Nash-Sutcliffe Coefficient Values Research Articles

Adaptive Parameter Estimation of the Generalized Extreme Value Distribution Using Artificial Neural Network Approach

Parameter estimation strategies have long been a focal point in research due to their significant implications for understanding data behavior, including the dynamics of big data. This study offers an advancement in these strategies by proposing an adaptive parameter estimation approach for the Generalized Extreme Value distribution (GEVD) using an artificial neural network (ANN). Through the proposed adaptive parameter estimation approach, based on ANNs, this study addresses the parameter estimation challenges associated with the GEVD. By harnessing the power of ANNs, the proposed methodology provides an innovative and effective solution for estimating the parameters of the GEVD, enhancing our understanding of extreme value analysis. To predict the flood risk areas in the Chi river watershed in Thailand, we first determine the variables that are significant in estimation of the three GEVD parameters μ,σ, and ξ by considering the respective correlation coefficient and then estimating these parameters. The data were compiled from satellite and meteorological data in the Chi watershed gathered from the Meteorological Department and 92 meteorological stations from 2010 to 2021, and consist of such variables as the Normalized Difference Vegetation Index (NDVI), climate, rainfall, runoff, and so on. The parameter estimation focuses on the GEVD. Taking into consideration that the processes could be stationary (parameters are constant over time, S) or non-stationary (parameters change over time, NS), maximum likelihood estimation and ANN approaches are applied, respectively. Both cases are modeled with the GEVD for the monthly maximum rainfall. The Nash-Sutcliffe coefficient (NSE), is used to compare the performance and accuracy of the models. The results illustrate that the non-stationary model was suitable for 82 stations, while the stationary model was suitable for only 10 stations. The NSE values in each model range from 0.6 to 0.9. This indicated that all 92 models were highly accurate. Furthermore, it is found that meteorological variables, geographical coordinates, and NDVI, that are correlated with the shape parameter in the ANN model, are more significant than others. Finally, two-dimensional maps of the return levels in the 2, 5, 10, 20, 50, and 100-year return periods are presented for further application. Overall, this study contributes to the advancement of parameter estimation strategies in the context of extreme value analysis and offers practical implications for water resource management and flood risk mitigation.

Daily suspended sediment concentration forecast in the upper reach of Yellow River using a comprehensive integrated deep learning model

The precise prediction of suspended sediment concentration (SSC) is of great importance for river reservoir construction planning, water resource management, and ecological environment restoration. This research aims to improve SSC prediction accuracy by constructing a comprehensive and integrated deep learning model Wavelet-MGGP-CNN-LSTM (ICNN-LSTM), combining wavelet transformation (WT), multi-gene genetic programming (MGGP), convolutional neural network (CNN), and long short-term memory (LSTM) simultaneously. In ICNN-LSTM, the WT decomposes the signal and extracts time and frequency domain information, while the MGGP filters out redundant information. Then, the CNN and LSTM are integrated in a parallel and loosely coupled manner to form an initial combined model CNN-LSTM (CNN combined with LSTM) to process filtered information by WT and MGGP. Furthermore, this study compares the performance of ICNN-LSTM with CNN, LSTM, CNN-LSTM, ICNN (CNN embedded with WT and MGGP), ILSTM (LSTM embedded with WT and MGGP), artificial neural network (ANN), and the traditional sediment rating curve (SRC). The evaluation of prediction accuracy for all models was conducted using root mean square error (RMSE), Nash-Sutcliffe coefficient (NSC), coefficient of determination (R2), and mean absolute error (MAE) as performance indicators. The daily discharge and suspended sediment concentration series data from Tangnaihai Hydrological Station in the upper reaches of the Yellow River spanning from 1977 to 1987 were selected to train and test the models. Results show that, first, deep learning networks such as CNN and LSTM outperform the shallow neural network ANN, with LSTM providing higher accuracy than CNN. Second, the CNN-LSTM hybrid outperforms both CNN and LSTM models, exhibiting a nearly 89% improvement in NSC value compared to SRC in the test phase. Third, deep learning models such as ICNN, ILSTM, and ICNN-LSTM show significantly higher NSC values than CNN, LSTM, and CNN-LSTM models in the test phase, with improvements of 13.8%, 5.7%, and 12.1%, respectively. Moreover, compared to SRC, the proposed ICNN-LSTM model improves NSC value by nearly 140% in the test phase. The proposed ICNN-LSTM model, integrating the advantages of WT, deep learning, and ensemble learning, provides accurate and reliable predictions and serves as a reference for time series prediction modeling.

