Nash-Sutcliffe Efficiency Coefficient Values Research Articles

Application of a hybrid algorithm of LSTM and Transformer based on random search optimization for improving rainfall-runoff simulation

Flood forecasting using traditional physical hydrology models requires consideration of multiple complex physical processes including the spatio-temporal distribution of rainfall, the spatial heterogeneity of watershed sub-surface characteristics, and runoff generation and routing behaviours. Data-driven models offer novel solutions to these challenges, though they are hindered by difficulties in hyperparameter selection and a decline in prediction stability as the lead time extends. This study introduces a hybrid model, the RS-LSTM-Transformer, which combines Random Search (RS), Long Short-Term Memory networks (LSTM), and the Transformer architecture. Applied to the typical Jingle watershed in the middle reaches of the Yellow River, this model utilises rainfall and runoff data from basin sites to simulate flood processes, and its outcomes are compared against those from RS-LSTM, RS-Transformer, RS-BP, and RS-MLP models. It was evaluated against RS-LSTM, RS-Transformer, RS-BP, and RS-MLP models using the Nash–Sutcliffe Efficiency Coefficient (NSE), Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Bias percentage as metrics. At a 1-h lead time during calibration and validation, the RS-LSTM-Transformer model achieved NSE, RMSE, MAE, and Bias values of 0.970, 14.001m3/s, 5.304m3/s, 0.501% and 0.953, 14.124m3/s, 6.365m3/s, 0.523%, respectively. These results demonstrate the model's superior simulation capabilities and robustness, providing more accurate peak flow forecasts as the lead time increases. The study highlights the RS-LSTM-Transformer model's potential in flood forecasting and the advantages of integrating various data-driven approaches for innovative modelling.

Analysis of Meteorological Element Variation Characteristics in the Heilongjiang (Amur) River Basin

Located in the Heilongjiang (Amur) River in north-east Asia, spanning four countries, plays a crucial role as an international border river, and its meteorological changes significantly impact the variation in water resources in the basin. This study utilizes daily average temperature and precipitation data from 282 meteorological stations in the Heilongjiang (Amur) River Basin and its surrounding areas for the period 1980–2022. The analysis employs spatial interpolation, change point testing, and model construction prediction methods. The results indicate a significant increasing trend in both overall temperature and precipitation changes within the Heilongjiang (Amur) River Basin. At the spatial scale, the annual warming rate increases gradually from the southeastern coastal region to the northwestern plateau region, while the rate of precipitation increase decreases from the southern area towards its surroundings. Temporally, the warming amplitude during the growing season decreases gradually from east to west, and the trend in precipitation changes during the growing season aligns with the overall annual precipitation trend. During the non-growing season, the warming trend shows a decrease in the plains and an increase in the plateau, while precipitation increase concentrates in the central and southern plains, and precipitation decrease predominantly occurs in the northwestern plateau region. Temperature and precipitation change points occurred in the years 2001 and 2012, respectively. In precipitation prediction, the Long Short-Term Memory (LSTM) model exhibits higher accuracy, with R (Pearson correlation coefficient) and NSE (Nash-Sutcliffe efficiency coefficient) values approaching 1 and lower NRSME values. This study provides a research foundation for the rational development and utilization of water resources in the Heilongjiang (Amur) River Basin and offers valuable insights for research on climate change characteristics in large transboundary river systems.

Modeling the spatial dependence of quality data using distance-based inputs and a data-driven model

In the current study, which aimed to model groundwater quality, a method has been proposed to account for the spatial structure of the data in the modeling process. This method incorporates the distance between observation points and the estimation point as one of the model inputs. The Multilayer Perceptron (MLP) model was utilized, and the models were trained and tested using groundwater quality data obtained from 153 observation wells in a region in northwestern Iran. Two distinct datasets were used. The mean error (ME), R-squared, and the Nash-Sutcliffe efficiency coefficient (NSE) values for models with eight or more neighboring wells indicated satisfactory performance. For the first dataset and training phase, the models achieved ME values less than 15 dS/m, R-squared values ranging from 0.57 to 0.73, and NSE values ranging from 0.57 to 0.72. In the testing phase, the ME for the best-trained models was −0.2, highlighting the absence of underestimation or overestimation in the results. Additionally, the R-squared and NSE for the best model were calculated as 0.54 and 0.53, respectively. Similar results were achieved for the second dataset. It was also observed that incorporating the distance to the observation wells improves the outcomes compared to using only electrical conductivity (EC) as an input, and additionally, considering more neighboring wells positively affected the performance of the models and allowed for more precise salinity estimation. For the first dataset and based on the calculated metrics, models using 8, 16, or 20 neighboring wells, with EC and distance as inputs, demonstrated comparable performance. For the second dataset, the model with 12 neighboring wells was selected as the best model. The findings suggest that models with a specific number of neighboring wells and appropriate input variables can accurately estimate EC in groundwater systems.

