In this contribution, the authors present their preliminary investigations into modeling the rainfall–runoff generation relation in a large subtropical catchment (Jiulong River catchment) on the southeast coast of China. Previous studies have mostly focused on modeling the streamflow and water quality of its small rural subcatchments. However, daily runoff on the scale of the whole catchment has not been modeled before, and hourly runoff data are desirable for some oceanographic applications. Three methods are proposed for modeling streamflow using rainfall outputted by the Weather Research Forecast (WRF) model, calculated potential evaporation (PET), and land cover type: (i) a ridge regression model; (ii) NPRED-KNN: a nonparametric k-nearest neighbor model (KNN) employing a parameter selection method (NPRED) based on partial information coefficient; (iii) the Hydrological Simulation Program-Fortran (HSPF) model with an hourly time step. Results show that the NPRED-KNN approach is the most unsuitable method of those tested. The HSPF model was manually calibrated, and ridge regression performs no worse than HSPF based on daily verification, whilst HSPF can produce realist daily flow time series, of which ridge regression is incapable. The HSPF model seems less prone to systematic underprediction when replicating monthly-annual water balance, and it successfully replicates the baseflow index (the flow intensity) of the Jiulong River catchment system. Full article

Consistent streamflow forecasts play a fundamental part in flood risk mitigation. Population increase and water cycle intensification are extending not only globally but also among Pakistan’s water resources. The frequency of floods has increased in the last few decades in the country, which emphasizes the importance of efficient practices needed to adopt for various aspects of water resource management such as reservoir scheduling, water sustainability, and water supply. The purpose of this study is to develop a novel hybrid model for streamflow forecasting and validate its efficiency at the upper Indus basin (UIB), Pakistan. Maximum streamflow in the River Indus from its upper mountain basin results from melting snow or glaciers and climatic unevenness of both precipitation and temperature inputs, which will, therefore, affect rural livelihoods at both a local and a regional scale through effects on runoff in the Upper Indus basin (UIB). This indicates that basins receive the bulk of snowfall input to sustain the glacier system. The present study will help find the runoff from high altitude catchments and estimated flood occurrence for the proposed and constructed hydropower projects of the Upper Indus basin (UIB). Due to climate variability, the upper Indus basin (UIB) was further divided into three zone named as sub-zones, zone one (z1), zone two (z2), and zone three (z3). The hybrid models are designed by incorporating artificial intelligence (AI) models, which includes Feedforward backpropagation (FFBP) and Radial basis function (RBF) with decomposition methods. This includes a discrete wavelet transform (DWT) and ensemble empirical mode decomposition (EEMD). On the basis of the autocorrelation function and the cross-correlation function of streamflow, precipitation and temperature inputs are selected for all developed models. Data have been analyzed by comparing the simulation outputs of the models with a correlation coefficient (R), root mean square errors (RMSE), Nash-Sutcliffe Efficiency (NSE), mean absolute percentage error (MAPE), and mean absolute errors (MAE). The proposed hybrid models have been applied to monthly streamflow observations from three hydrological stations and 17 meteorological stations in the UIB. The results show that the prediction accuracy of the decomposition-based models is usually better than those of AI-based models. Among the DWT and EEMD based hybrid model, EEMD has performed significantly well when compared to all other hybrid and individual AI models. The peak value analysis is also performed to confirm the results’ precision rate during the flood season (May-October). The detailed comparative analysis showed that the RBFNN integrated with EEMD has better forecasting capabilities as compared to other developed models and EEMD-RBF can capture the nonlinear characteristics of the streamflow time series during the flood season with more precision. Full article

Rainfall during typhoons is one of the most important water resources in Taiwan, but heavy typhoon rainfall often leads to serious disasters and consequently results in loss of lives and property. Hence, accurate forecasts of typhoon rainfall are always required as important information for water resources management and rainfall-induced disaster warning system. In this study, a methodology is proposed for providing quantitative forecasts of 24 h cumulative rainfall during typhoons. Firstly, ensemble forecasts of typhoon rainfall are obtained from an ensemble numerical weather prediction (NWP) system. Then, an evolutionary algorithm, i.e., genetic algorithm (GA), is adopted to real-time decide the weights for optimally combining these ensemble forecasts. That is, the novelty of this proposed methodology is the effective integration of the NWP-based ensemble forecasts through an evolutionary algorithm-based strategy. An actual application is conducted to verify the forecasts resulting from the proposed methodology, namely NWP-based ensemble forecasts with a GA-based integration strategy. The results confirm that the forecasts from the proposed methodology are in good agreement with observations. Besides, the results from the GA-based strategy are more accurate as compared to those by simply averaging all ensemble forecasts. On average, the root mean square error decreases about 7%. In conclusion, more accurate typhoon rainfall forecasts are obtained by the proposed methodology, and they are expected to be useful for disaster warning system and water resources management during typhoons. Full article

Accurate modeling for nonlinear and nonstationary rainfall-runoff processes is essential for performing hydrologic practices effectively. This paper proposes two hybrid machine learning models (MLMs) coupled with variational mode decomposition (VMD) to enhance the accuracy for daily rainfall-runoff modeling. These hybrid MLMs consist of VMD-based extreme learning machine (VMD-ELM) and VMD-based least squares support vector regression (VMD-LSSVR). The VMD is employed to decompose original input and target time series into sub-time series called intrinsic mode functions (IMFs). The ELM and LSSVR models are selected for developing daily rainfall-runoff models utilizing the IMFs as inputs. The performances of VMD-ELM and VMD-LSSVR models are evaluated utilizing efficiency and effectiveness indices. Their performances are also compared with those of VMD-based artificial neural network (VMD-ANN), discrete wavelet transform (DWT)-based MLMs (DWT-ELM, DWT-LSSVR, and DWT-ANN) and single MLMs (ELM, LSSVR, and ANN). As a result, the VMD-based MLMs provide better accuracy compared with the single MLMs and yield slightly better performance than the DWT-based MLMs. Among all models, the VMD-ELM and VMD-LSSVR models achieve the best performance in daily rainfall-runoff modeling with respect to efficiency and effectiveness. Therefore, the VMD-ELM and VMD-LSSVR models can be an alternative tool for reliable and accurate daily rainfall-runoff modeling. Full article

Rainfall stations of a certain number and spatial distribution supply sampling records of rainfall processes in a river basin. Uncertainty may be introduced when the station records are spatially interpolated for the purpose of hydrological simulations. This study adopts a bootstrap method to quantitatively estimate the uncertainty of areal rainfall estimates and its effects on hydrological simulations. The observed rainfall records are first analyzed using clustering and correlation methods and possible average basin rainfall amounts are calculated with a bootstrap method using various combinations of rainfall station subsets. Then, the uncertainty of simulated runoff, which is propagated through a hydrological model from the spatial uncertainty of rainfall estimates, is analyzed with the bootstrapped rainfall inputs. By comparing the uncertainties of rainfall and runoff, the responses of the hydrological simulation to the rainfall spatial uncertainty are discussed. Analyses are primarily performed for three rainfall events in the upstream of the Qingjian River basin, a sub-basin of the middle Yellow River; moreover, one rainfall event in the Longxi River basin is selected for the analysis of the areal representation of rainfall stations. Using the Digital Yellow River Integrated Model, the results show that the uncertainty of rainfall estimates derived from rainfall station network has a direct influence on model simulation, which can be conducive to better understand of rainfall spatial characteristic. The proposed method can be a guide to quantify an approximate range of simulated error caused by the spatial uncertainty of rainfall input and the quantified relationship between rainfall input and simulation performance can provide useful information about rainfall station network management in river basins. Full article