Comparison of two hydro-sediment models based on energy principle of water movement in saturated/unsaturated soils

To study the impact of complex climatic conditions on runoff and sediment, physics-based distributed hydrological models are frequently employed for representation and analysis. In this work, two physics-based hydro-sediment models, namely CASC2D-SED (CASCade 2 Dimensional SEDiment) and GSSHA (Gridded Surface Subsurface Hydrologic Analysis), are used to simulate runoff and sediment in a semi-arid watershed on the Loess Plateau. Examining the degree to which the two models provide comparable or dissimilar outcomes with regard to runoff and sediment casting is an aspect of this investigation that is of utmost importance. Both the models are derived from the principle of mass and energy balance, and arrived at very similar conclusions regarding the infiltration of precipitation and the transport of silt in surface runoff. For typical flood events that occurred between 1981 and 2017, a comparison of modeled data with measured runoff and sediment is carried out. Both models achieved satisfactory results when runoff and sediment simulations were performed over three separate time periods, with average NSE (Nash–Sutcliffe coefficient) values of runoff and sediment are 0.81 and 0.70 respectively. The GSSHA model performs better in terms of the maximum and total amount of runoff and sediment, with an 6.8%, 8.6%, and 6.7% increase, respectively, in the average Rq (relative error of peak flow) Rd (relative error of runoff depth), and Rs (relative error of peak sediment values). In contrast, the CASC2D-SED model simulates better in terms of the Tq (time difference of peak discharge), and Ts (time difference of peak sediment), with the mean time difference of peak flow and sediment values being 0.27 and 1.02 higher, respectively. This shows that the CASC2D-SED model can simulate the timing of runoff and sediment more accurate.

Modelling and prediction of binder content using latest intelligent machine learning algorithms in carbon fiber reinforced asphalt concrete

In the present work, an attempt is made to find the most suitable prediction model for Marshall Stability and the optimistic Bitumen Content (BC) in carbon fiber reinforced asphalt concrete for flexible pavements by performing Marshall Stability tests. Further the prediction analysis is performed by taking the cognizance of the published research articles. Twofold approaches are adopted; first, to find the most suitable model to predict the Marshall Stability and second to obtain the optimum binder content responsible for the highest strength. Further, to find the most suitable model for closer prediction of Marshall Stability, eighteen input parameters i.e., Binder Content (BC with fifteen variations); 4.20%, 4.30%, 4.50%, 4.90%, 5.00%, 5.10%, 5.15%, 5.20%, 5.23%, 5.30%, 5.34%, 5.40%, 5.50%, 6.00%, 6.50%, and three others i.e., Carbon fiber, Bitumen grade and Fiber length are applied in the modelling algorithm. Five Machine learning techniques viz., Support Vector Machine, Gaussian Process, Random Forest, Random Tree, and M5P model were employed to find the most suitable prediction model. Seven statistical metrices i.e., Coefficient of correlation (CC), Mean absolute error (MAE), Root mean squared error (RMSE), Relative absolute error (RAE), Root relative squared error (RRSE), Willmott's index (WI), and Nash- Sutcliffe coefficient (NSE) were used to evaluate the performance of the applied models. After performing modelling analysis, it has been found that the Random Forest-based model is outperforming amongst all applied models with CC as 0.9735, MAE as 1.1755, RMSE as 1.5046, RAE as 25.68%, RRSE as 26.93%, WI values as 0.9351, and NSE values as 0.9272 in the testing stage. The Taylor diagram of the testing dataset also conforms to the results of RF-based model. The sensitivity analysis demonstrates that binder content (BC) of about 5.0% has a significant influence on the Marshall Stability in the asphalt mix used with carbon fibers.