An assessment of Sentinel‐1 synthetic aperture radar, geophysical and topographical covariates for estimating topsoil particle‐size fractions

AbstractData derived from synthetic aperture radar (SAR) are widely employed to predict soil properties, particularly soil moisture and soil carbon content. However, few studies address the use of microwave sensors for soil texture retrieval and those that do are typically constrained to bare soil conditions. Here, we test two statistical modelling approaches—linear (with and without interaction terms) and tree‐based models, namely compositional linear regression model (LRM) and random forest (RF)—and both nongeophysical (e.g., surface soil moisture, topographic, etc) and geophysical‐based (electromagnetic, magnetic and radiometric) covariates to estimate soil texture (sand %, silt % and clay %), using microwave remote sensing data (ESA Sentinel‐1). The statistical models evaluated explicitly consider the compositional nature of soil texture and were evaluated with leave‐one‐out cross‐validation (LOOCV). Our findings indicate that both modelling approaches yielded better estimates when fitted without the geophysical covariates. Based on the Nash–Sutcliffe efficiency coefficient (NSE), LRM slightly outperformed RF, with NSE values for sand, silt and clay of 0.94, 0.62 and 0.46, respectively; for RF, the NSE values were 0.93, 0.59 and 0.44. When interaction terms were included, RF was found to outperform LRM. The inclusion of interactions in the LRM resulted in a decrease in NSE value and an increase in the size of the residuals. Findings also indicate that the use of radar‐derived variables (e.g., VV, VH, RVI) alone was not able to predict soil particle size without the aid of other covariates. Our findings highlight the importance of explicitly considering the compositional nature of soil texture information in statistical analysis and regression modelling. As part of the continued assessment of microwave remote sensing data (e.g., ESA Sentinel‐1) for predicting topsoil particle size, we intend to test surface scattering information derived from the dual‐polarimetric decomposition technique and integrate that predictor into the models in order to deal with the effects of vegetation cover on topsoil backscattering.

Early estimation of 28-day compressive strength of mortars using regression and neural network-based models

Concrete, the most commonly used construction material, presents a challenge in determining its compressive strength (CS) after 28 days of manufacture. This study aims to predict the compressive strength of conventionally cured (CC) mortar specimens using data from accelerated microwave-cured (MC) specimens. Two hundred and seventy-six specimens with different mixes and admixtures were divided into two groups: one half was subjected to MC and the other to CC. The CSs of the CC specimens were determined after 28 days, while the MC specimens were subjected to additional tests such as ultrasonics pulse velocity and three-point bending after microwave curing. Regression and artificial neural network models (ANN) were developed using the obtained data. While the data of the specimens subjected to MC were used as independent variables of the developed models, the compressive strength of the conventionally cured specimens (CC) was used as the dependent variable. In the models, 60% of the data were used for training, 20% for validation, and 20% for testing. The predictive capabilities of the developed models were compared using performance statistics such as root mean square error (RMSE), root mean absolute error (MAE), and Nash-Sutcliffe efficiency coefficient (NSE). As a result of the study, it was found that compared to all models, the highest performance values were obtained by the ANN-based models. Among the regression-based models, the multivariate adaptive regression splines method (MARS) obtained the highest performance values. Considering the calculated NSE values, it was determined that the ANN-based model gave similar results to the MARS model for the training and validation data sets but provided performance values approximately 25% higher than the MARS model for the test data set. In addition, the results obtained from the modeling studies showed that the CSs of concrete could be predicted very accurately and well in advance.

