Search

Water quality in irrigation areas is related to food security and many other national strategies. This study takes China’s Yingkou Irrigation District as the research object and analyzes 4 water quality indicators, including NH4+-N, TP, DO and COD, from 2017 to 2019 at 3 water quality monitoring points set up in the irrigation area. Simultaneously, a comprehensive pollution index is introduced to evaluate the quality of returned water in the irrigation area. Using Redundancy Analysis (RDA), the correlation between water quality indicators and Land-use and socio-economic indicators are analyzed separately. The results show that the water quality of the Yingkou Irrigation District varies greatly within and between years, and the months with poor water quality are concentrated in autumn and winter. This is mainly related to the geographical location of the irrigation area and the drainage cycle of rice planting. The quality of receding water shows a positive correlation with the water area and a negative correlation with the unused land. The RDA analysis with socio-economic indicators shows that PD and IP are the ones that have a more significant impact on the quality of return water in irrigation areas. The prevention and control of industrial point source pollution, as well as the interception of non-point source pollution in water regions, should be the focus of water pollution prevention and control in irrigation areas.

To further increase the forecast precision of non-stationary non-linear monthly runoff series and improve the effectiveness of pretreatment of monthly runoff series, the whale optimization algorithm (WOA) is introduced to optimize the variational mode decomposition (VMD), and the WOA-VMD-GRU prediction model is constructed by coupling with the gating cycle unit (GRU) neural network. First, the variation modal decomposition is optimized by the whale optimization algorithm, to find the best decomposition modal number k and penalty factor α, then several IMF components are obtained according to VMD processing runoff sequences; finally, results are obtained by adding those of each component. Taking Manwan Hydropower, Hongjiadu Hydropower, and Changshui hydrological station as examples, the BP model, the GRU model, the EMD-GRU model, the CEEMDAN-GRU model, and the VMD-GRU model are compared. Four quantitative indexes were used to estimate the model performance. The results show that the WOA-VMD-GRU model has the best prediction accuracy, with correlation coefficients and Nash coefficients above 0.99 and 0.97 in the prediction results of the three hydrological stations, respectively, and avoids the low efficiency of VMD decomposition parameters in manual trial computation, providing a new way for monthly runoff prediction.

Accurate and consistent annual runoff prediction in regions is a hot topic in the management, optimization, and monitoring of water resources. A novel prediction model (ESMD-SE-WPD-LSTM) is presented in this study. Firstly, the extreme-point symmetric mode decomposition (ESMD) is used to produce several intrinsic mode functions (IMF) and a residual (Res) by decomposing the original runoff series. Secondly, the sample entropy (SE) method is employed to measure the complexity of each IMF. Thirdly, we adopt wavelet packet decomposition (WPD) to further decompose the IMF with the maximum SE into several appropriate components and detailed components. Then the LSTM model, a deep learning algorithm based recurrent approach, is employed to predict all components obtained in the previous step. Finally, the forecasting results of all components are aggregated to generate the final prediction. The proposed model, which is applied to five annual series from different areas in China, is evaluated based on four quantitative indexes (R, NSEC, MAPE and RMSE). The results indicate that the ESMD-SE-WPD-LSTM outperforms other benchmark models in terms of four quantitative indexes. Hence the proposed model can provide higher accuracy and consistency for annual runoff prediction, making it an efficient instrument for scientific management and planning of water resources.

The Hanjiang-to-Weihe River Water Transfer Project is a systematic water supply project which involves the water supply system, the water storage system and the water conveyance system. The water storage system of the reservoir involves the water retaining system and the discharge system. The function of the discharge system is an important guarantee for the safety of the project. Regarding the Hanjiang-to-Weihe River Water Transfer Project’s discharge system risk control as the goal, on the basis of risk events, the fault tree analysis of the risk causes and Bayesian network theory, a system of Bayesian networks is established. A reliable risk rate calculation method is chosen based on the feature of the risk. The Hanjiang-to-Weihe River Water Transfer Project discharge system failure risk calculation theory system is built based on the Bayesian network inference ability. The engineering of the Sanhekou reservoir drain system is taken as the research object. Based on the Monte Carlo simulation and the method of the reliability derivation, the component failure risk is calculated and the Bayesian network is constructed to compute the risk of system failure rate. The scenario analysis is used to judge the main risk factors that affect the drain system security. If the risk rate of floating debris and leakage increases by 50%, the rate of failure risk of the discharge system would increase by 1.35 × 10–5. If the risk rate of the discharge system is reduced to 0, the risk rate of floating objects and leakage should have to be reduced by 1.35 × 10–5. If the discharge system fails, the failure risk rate of floating objects and leaks is 1.82 × 10–1. It shows that the theory of Bayesian network system can dynamically identify, evaluate, reduce and control the risk of discharge system, which is of guiding significance for risk management of inter-basin water transfer projects.