Predicting hydraulic jump characteristics in a gradually expanding stilling basin with roughness elements by Sugeno Fuzzy Logic

Abstract The hydraulic behavior of a hydraulic jump can be challenging to estimate in order to design gradually expanding stilling basins with roughness elements on the bed. In this study, five dependent variables were identified that comprise: (i) the sequent depth ratio; (ii) the relative length of the jump; (iii) the relative roller length of the jump; (iv) the relative energy dissipation; and (v) the water surface profile. This study undertook a set of formulations based on the regression analysis and Sugeno Fuzzy Logic (SFL) to predict these variables based upon experimental data. Results demonstrate that the trained SFLs predicted the behavior of the dependent variables with the Nash–Sutcliffe Coefficient (NSC) greater than 0.96 in the testing phase. In contrast, the NSC values for the regression models are greater than 0.79. The higher accuracy of SFL is attributed to its capability in managing uncertainty and imprecise data owing to water surface profile oscillations of the hydraulic jump. Also, the results indicate that the prediction residuals for SFL are homoscedastic for all hydraulic parameters investigated except for the water surface profile, the prediction residuals of which for the regression equations are heteroscedastic.

Impact of riverbed topography on hydrology in small watersheds using Soil and Water Assessment Tool

In various watershed-based hydrological models, river topography information is generated by utilizing empirical expressions developed for large catchments. Therefore, identified the impact of channel geometry factors on hydrologic simulation in small watersheds and evaluation of reproducibility of river flow rate are required. We applied the Soil and Water Assessment Tool (SWAT) to small agricultural watersheds and compared measurements obtained from a river with DEM-generated values. The channel geometry factors determined using DEM were underestimated in comparison with measurements; however, using the measured channel geometry as the model input led to a 14.7% increase in the R2 and Nash–Sutcliffe coefficient (NSE) values in the daily average flow. In addition, the R2 and NSE values were more accurate by approximately 26% and 47% than the DEM-based optimized values obtained through SWAT-Calibration and Uncertainty Program (CUP)/Sequential Uncertainty Fitting ver.2 (SUFI-2) optimization, based on measured riverbed topography.

Spatio-temporal modeling of surface runoff in ungauged sub-catchments of Subarnarekha river basin using SWAT

In this present study, Soil and Water Assessment Tool (SWAT) embedded with ArcGIS interface has been used to simulate the surface runoff from the un-gauged sub-catchments in the upper catchment of Subarnarekha basin. Model calibration and validation were performed with the help of Sequential Uncertainty Fitting (SUFI-2) in-built in the SWAT-CUP package (SWAT Calibration Uncertainty Programs). The model was calibrated for a period from 1996 to 2008 with 3 years warm up period (1996-1998) and validated for a period of 5 years from 2009 to 2013. The model evaluation was performed by Nash - Sutcliffe coefficient (NSE), Coefficient of determination (R2) and Percentage Bias (PBIAS). The degree of uncertainty was evaluated by P and R factors. Basing upon the R2, NSE and PBIAS values respectively, of the order of 0.90, 0.90 and -12%, during calibration and 0.85, 0.83 and -15% during validation, substantiate performance of the model. All uncertainties of model parameters have been well taken by the P and R factors respectively, of the order of 0.95 and 0.77 during calibration and 0.82 and 0.87 during validation. The runoff generation from 19 sub-catchments of Adityapur catchment varies from 29.2-44.1% of the annual rainfall and average surface runoff simulated for the entire catchment is 545 mm. As the surface runoff generated in most of the sub-catchments amounts to above 30% of rainfall, it is recommended for adequate number of structural interventions at appropriate locations in the catchment to store the rainfall excess for providing irrigation, recharging groundwater and restricting the sediment and nutrient loss.

APPLICATION OF HYDROLOGICAL MODELS IN A SMALL AGRICULTURAL CATCHMENT FOR WATER MANAGEMENT: CASE STUDY ON WADI EL-RAML WATERSHED-NORTH WESTERN COAST OF EGYPT