Pitfalls in hydrologic model calibration in a data scarce environment with a strong seasonality: experience from the Adyar catchment, India

Process-based hydrologic models can provide necessary information for water resources management. However, the reliability of hydrological models depends on the availability of appropriate input data and proper model calibration. In this study, we demonstrate that common calibration procedures that assume stationarity of hydrological processes can lead to unsatisfactory model performance in areas that experience a strong seasonal climate. Moreover, we develop a more robust calibration procedure for the Soil and Water Assessment Tool (SWAT) in the Adyar catchment of Chennai, India. Calibration was carried out based on seasonal decomposition and by successively shifting the calibration period. Daily and monthly streamflow records were used to investigate how these different calibration procedures influence model parameterization. Results show that SWAT model performance improved when calibrated after separating the streamflow into wet and dry seasons. The wet season calibration increased the Kling Gupta Efficiency coefficient and Nash–Sutcliffe Efficiency coefficient values from 0.56 to 0.68 and 0.19 to 0.51, respectively, compared to calibration based on wet and dry seasons together. In addition, when calibration time periods were shifted, resultant sets of model parameter values and performance metrics differed. Calibration based on the 2004–2009 period resulted in an overestimation of streamflow by 8.2%, whereas the overestimation was 12.1%, 18.3%, and 20.0% for the 2004–2010, 2004–2011, and 2004–2012 periods, respectively. This study underlines that both the availability of observed streamflow data and the way these data are applied to calibration have a strong impact on model parameterization and performance.

Deep learning approach for one-hour ahead forecasting of weather data

ABSTRACT Weather is made up of multiple parameters, including solar radiation (SR), atmospheric pressure (AP), soil temperature (ST), atmospheric temperature (AT), wind speed (WS), relative humidity (RH), and sunshine duration (SD). These factors are also crucial for the renewable energy sector, solar simulation, agriculture, air pollution, water supply and distribution, avalanche warning, forestry, and town and regional planning. A deep learning method based on a neural network with Long Short-Term Memory (LSTM) was employed in this investigation for one-hour-ahead weather data forecasting. The ability of the LSTM model was compared with the Adaptive Neuro-Fuzzy Inference System (ANFIS) with that of the fuzzy c-means (FCM), Autoregressive Integrated Moving Average (ARIMA) model, and the Autoregressive Moving Average (ARMA) model. Mean absolute error (MAE), correlation coefficient (R), root means square error (RMSE), average bias, Nash – Sutcliffe efficiency coefficient (NSE), and mean absolute percentage error (MAPE) were selected as evaluation criteria. Results indicated that the proposed LSTM model presented good enough results compared to other used methods. 7 different types of meteorological data from a total of 4 years (35040 hours) were divided into 25% test data and 75% training data for the models. The best result was obtained for the hourly ST estimation of Adana province using the LSTM method, the MAE, RMSE, R, bias, NSE, and MAPE values were computed as 0.016°C, 0.078°C, 0.9999, −0.00018°C, 0.0805%, and 0.9999, respectively. On the other hand, the worst result was obtained for the hourly SD for Mardin province when ARIMA was used, and the statistical measures were derived as 0.128 hours for MAE, 0.215 hours for RMSE, 0.8851 for R, 0.00091 hours for bias, and 0.7657 for NSE. In this regard, it is demonstrated that the LSTM technique outperformed the other models in terms of all-weather data estimates and delivered highly sensitive outcomes.

A data‐driven approach for flood prediction using grid‐based meteorological data

AbstractEstablishing a physically‐based hydrological model for flood prediction in ungauged or data‐limited catchments has always been a difficult problem. In this study, a data‐driven approach based on the NASA Global Land Data Assimilation System (GLDAS) data is proposed for flood prediction with the assistance of the Gamma Test. A runoff generation model together with a routing model based on the most advanced deep learning model, i.e. Long Short‐Term Memory (LSTM) network is established. By calculating the noise of the input data, Gamma Test can effectively avoid the overfitting phenomenon with the LSTM network. Taking the a small‐scale mountainous catchment from northern China as an example, Gamma Test in this study is used to help select the optimal combination of the GLDAS inputs for the runoff generation model, and meanwhile the input grids involved in the routing model. The established models are then verified based on the Nash‐Sutcliffe Efficiency coefficient (NSE), and it is found that the surface runoff generation model shows a good performance, with an average NSE value of 0.8708, while the baseflow‐groundwater runoff generation model also results in an acceptable performance with an average NSE value of 0.6320. After the involvement of the observed rainfall data, the model performance (NSE) of the routing model has increased from 0.5738 to 0.7144. Finally, the runoff generation models and the routing model are integrated, and the grid‐based GLDAS meteorological data are used directly to simulate the streamflow at the catchment outlet. The integrated model performs well with an NSE value of 0.7909, which indicates the feasibility of this data‐driven approach for flood prediction using the gird‐based meteorological data. The methodology adopted in this study provides a reference for flood prediction using data‐driven models in ungauged or data‐limited catchments.

