflow predictions at interior points. In addition, the results of the complementary study, which uses interior points during the model calibration, suggest that model performance at the outlet can be further improved by using a semi-distributed structure calibrated at both interior points and the outlet, even when only a few years of historical record are available. © 2009 Elsevier B.V.

• ERA5-Land has great potential for LSTM-based streamflow prediction across large and data-sparse catchments. • Multiscale attributes can substantially improve LSTM performance even for catchments with dams and reservoirs. • LSTM with multiscale attributes outperforms a global process-based hydrological model (LISFLOOD) in human-regulated basins. Streamflow simulation in human-regulated catchments is a great challenge for both process-based hydrological models and deep learning (DL) methods, mainly because human-regulation rules are difficult to parameterize in these models. In this study, we investigate the roles of river and catchment attributes in DL for streamflow prediction. We evaluate a typical DL method, i.e., long short-term memory (LSTM), and evaluate its performance in 25 large catchments across the Yellow River Basin where human activities are intensive, especially with large numbers of dams and reservoirs influencing streamflow processes. For the LSTM forcing data, we compare two forcing datasets: the Fifth Generation of European Reanalysis (ERA5-Land) and meteorological station-based data. The results show that the LSTM forced by ERA5-Land achieves improved performance, as its mean Kling–Gupta efficiency (KGE) is 0.21 relative to the mean KGE of 0.08 from the meteorological station forced LSTM. Integrating different types of hydrological attributes (catchment and river characteristics) can substantially improve LSTM performance even for catchments with dams and reservoirs. The river-reach attributes show the largest contribution to the LSTM model improvement. Moreover, LSTM with multiscale attributes outperforms a global process-based hydrological model (LISFLOOD) in the middle and lower reaches of the Yellow River Basin. Our study indicates that multiscale attributes are promising pivots for DL methods to improve streamflow prediction in human-regulated basins. [ABSTRACT FROM AUTHOR]

• The impact of satellite actual evapotranspiration assimilation on runoff simulations is assessed for 149 CAMELS basins. • The assimilation of GLEAM, REA-ET, and GLDAS ETa products shows relatively small improvements in runoff simulations. • The data assimilation using GLEAM ETa data yielded the best performance in runoff simulations. • The influence of multiple catchment characteristics is evaluated for the runoff simulation. Data assimilation (DA) techniques that integrate remotely sensed observations and in-situ measurements into hydrological models have been extensively implemented, but few studies have applied multisource actual evapotranspiration (ETa) products to evaluate its impacts on the DA performances through a large number of basins with different catchment conditions. This study performs a DA experiment to assess the influence of remotely sensed ETa products on streamflow simulations over 149 Catchment Attributes and Meteorology for Large-sample Studies basins across the contiguous United States, and the relationships between the DA results and multiple catchment characteristics are further analyzed. In this experiment, three satellite ETa products from the GLEAM, REA-ET, and GLDAS are assimilated into a conceptual rainfall-runoff model (i.e., HYMOD) using the ensemble Kalman filter. The results show that assimilating the three ETa products provides relatively small improvements compared to the open loop (OL) scenario in approximately 48%-58% of the study basins, and around 14%-19% of these basins exhibit an increase in Nash-Sutcliffe Efficiency over 0.01. Specifically, the GLEAM ETa shows the greatest improvements in runoff, followed by the GLDAS and REA-ET products. The OL performance plays a significant role in runoff improvements, greater improvements are achieved for basins with poor OL runoff simulations. The assimilation efficiency is revealed to be relevant to the accuracy of the satellite ETa products. While the catchment conditions such as the aridity index, mean slope, the fraction of forest, and catchment area are found to have a limited effect on the spatial pattern of ET assimilation performances. [ABSTRACT FROM AUTHOR]

The Zambezi River basin is the fourth largest basin in Africa and the largest in southern Africa, comprising 5% of the total area of the continent. The basin is extremely vulnerable to climate change effects due to its highly variable climate. The purpose of this study was to evaluate the impact of climate change on streamflow in one of the sub-basins, the Kabombo basin. The multi- global climate model projections were used as input to the Soil Water Assessment Tool (SWAT) hydrological model for simulation of streamflow under RCP 4.5 and RCP 8.5 climate scenarios. The model predicted an annual streamflow increase of 85% and 6% for high uncertainty and strong consensus, respectively, under RCP 8.5. The model predicted a slightly reduced annual streamflow of less than 3% under RCP 4.5. The majority of simulations indicated that intra-annual and inter-annual streamflow variability will increase in the future for RCP 8.5 while it will reduce for the RCP 4.5 scenario. The predicted high and moderate rise in streamflow for RCP 8.5 suggests the need for adaptation plans and mitigation strategies. In contrast, the streamflow predicted for RCP 4.5 indicates that there may be a need to review the current management strategies of the water resources in the basin.

A stochastic streamflow program was developed to simulate inflow to a large river from a network of gauged tributaries. The program uses historical streamflow data from major tributary gauges near their confluence with the main stem and combines them stochastically to represent spatial and temporal patterns in flood events. It incorporates seasonality, event basis, and correlation in flood occurrence and flood peak magnitude between basins. The program produces synchronous tributary inflow hydrographs, which when combined and routed, reproduce observed main stem hydrograph characteristics, including peak, volume, shape, duration, and timing. Verification of the program is demonstrated using daily streamflow data from primary tributary and main stem gauges in the Sacramento River basin, California. The program is applied to simulating flow at ungauged main stem locations, assessing risk in fluvial systems, and detecting bed level change.