North Western Coast of Egypt as a semi-arid region suffers from lack of rainfall most of the year except during rainfall events during the winter season which may even culminate with flash floods that causes acceleration of soil erosion and water losses resulting in degradation of the cultivation lands and consequently the agri-food productivity. This region Suffer lack of water reservoirs and terraces construction. Hydrological models are very useful tools for simulating the effect of natural processes and management practices on soil and water resources. Hence, the aim of study is calibrating and validating hydrological models KINEROS2(K2), ARCSWAT, to select the appropriate model and applying it to establish water and Soil conservation strategies such as terraces construction and building reservoirs. A yearly data set was used consisting of weather data, water content measurements and Gerlesh trough technique. Satellite image Digital elevation model (DEM), Land Cover classifications and soil map were used as layers input in models. The Nash–Sutcliffe coefficient (NSE) and coefficient of determination (R²) were used to evaluate model’s performance. The results reveal that the K2 are better than ARCSWAT, with acceptable NSE and R2 values, where the NSE, R2 were 0.86, 0.84 fork 2 and NSE, R2 were 0.55, 0.47 for ARCSWAT respectively. The K2 model was applied in sub-catchment where the estimated result of surface runoff was ranged from 21.92 to 169.695 m 3 ha-1 year-1. Hence, it is recommended that a reservoir can be constructed inside field approximately about 200 m3 ha-1 y-1 for using as Supplementary irrigation during dry period and the result indicated to areas with a high risk of soil loss that ranged from 38.88 to 3,860 kg. ha-1. year-1 .

Effect of sensitivity analysis on parameter optimization: Case study based on streamflow simulations using the SWAT model in China

Parameter optimization is an essential step in hydrological simulations, especially for solving practical problems. However, parameter optimization is usually intractable for complex models with a large number of parameters. In this study, a parameter optimization system based on Sensitive Parameter Combinations (SPCs) was developed, which comprised four parameter sensitivity analysis (SA) methods and a sensitive parameter optimization method. In particular, parameter SA was used to screen out the relatively sensitive parameters with significant impacts on the model output, and instead of using All Parameter Combinations (APCs), the SPCs were optimized with a global optimization method. This system was applied to the Soil and Water Assessment Tool (SWAT) model for daily streamflow simulation and monthly evaluation in four watersheds of China. The results showed that no more than 10 sensitive parameters were identified from 27 adjustable parameters for each watershed. In particular, four parameters (CN2, SOL_K, ALPHA_BNK, and SLSUBBSN) were relatively sensitive in all watersheds. Compared with optimizing APCs, despite the number of parameters was reduced by almost 2/3 in the optimization of SPCs, the accuracy was still very close (the maximum Nash–Sutcliffe coefficient (NSE) difference was 0.024 and the minimum difference was 0.002) and the optimization speed was doubled. In the comparison of monthly streamflow optimization, the SPCs were in good agreement with the APCs and had an obvious improvement for the default simulation. The NSE values of the SPCs optimization were greater than 0.88 during the calibration period in all watersheds and greater than 0.83 during the validation period in three watersheds. These findings indicate that optimizing the sensitivity parameters can greatly reduce the computational costs of SWAT streamflow simulations while ensuring their accuracy.

New Approach for Sediment Yield Forecasting with a Two-Phase Feedforward Neuron Network-Particle Swarm Optimization Model Integrated with the Gravitational Search Algorithm

Predicting sediment yield is an important task for decision-makers in environmental monitoring and water management since the benefits of applying non-linear, artificial intelligence (AI) models for optimal prediction can be far reaching in real-life decision support systems. AI-based models are considered to be favorable predictive tools since the nonlinear nature of suspended sediment data series warrants the utilization of nonlinear predictive methods for feature extraction, and for accurate simulation of suspended sediment load. In this study, Artificial Neural Network (ANN) approaches are employed to estimate the monthly sediment load where the two-phase Feed-forward Neuron Network Particle Swarm Optimization Gravitational Search Algorithm (FNN-PSOGSA) is developed, and then evaluated in respect to 3 distinct algorithms: the Adaptive Neuro-Fuzzy Inference System (ANFIS), Feed-forward Neuron Network (FNN) and the single-phase Feed-forward Neuron Network Particle Swarm Optimization (FNN-PSO). The study is carried out using the monthly rainfall, runoff and sediment data spanning a 10 year period (2000–2009) where about 75% of data are used in model training phase, 25% in testing phase. Three statistical performance criteria namely: the mean absolute error (MAE), Nash-Sutcliffe coefficient (NSE) and the Willmott’s Index (WI) and diagnostic plots visualizing the tested results are used to evaluate the performance of four AI-based models. The results reveal that the objective model (the two-phase FNN-PSOGSA model) and the single-phase FNN-PSO model yielded more precise results compared to the other forecast models. This result accorded to an NSE value of 0.612 (for the FNN-PSOGSA model) vs. an NS value of 0.500, 0.331 and 0.244 for the FNN-PSO, FNN and ANFIS models, and WI = 0.832 vs. 0.771, 0.692 and 0.726, respectively The study also demonstrated that the FNN model generated slightly better results than the ANFIS model for the estimation of sediment load data but overall, the two-phase FNN-PSOGSA model outperformed all comparison models. In light of the superior performance, this research advocates that the fully-optimized two-phase FNN-PSOGSA model can be explored as a decision-support tool for monthly sediment load forecasting using the rainfall and runoff values as the predictor datasets.