Understanding the Way Machines Simulate Hydrological Processes—A Case Study of Predicting Fine-Scale Watershed Response on a Distributed Framework

This study developed a Deep Neural Network (DNN) based distributed hydrologic model for an urban watershed in Republic of Korea. The developed model is composed of multiple Long Short-Term Memory (LSTM) hidden units connected by a fully connected layer. To examine the study area using the developed model, time series of 10-minute radar-gauge composite precipitation data and 10-minute temperature data at 239 model grid cells with 1km resolution were used as inputs to simulate 10-minute watershed flow discharge as an output. The model performed well for the calibration period (2013-2016) and validation period (2017-2019), with Nash-Sutcliffe Efficiency coefficient values being 0.99 and 0.67, respectively. Further in-depth analyses were performed to derive the following conclusions: (1) the map of runoff-precipitation ratios produced using the developed DNN model resembled imperviousness ratio map of the study area from the land cover data, revealing that the DNN successfully deep-learned the precipitation partitioning processes only with the input and output data without depending on any priori information about hydrology; (2) the model successfully reproduced the soil moisture dependent runoff process, an essential prerequisite of continuous hydrologic models; (3) each LSTM unit has different temporal sensitivity to the precipitation stimulus, with fast-response LSTM units having greater output weight factors near the watershed outlet, which implies that the developed model has a mechanism to separately consider the hydrological components with distinct response time such as direct runoff and the groundwater-driven baseflow.

Response of Runoff Yield to Land Use Changes in the Small Watershed of Core Area for 2022 Winter Olympic Games in Zhangjiakou City Based on SWAT Model

Based on an improved high-precision land use map, the SWAT model of the small watershed of the core area for the 2022 Winter Olympic Games in Zhangjiakou City was established. The model was set up, calibrated, and validated with data from 2017 to 2019. In comparison with the measured flow discharge, all the coefficient of determination values of the simulated results at the upstream, midstream, and downstream (outlet) points were larger than 0.80 in both calibration and validation periods, and the relevant Nash–Sutcliffe efficiency coefficient values were above 0.62. With the model, the response of runoff yield in the small watershed to the land use change before (2015) and after (2019) Olympic construction was simulated and analyzed. The runoff yield change was only −5.1% from 2015 to 2019, which was not statistically significant (p = 0.87 > 0.05). Through simulation of two extreme scenarios, two runoff yield change coefficients were introduced to explore the effects of land use change on runoff yield. The results revealed that the neutralization effects of the land use change from grassland to forest (with strong water conservation capacity) and residential and bare land (with poor water conservation capacity) were the main reasons for the insignificant change of runoff yield. The results of this research may provide some inspiration to the application of SWAT model in small watershed and some guidance to the vegetation restoration practices for water conservation. These results can promote understanding on the response of runoff yield to the land use changes towards two extreme directions of forest land and residential and barren land, and provide some guidance for the vegetation restoration practices in the small watersheds hosting major events.

High-resolution satellite images combined with hydrological modeling derive river discharge for headwaters: A step toward discharge estimation in ungauged basins