Simulation of Flow and Agricultural Non-Point Source Pollutant Transport in a Tibetan Plateau Irrigation District

Flow and transport processes in soil and rock play a critical role in agricultural non-point source pollution (ANPS) loads. In this study, we investigated the ANPS load discharged into rivers from an irrigation district in the Tibetan Plateau and simulated ANPS load using a distributed model. Experiments were conducted for two years to measure soil water content and nitrogen concentrations in soil and the quality and quantity of subsurface lateral flow in the rock and at the drainage canal outlet during the highland barley growing period. A distributed model, in which the subsurface lateral flow in the rock was described using a stepwise method, was developed to simulate flow and ammonium nitrogen (NH4+-N) and nitrate nitrogen (NO3−-N) transport processes. Sobol’s method was used to evaluate the sensitivity of simulated flow and transport processes to the model inputs. The results showed that with a 21.2% increase of rainfall and irrigation in the highland barley growing period, the average NH4+-N and NO3−-N concentrations in the soil layer decreased by 10.8% and 14.3%, respectively, due to increased deep seepage. Deep seepage of rainfall water accounted for 0–52.4% of total rainfall, whereas deep seepage of irrigation water accounted for 36.6–45.3% of total irrigation. NH4+-N and NO3−-N discharged into the drainage canal represented 19.9–30.4% and 19.4–26.7% of the deep seepage, respectively. The mean Nash–Sutcliffe coefficient value, which was close to 0.8, and the lowest values of root mean square errors, the fraction bias, and the fractional gross error indicated that the simulated flow rates and nitrogen concentrations using the proposed method were very accurate. The Sobol’s sensitivity analysis results demonstrated that subsurface lateral flow had the most important first-order and total-order effect on the simulated flow and NH4+-N and NO3−-N concentrations at the surface drainage outlet.

Estimation of soil temperature from meteorological data using different machine learning models

Soil temperature (Ts) plays a key role in physical, biological and chemical processes in terrestrial ecosystems. Accurate estimation of Ts at various soil depths is crucial for land-atmosphere interactions. This study investigated the applicability of four different machine learning models, extreme learning machine (ELM), generalized regression neural networks (GRNN), backpropagation neural networks (BPNN) and random forests (RF), for modeling half-hourly Ts at four different depths of 2 cm, 5 cm, 10 cm, and 20 cm on the Loess Plateau of China. A field experiment was conducted to measure half-hourly Ts and meteorological variables. Air temperature, wind speed, relative humidity, solar radiation, and vapor pressure deficit were used as inputs to train the models for estimation of half-hourly Ts. The results showed ELM, GRNN, BPNN and RF models provided desirable performance in modeling half-hourly Ts at all depths, with root mean square error values ranging 2.26–2.95, 2.36–3.10, 2.32–3.04 and 2.31–3.00 °C, mean absolute error values ranging 1.76–2.26, 1.83–2.31, 1.80–2.32 and 1.79–2.26 °C, Nash–Sutcliffe coefficient values ranging 0.856–0.930, 0.841–0.924, 0.847–0.927 and 0.850–0.927, and concordance correlation coefficient values ranging 0.925–0.965, 0.925–0.963, 0.928–0.963, and 0.924–0.961 for the ELM, GRNN, BPNN, and RF models, respectively. There was a statistically significant agreement (P < 0.001) between the measured and modeled values at both half-hour and daily timescales, and the box plots showed the distributional differences between the measured and modeled values were small. Generally, the ELM model had slightly better performance with much better computation speed than GRNN, BPNN as well as RF models at half-hourly timescales, thus the ELM model was highly recommended to estimate Ts at different soil depths.