The reliance on ground-based measurements has been a long-standing barrier toward deriving river discharge for ungauged headwaters, which are quite sensitive to climate change due to relatively high contributions of snow and glacier meltwater to river discharge. In addition, information pertaining to discharge is scarce and inconsistent across the globe due to economic, geographic, and political reasons. Therefore, understanding spatiotemporal dynamics of river discharge in ungauged headwaters is extremely challenging, particularly in high-mountain regions such as the Tibetan Plateau (TP). This study presents a methodology of estimating daily continuous discharge using multisource remote sensing and a hydrological model that can simulate cryospheric processes including snow and glacier accumulation and melt. The hydrological model is forced and calibrated purely by satellite-observed information, particularly by river widths derived from high-spatial-resolution images without using in situ river discharge measurements. Poorly gauged or ungauged headwaters with river widths on the order of 100 m, including the Lhasa (the tributary of the upper Brahmaputra River), Salween, Mekong, and Yangtze rivers originating on the TP were chosen to test the methodology. At-a-section river widths were derived from high-resolution satellite images (i.e., IKONOS, QuickBird, RapidEye-1/2/4/5, GeoEye-1, and WorldView-2) and Landsat archives. As exemplified by the Nash–Sutcliffe efficiency coefficient values (NSE) (up to 0.8 and higher than 0.5 overall), simulated daily continuous discharge was reliable in the four testbeds with different river bathymetry, narrow river widths, and complex terrain. Uncertainties associated with the accuracy, number, and variability of calibration references (i.e., remotely sensed widths) and optimization algorithms were fully discussed. This study highlights the potential of deriving daily continuous discharge without a priori information on river flow, paving the way for discharge estimation in ungauged headwaters globally and providing important implications for satellite-based discharge estimation in narrow rivers across high-mountain regions.

Improving the accuracy of streamflow data acquired from the acoustic tomography technique using data despiking algorithms

Streamflow is an important hydrological factor in water resources management. The Fluvial Acoustic Tomography (FAT) as a cutting-edge river monitoring technology measures the flow velocity using the travel time of the acoustic signals. The velocity is then multiplied by the river cross-section to obtain accurate and continuous streamflow data. Similar to all acoustic instruments, the FAT output velocity contaminated by spike data. Standard deviation (STD) as a simple method of spike removing, was used in previous FAT studies. The disadvantages of the STD method are that it not only does not replace the identified spikes with the appropriate values and therefore results in measurement discontinuity, but it also requires engineering judgment that may increase the measurement error. In this study, two despiking methods, i.e., Phase-Space Thresholding (PST) and three-dimensional Rousseeuw Phase-Space thresholding (3D-RPS) are applied to detect/modify outliers of a 6-day FAT velocity signals. The results showed that the 3D-RPS method has the best performance with a Nash-Sutcliffe efficiency coefficient (NSC) of 0.731 compared to 0.572 and 0.44 NSC values of STD and PST methods, respectively. The FAT-based streamflow data were also compared with the reference streamflow obtained from the Rating-Curve method. The results reveal that the streamflow measurement accuracy improves significantly, where the relative errors decrease from ± 40% for the STD method to ± 15% for the 3D-RPS approach. As a result, the 3D-RPS approach appears to be an effective method for improving the accuracy of the FAT-acquired streamflow data.

Multi-step-ahead prediction of river flow using NARX neural networks and deep learning LSTM

Abstract Kelantan river (Sungai Kelantan in Malaysia) basin is one of the essential catchments as it has a history of flood events. Numerous studies have been conducted in river basin modelling for the prediction of flow and mitigation of flooding events as well as water resource management. Therefore, having multi-step-ahead forecasting for river flow (RF) is of important research interest in this regard. This study presents four different approaches for multi-step-ahead forecasting for the Kelantan RF, using NARX (nonlinear autoregressive with exogenous inputs) neural networks and deep learning recurrent neural networks called LSTM (long short-term memory). The dataset used was obtained in monthly record for 29 years between January 1988 and December 2016. The results show that two recursive methods using NARX and LSTM are able to do multi-step-ahead forecasting on 52 series of test datasets with NSE (Nash–Sutcliffe efficiency coefficient) values of 0.44 and 0.59 for NARX and LSTM, respectively. For few-step-ahead forecasting, LSTM with direct sequence-to-sequence produces promising results with a good NSE value of 0.75 (in case of two-step-ahead forecasting). However, it needs a larger data size to have better performance in longer-step-ahead forecasting. Compared with other studies, the data used in this study is much smaller.

Combined Wavelet Transform With Long Short-Term Memory Neural Network for Water Table Depth Prediction in Baoding City, North China Plain

Accurate estimation of water table depth dynamics is essential for water resource management, especially in areas where groundwater is overexploited. In recent years, as a data-driven model, artificial neural networks (NNs) have been widely used in hydrological modeling. However, due to the non-stationarity of water table depth data, the performance of NNs in areas of over-exploitation is challenging. Therefore, reducing data noise is an essential step before simulating the water table depth. This research proposed a novel method to model the non-stationary time series data of water table depth through combing the advantages of wavelet analysis and Long Short-Term Memory (LSTM) neural network (NN). A typical groundwater over-exploitation area, Baoding, North China Plain (NCP), was selected as a study area. To reflect the impact of anthropogenic activities, the variables harnessed to develop the model includes temperature, precipitation, evaporation, and some socio-economic data. The results show that decomposing the time series of the water table depth into three sub-temporal components by Meyer wavelets can significantly improve the simulation effect of LSTM on the water table depth. The average NSE (Nash-Sutcliffe efficiency coefficient) value of all the sites increased from 0.432 to 0.819. Additionally, a feedforward neural network (FNN) is used to compare forecasts over 12-months. As expected, wavelet-LSTM outperforms wavelet-FNN. As the prediction time increases, the advantages of wavelet-LSTM become more evident. The wavelet-LSTM is satisfactory for forecasting the water table depth at most in 6 months. Furthermore, the importance of input variables of wavelet-LSTM is analysed by the weights of the model. The results indicate that anthropogenic activities influence the water table depth significantly, especially in the sites close to the Baiyangdian Lake, the largest lake in the North China Plain. This study demonstrates that the wavelet-LSTM model provides an option for water table depth simulation and predicting areas of over-exploitation of groundwater.

Investigation of satellite precipitation product driven rainfall-runoff model using deep learning approaches in two different catchments of India

Abstract Rainfall–runoff models are valuable tools for flood forecasting, management of water resources, and drought warning. With the advancement in space technology, a plethora of satellite precipitation products (SPPs) are available publicly. However, the application of the satellite data for the data-driven rainfall–runoff model is emerging and requires careful investigation. In this work, two satellite rainfall data sets, namely Global Precipitation Measurement-Integrated Multi-Satellite Retrieval Product V6 (GPM-IMERG) and Climate Hazards Group Infrared Precipitation with Station (CHIRPS), are evaluated for the development of rainfall–runoff models and the prediction of 1-day ahead streamflow. The accuracy of the data from the SPPs is compared to the India Meteorological Department (IMD)-gridded precipitation data set. Detection metrics showed that for light rainfall (1–10 mm), the probability of detection (POD) value ranges between 0.67 and 0.75 and with an increasing rainfall range, i.e., medium and heavy rainfall (10–50 mm and >50 mm), the POD values ranged from 0.24 to 0.45. These results indicate that the satellite precipitation performs satisfactorily with reference to the IMD-gridded data set. Using the daily precipitation data of nearly two decades (2000–2018) over two river basins in India's eastern part, artificial neural network, extreme learning machine (ELM), and long short-time memory (LSTM) models are developed for rainfall–runoff modelling. One-day ahead runoff prediction using the developed rainfall–runoff modelling confirmed that both the SPPs are sufficient to drive the rainfall–runoff models with a reasonable accuracy estimated using the Nash–Sutcliffe Efficiency coefficient, correlation coefficient, and the root-mean-squared error. In particular, the 1-day streamflow forecasts for the Vamsadhara river basin (VRB) using LSTM with GPM-IMERG inputs resulted in Nash-Sutcliffe Efficiency Coefficient (NSC) values of 0.68 and 0.67, while ELM models for Mahanadhi river basin (MRB) with the same input resulted in NSC values of 0.86 and 0.87, respectively, during training and validation stages. At the same time, the LSTM model with CHIRPS inputs for the VRB resulted in NSC values of 0.68 and 0.65, and the ELM model with CHIRPS inputs for the MRB resulted in NSC values of 0.89 and 0.88, respectively, in training and validation stages. These results indicated that both the SPPs could reliably be used with LSTM and ELM models for rainfall–runoff modelling and streamflow prediction. This paper highlights that deep learning models, such as ELM and LSTM, with the GPM-IMERG products can lead to a new horizon to provide flood forecasting in flood-prone catchments.

Flood routing via a copula-based approach

Abstract Floods are among the most common natural disasters that if not controlled may cause severe damage and high costs. Flood control and management can be done using structural measures that should be designed based on the flood design studies. The simulation of outflow hydrograph using inflow hydrograph can provide useful information. In this study, a copula-based approach was applied to simulate the outflow hydrograph of various floods, including the Wilson River flood, the River Wye flood and the Karun River flood. In this regard, two-dimensional copula functions and their conditional density were used. The results of evaluating the dependence structure of the studied variables (inflow and outflow hydrographs) using Kendall's tau confirmed the applicability of copula functions for bivariate modeling of inflow and outflow hydrographs. The simulation results were evaluated using the root-mean-square error, the sum of squared errors and the Nash–Sutcliffe efficiency coefficient (NSE). The results showed that the copula-based approach has high performance. In general, the copula-based approach has been able to simulate the peak flow and the rising and falling limbs of the outflow hydrographs well. Also, all simulated data are at the 95% confidence interval. The NSE values for the copula-based approach are 0.99 for all three case studies. According to NSE values and violin plots, it can be seen that the performance of the copula-based approach in simulating the outflow hydrograph in all three case studies is acceptable and shows a good performance.