Flow Routing and Surface Water Balance Simulation of a Tank Cascaded Catchment Using Coupled MIKE SHE/MIKE 11 Modeling

Hydrology of a tank cascaded basin in semiarid region is unique and different from a plain terrain hydrology as it consists of intervening surface storage structures called tanks. The incorporation of physical processes of irrigation tank cascaded system in hydrological modeling is essential for accurate estimation and simulation of its water balance components. The earlier studies of rainfall–runoff modeling in such catchments and the effects and processes of these tanks were lumped while modeling as it requires extensive data. Routing through the streams without considering the intervening tank storages will not give accurate estimation of surface water resource potential. Hence, an attempt has been made in this study to characterize the physical processes and to simulate the flow routing through an irrigation tank cascaded system with the application of coupled MIKE SHE and MIKE 11, a fully integrated physically based deterministic and distributed model. With the help of primary and secondary data, the model has been set up by coupling MIKE SHE and MIKE 11 model and the dynamics of water movement has been simulated. The model simulation results have been evaluated through calibration and validation of simulated tank yield with that of observed values. Water level at three tank outlets located at the upper, middle and tail reach of the tank cascaded catchment was used to study the model performance evaluation. The Nash–Sutcliffe coefficients obtained are 0.91, 0.94 and 0.92 for three tank water levels showing satisfactory goodness of fit between the observed and simulated water levels. The statistical parameters obtained from the validation process, such as error matrix, correlation coefficient and Nash–Sutcliffe coefficient values, indicate good simulation and prediction capability of the developed coupled model that incorporates a methodology which takes into account all the physical processes of the tank cascaded system.

Evaluating the Performance of Remotely Sensed Precipitation Estimates against In-Situ Observations during the September 2014 Mega-Flood in the Kashmir Valley

In the present study, 4 gridded satellite precipitation data products for September 2014 flood, IMERG (Integrated Multi-satellitE Retrievals for GPM), GSMaP (Global Satellite Mapping of Precipitation), TRMM-3B42 (Tropical Rainfall Measuring Mission) and INSAT-3D-IMR (INSAT Multispectral Rain), were evaluated against the Indian Meteorological Department rain-gauge data from Sep-1st to Sep-7th 2014. Three evaluation indices; Correlation coefficient (CC), the Relative bias (RB) and the Nash-Sutcliffe coefficient (NSC), were used to evaluate the robustness of satellite precipitation estimates with actual rainfall measurements. IMERG precipitation product has a near perfect positive CC and NSC values of 0.94 and 0.99 respectively; while the CC and NSC values are 0.7 and 0.5 for GSMaP_Gauge; 0.69 and 0.05 for INSAT-3D-IMR; and 0.9 and 0.8 for TRMM-3B42 respectively. The RB estimates indicate that IMERG, with a bias of 2%, is a best-fit dataset when compared to the surface rain-gauge observations. In contrast, TRMM-3B42, GSMaP and INSAT-3D-IMR have underestimation biases of −31%, −58%, and − 86% respectively. Analysis of the indices indicates that IMERG precipitation product performed better than other three satellite precipitation products owing to the closeness of values with surface gauge station data over Kashmir. Owing to scanty observation of rainfall in the region, IMERG has a potential to become a cost effective input data source for designing a flood early warning system (FEWS) for Kashmir. However, it is suggested to evaluate the robustness of different satellite-derived precipitation estimates compared to rain gauge observations by incorporating more extreme events from different mountain regions globally for establishing the best satellite derived precipitation product.

Estimating monthly evapotranspiration by assimilating remotely sensed water storage data into the extended Budyko framework across different climatic regions

The estimates of actual evapotranspiration (E) across basins and/or regions are still facing some difficulties because of the complex interactions amongst the components of the land-plant-atmosphere system, even though the basic physical mechanism of E is well understood. Presenting as a nonlinear relationship constrained by physical limits, the Budyko framework serves as a powerful tool used to estimate the averaged E at long-term scale. Given the ability of Gravity Recovery And Climate Experiment (GRACE) in effectively simulating the terrestrial water storage change (ΔTWS) at monthly scale, a model to estimate E was developed based on the Budyko framework with mean monthly parameter of multi-years, and was applied to the estimation of E for the 24 selected catchments in different climatic regions across China. Results indicate that for the majority of 24 catchments, the positive or negative trends in precipitation (P), potential evapotranspiration (E0), runoff (R) and ΔTWS were not statistically significant during 2003–2013 at both annual and monthly scales, but the ΔTWS occupied a large proportion in the partitioning of P into R and E at monthly scale. The monthly Budyko parameter showed large variation within the year and also across these 24 catchments. The Budyko-modeled monthly E (Budyko-E) represented the GRACE-derived ones (GRACE-E) across these catchments well with both modeled E volumes and hydrograph shapes which were also consistent with GRACE-E series except some underestimation for peak E. Overall, the Budyko-E represented GRACE-E in the arid and semi-arid catchments comparatively better than it in the humid and semi-humid catchments due to the larger proportion of ΔTWS in the P partition. Among all the 24 selected catchments, the monthly Budyko-type E model represented the GRACE-E across Shixiali catchment (located in the Hai River basin) best, with the correlation coefficient (r), Nash-Sutcliffe coefficient (NSCE) and relative error (RE) between GRACE-E and simulated results being 0.996, 0.988, −0.033 in calibration and 0.994, 0.951, −0.111 in validation, respectively. Furthermore, using GRACE-E as the benchmark values, the Budyko-E outperformed four other global E products from the newly published Global Land Data Assimilation System with Noah Land Surface Model-2 (GLDAS_E), MODIS (MODIS_E), Japanese 25-year reanalysis product (JRA_E), and Zhang’s method (Zhang_E). GLDAS_E was the best of four global E products at reproducing monthly variations, with relatively high r and NSCE values, small RE. While MODIS_E showed the poorest performance in reproducing the inter-annual E variations of these catchments, which was reflected by low r and NSCE values. This extended Budyko-E model can apply in other similar climatic regions and also may provide skill in evaluating the water resources since the preferable accuracy of E estimate.

SWAT manual calibration and parameters sensitivity analysis in a semi-arid watershed in North-western Morocco

Being a laborious approach, manual calibration of hydrologic model in a semi-arid context requires in-depth knowledge of the watershed and as much as possible field input data to obtain reliable simulations. In this study, manual calibration and relative sensitivity analysis approaches of the SWAT model (Soil and Water Assessment Tool) were applied for water balance in a 1993 km2 watershed (on the R’dom river) located in North-western Morocco. The watershed is located in a semi-arid area dominated by agro-forestry activities. The objectives of this study were (i) to perform a local sensitivity analysis of the SWAT model taking into consideration the watershed characteristics and (ii) to implement a detailed methodology of manual calibration and validation of the model in a semi-arid context. Sensitivity analysis has been carried out on 12 different SWAT input parameters, and has revealed that 4 input parameters only were the most influential ones on flow components of the R’dom watershed. Model manual calibration was conducted along 2006 and 2007 by comparing measured and predicted monthly and daily discharges and taking Nash-Sutcliffe coefficient (NSE), determination coefficient (R 2), and percent bias (PBIAS) as goodness-of-fit indicators. Validation has been performed by the same approach through 2008 and 2009 period. All final NSE values were above 0.5, R 2 values exceeded 0.7, and PBIAS lower than 25% demonstrating satisfactory model performances over the study watershed conditions. The SWAT model set-up with measured input data, manually calibrated and validated, reflects well the real hydrologic processes occurring in the R’dom watershed and can be used to assess current and future conditions and to evaluate alternative management practices.

Determination of Monthly Hydrological Erosion Severity and Runoff in Torogh Dam Watershed Basin Using SWAT and WEPP Models

Evaluation of runoff and sediment load is the main problem that affects the performance of dams due to the reduction in the storage capacity of their reservoirs and their effect on dam efficiency and operation schedule. Hydrologic models are increasingly used for simulation of spatially varied hydrologic processes with the availability of spatially distributed data to understand and manage natural and activities that affect watershed systems if the continuous field measurements are not available. Soil and water conservation and also quantification of soil loss in watershed basins are a significant issue. In this study, the Soil and Water Assessment Tool (SWAT) and the Water Erosion Prediction Project (WEPP) are applied to estimate runoff volume and sediment load for Torogh dam reservoir area that is located in Kashafrood Watershed Basin in northeastern Iran. Simulated and observed runoff and sediment load are compared with these models. In the calibration period, the Nash–Sutcliffe coefficient (NSE) values for the SWAT and WEPP were 0.698 and 0.854 for runoff, and 0.667 and 0.832 for sediment load, respectively. In the validation period, the NSE values for SWAT and WEPP were 0.678 and 0.824 for runoff, and 0.809 and 0.816 for sediment load, respectively. The results indicate that both models gave reasonable results in comparison with measured values. Simulation with WEPP model was better than SWAT in some cases and with reasonable confidence could be used for soil loss quantification in the watershed basin of Torogh dam.