Large-scale monitoring of soil moisture using Temperature Vegetation Quantitative Index (TVQI) and exponential filtering: A case study in Beijing

In this study, to understand the variation trend of soil moisture in Beijing, surface soil moisture retrieval, deep soil moisture estimation and spatiotemporal distribution characteristics from 2013 to 2014 were analyzed. The results showed that the Temperature Vegetation Quantitative Index (TVQI), which was modified from the TVDI by the quantitative dry edge and wet edge, is effective for predicting surface soil moisture. The coefficient of determination (R2), root man squared error (RMSE), and Nash-Sutcliffe efficiency coefficient (NSE) of the TVQI increased by 0.2, 0.3%, and 4.6, respectively, compared to those of the TVDI. Moreover, the TVQI is closely related to cumulative precipitation within 1 month, as the correlation is the highest in this time interval. The accuracy of surface soil moisture retrieval was good, with relative error (RE), R2, and NSE values of 6%, 0.9 and 0.7, respectively. The deep soil moisture estimations performed well at both 20 cm and 40 cm, with an average R2 above 0.9, an average RE below 0.1 and an average NSE above 0.7. However, the prediction accuracy decreased as the soil depth increased. Topt, the best characteristic time length of the exponential filter, was significantly related to the soil bulk density, while the precipitation meteorological condition had little effect. Moreover, the average Topt of all stations can be used instead of the Topt of each station. Beijing, the study area, is dry throughout the year, with soil moisture reaching a minimum in May and a maximum in July. The changes in soil moisture at 10 cm, 20 cm and 40 cm were consistent, and soil moisture increased with soil depth. Furthermore, the variation trends of soil moisture were high in the east and low in the west. Not only is this study important for understanding soil moisture variation trends; it also provides a preliminary reference for water management in Beijing.

Uncertainty Analysis of SWAT Modeling in the Lancang River Basin Using Four Different Algorithms

The hydrological model is the primary tool for regional water resources management, allocation, and prediction. However, these models always suffer from large uncertainties from multiple sources. Therefore, it is necessary to conduct an uncertainty analysis before performing hydrological simulation. Sequential Uncertainty Fitting (SUFI-2), Parameter Solution (ParaSol), Generalized Likelihood Uncertainty Estimation (GLUE), and Particle Swarm Optimization (PSO) integrated with the SWAT-CUP software were used to calibrate the Soil and Water Assessment Tool (SWAT) model and quantify the parameter sensitivity and prediction uncertainty of the SWAT in the Lancang River (LR) Basin, which is located in the southwest of China. This model was calibrated and validated using the four algorithms both at the daily scale, and the optimal simulation results derived by the four methods showed that the SWAT model performed well over the Yunjinghong station with Nash–Sutcliffe efficiency coefficient (NSE) and coefficient of determination (R2) values greater than 0.8 both in the calibration (1975 to 1989) and validation (1990 to 2004) periods. Among the four algorithms, the ParaSol algorithm produced the best simulation result at the daily scale with NSE values of 0.89 and 0.90 for the calibration and validation periods, respectively. Furthermore, the ParaSol algorithm has the greatest proportion of simulations (94%) with an NSE greater than 0.5. Parameter sensitivity analysis results demonstrated that the four methods all can be used for parameter sensitivity analysis in streamflow simulation, and they all identified that the base flow factor for bank storage (ALPHA_BNK) and effective hydraulic conductivity in the main channel alluvium (CH_K2) were more sensitive. The uncertainty analysis of model parameters showed that the parameter 95PPU (95% prediction uncertainty) width yielded by the ParaSol algorithm was the smallest compared with that of the other methods, followed by PSO, SUFI-2, and GLUE. The uncertainty analysis of the model simulation indicated that the SUFI-2 and PSO methods can achieve satisfactory results (with P-factor > 0.7 and R-factor < 1.5) at the daily scale; among them, SUFI-2 (P-factor = 0.93, R-factor = 1.17) performed much better than PSO (P-factor = 0.78, R-factor = 1.14). In general, by comparing its evaluation criteria (NSE, R2, RE, P-factor, and R-factor) to other methods, ParaSol stood out as the most efficient tool for model calibration. However, SUFI-2 remains the most robust method to perform uncertainty analysis considering its uncertainties of model structure, model inputs, and parameters. This study provides insight into hydrological simulation of the LR Basin using the appropriate algorithm to calibrate the model and implement the uncertainty analysis.