Calibrating a flow model in an irrigation network: Case study in Alicante, Spain

The usefulness of models depends on their validation in a calibration process, ensuring that simulated flows and pressure values in any line are really occurring and, therefore, becoming a powerful decision tool for many aspects in the network management (i.e., selection of hydraulic machines in pumped systems, reduction of the installed power in operation, analysis of theoretical energy recovery). A new proposed method to assign consumptions patterns and to determine flows over time in irrigation networks is calibrated in the present research. As novelty, the present paper proposes a robust calibration strategy for flow assignment in lines, based on some key performance indicators (KPIF) coming from traditional hydrological models: Nash-Sutcliffe coefficient (non-dimensional index), root relative square error (error index) and percent bias (tendency index). The proposed strategy for calibration was applied to a real case in Alicante (Spain), with a goodness of fit considered as “very good” in many indicators. KPIF parameters observed present a satisfactory goodness of fit of the series, considering their repeatability. Average Nash-Sutcliffe coefficient value oscillated between 0.30 and 0.63, average percent bias values were below 10% in all the range, and average root relative square error values varied between 0.65 and 0.80.

Calibração hierárquica do modelo swat em uma bacia hidrográfica Catarinense

One of the main challenges of hydrological modeling is the effective calibration of model parameters. The aim of this study was to test a hierarchical calibration scheme for the Soil and Water Assessment Tool (SWAT) model by considering land use and climate changes. The study area was the Rio Negrinho watershed, located in Santa Catarina State, Brazil. Model calibration was performed with the Sequential Uncertainty Fitting (SUFI2) Algorithm through increasing complexity tests. The observed discharge data were first divided into calibration and validation periods(Split Sample Test) wherein the predicted monthly flow matched the measured values, with a Nash–Sutcliffe coefficient (NSE) ≥0.5 and Pbias<±25%. When wet and dry periods were considered separately (Differential Split-Sample Test) the NSE values of wet years (2008 a 2012) were higher than dry year values (2003 to 2007). In the SWAT validation for a watershed with similar land use (Proxy-Catchment Test), the model could not capture the flow peaks well (NSE of 0.0), however, the water balance volumes were compensated (Pbias of 0.1%). The case study demonstrates that SWAT performance varies greatly, depending on the spatial scale and calibration period. The SWAT model can be used in further studies to simulate land use and climate change scenarios however it must be calibrated and validated using spatially well distributed data and recorded series segment including dry and wet periods.

Prediction of dissolved oxygen in River Calder by noise elimination time series using wavelet transform

Prediction of dissolved oxygen (DO) plays an important role in water resources especially in surface waters such as rivers. The oxygen affects a vast number of other water indicators. In this study, the artificial neural network (ANN) and a hybrid wavelet-ANN (WANN) models were considered to predict thirty minutes dissolved oxygen in the River Calder at the Methley Bridge Station was located in the UK. For the proposed WANN model, the discrete wavelet transform (DWT) was linked to the ANN model for DO prediction. To achieve this aim, the original time series of thirty minutes DO and temperature (T) were decomposed in several sub-time series by DWT, and these new sub-series were imposed to the ANN model. The results were compared with single ANN model. The comparisons were done by some of the widely used relevant physical statistic indices. The Nash–Sutcliffe coefficient values were 0.998 and 0.740 for the WANN and ANN models, respectively. The model computed values of DO by the WANN model were in close agreement with respective measured values in the river water. Elimination noise by DWT model during pre-processing data is one of the abilities of the WANN model to better prediction. Since the results indicate closer approximation of the peak DO values by the WANN model, this model could be used for the simulation of cumulative DO data prediction in thirty minutes ahead.