Improving Runoff Simulation and Forecasting with Segmenting Delay of Baseflow from Fast Surface Flow in Montane High-Vegetation-Covered Catchments

Due to the complicated terrain conditions in montane catchments, runoff formation is fast and complicated, making accurate simulation and forecasting a significant hydrological challenge. In this study, the spatiotemporal variable source mixed runoff generation module (SVSMRG) was integrated with the long short-term memory (LSTM) method, to develop a semi-distributed model (SVSMRG)-based surface flow and baseflow segmentation (SVSMRG-SBS). Herein, the baseflow was treated as a black box and forecasted using LSTM, while the surface flow was simulated using the SVSMRG module based on hydrological response units (HRUs) constructed using eco-geomorphological units. In the case study, four typical montane catchments with different climatic conditions and high vegetation coverage, located in the topographically varying mountains of the eastern Tibetan Plateau, were selected for runoff and flood process simulations using the proposed SVSMRG-SBS model. The results showed that this model had good performance in hourly runoff and flood process simulations for montane catchments. Regarding runoff simulations, the Nash–Sutcliffe efficiency coefficient (NSE) and correlation coefficient (R2) reached 0.8241 and 0.9097, respectively. Meanwhile, for the flood simulations, the NSE ranged from 0.5923 to 0.7467, and R2 ranged from 0.6669 to 0.8092. For the 1-, 3-, and 5-h baseflow forecasting with the LSTM method, it was found that model performances declined when simulating the runoff processes, wherein the NSE and R2 between the measured and modeled runoff decreased from 0.8216 to 0.8087 and from 0.9095 to 0.8871, respectively. Similar results were found in the flood simulations, the NSE and R2 values declined from 0.7414–0.5885 to 0.7429–0.5716 and from 0.8042–0.6547 to 0.7936–0.6067, respectively. This means that this new model achieved perfect performance in montane catchment runoff and flood simulation and forecasting with 1-, 3-, 5-h steps. Therefore, as it considers vegetation regulation, the SVSMRG-SBS model is expected to improve runoff and flood simulation accuracy in montane high-vegetation-covered catchments.

Drought prediction using hybrid soft-computing methods for semi-arid region

Drought is one of the most significant natural disaster and prediction of drought is a key aspect in effective management of water resources and reducing the effect of a drought with preliminary studies plays significant role. In this study, we predicted one of the meteorological drought indices, the self-calibrated Palmer Drought Severity Index (sc-PDSI), values for Adana, Turkey. First, we used adaptive neuro fuzzy inference system (ANFIS) as a standalone technique to predict sc-PDSI. Second, we used empirical mode decomposition (EMD) as a pre-processing technique to decompose the sc-PDSI time series into the sub-series and applied ANFIS to each sub-series. Following the prediction, results are summed each other and final prediction of the hybrid EMD-ANFIS method is obtained. Within the scope of the study, 1, 3and 6-months lead time sc-PDSI values are predicted. We utilized the mean square error (MSE) and Nash–Sutcliffe efficiency coefficient (NSE) as performance indicators in order to perform statistical evaluation. For ANFIS, we obtained NSE = 0.52 and NSE = 0.17 for 3-month and 6-month lead times, respectively. Also, NSE values are obtained as 0.81 and 0.77 for the hybrid model in 3-month and 6-month lead time predictions, respectively. The results revealed that the hybrid EMD-ANFIS model outperforms the standalone ANFIS model. Also, the predicted and actual sc-PDSI series investigated according to the statistical distributions. At last, error histograms of both predicted and actual series are compared according to the Kolmogorov–Smirnov test and the p values are calculated. The results illustrated the predictions are statistically significant.