Drought generally has significant impacts on crops. It is essential to quantitatively evaluate the relationship between crop production and drought degree to provide technical support for drought disaster prevention. In this study, a drought degree index that can reflect the changes in precipitation, evapotranspiration, and soil moisture was developed on the basis of crop yield reduction rate. Four drought scenarios were set up to simulate the effects of meteorological drought on drought degree of crops at different growth stages. A cusp catastrophe model was constructed to analyze the mutation characteristics of the drought degree of maize at different growth stages under different meteorological drought conditions. Xi'an City in China was selected as the study area, and summer maize was selected as the research crop. Precipitation and crop yield data from 1951 to 2010 were used as the fundamental data to investigate drought degree mutation of summer maize. The results show that, under the meteorological drought conditions at the emergence-jointing stage, drought degree may change abruptly, and soil moisture content at the sowing-emergence, jointing-tasseling, and tasseling-mature stages should be kept higher than 39%.

Yin–Yang-pair Optimization (YYPO) is a recently developed philosophy-inspired meta-heuristic algorithm, which works with two main points for exploitation and exploration, respectively, and then generates more points via splitting to search the global optimum. However, it suffers from low quality of candidate solutions in its exploration process owing to the lack of elitism. Inspired by this, a new modified algorithm named orthogonal opposition-based-learning Yin–Yang-pair Optimization (OOYO) is proposed to enhance the performance of YYPO. First, the OOYO retains the normalization operation in YYPO and starts with a single point to exploit. A set of opposite points is designed by a method of opposition-based learning with split points generated from the current optimum for exploration. Then, the points, i.e., candidate solutions, are constructed by the randomly selected split point and opposite points through the idea of orthogonal experiment design to make full use of information from the space. The proposed OOYO does not add additional time complexity and eliminates a user-defined parameter in YYPO, which facilitates parameter adjustment. The novel orthogonal opposition-based learning strategy can provide inspirations for the improvement of other optimization algorithms. Extensive test functions containing a classic test suite of 23 standard benchmark functions and 2 test suites of Swarm Intelligence Symposium 2005 and Congress on Evolutionary Computation 2020 from Institute of Electrical and Electronics Engineers are employed to evaluate the proposed algorithm. Non-parametric statistical results demonstrate that OOYO outperforms YYPO and furnishes strong competitiveness compared with other state-of-the-art algorithms. In addition, we apply OOYO to solve four well-known constrained engineering problems and a practical problem of parameters optimization in a rainstorm intensity model.

Reservoirs play a key role in many infrastructure functions for people like flood control, irrigation, and water supply. In this work, we focused on the water quality evaluation model for Shimen Reservoir. Based on the monthly changes of factors such as pH, nitrate, ammonia nitrogen (NH3-N) and total nitrogen (TN) in 2013 and 2014, the information diffusion theory and fuzzy neural network technology were utilized to evaluate the water quality comprehensively. The probability distribution of these four factors in the reservoir was analysed and the water quality of the reservoir evaluated. The results show its reliability and these two methods can provide a basis for water quality control of Shimen Reservoir. Furthermore, the methods can be universally applied to the analysis and research of water quality in other regions.

The nitrogen concentration/flux of water is very sensitive to the response of Net Anthropogenic Nitrogen Inputs (NANI), but the research on the dynamic response of nitrogen concentration to NANI is still rarely reported. In this study, the source of Zhaoshandu water source in Zhejiang Province, China, was used as the research object. The dynamic response of total nitrogen (TN) concentration to NANI in the basin was quantitatively analysed from 2005 to 2014. The results show that the NANI of the water source has a growing trend. It has increased by 6.73% in ten years. The average annual average of NANI is 85.76 kg (hm2·a-1), which is 1.7 times the national average of 2009. The average contribution rate of atmospheric nitrogen, chemical fertilizer nitrogen, food and feed nitrogen and crop fixed nitrogen is 40.98%, 34.06%, 20.25% and 4.7% respectively. The spatial difference of NANI is large, showing an increasing trend from the upstream mountainous area along the downstream. The sub-basin 2 is the key source of nitrogen pollution in this water source. NANI is the only independent variable of the TN concentration regression model with a variance interpretation rate of 55.4%. In the future, it is necessary to reduce the nitrogen input intensity of nitrogen, food and feed for chemical fertilizers as the main target, and to achieve the goal of water quality improvement by reducing the NANI intensity in the basin.

River ecosystem is one of the most important ecosystems, and it provides many ecosystem services for human beings. However, river health has also been damaged by over-exploitation and water pollution. In the process of reservoir operation, the ecological flow demand of rivers should be fully considered and multi-objective ecological dispatch of reservoirs should be implemented. On basis of the traditional particle swarm optimization (PSO), the improved PSO with adaptive random inertia weights (ARIW) is proposed to deal with the problem of ecological optimal operation of the reservoir in the paper. According to the evolutionary process, based on the probability distribution density function of triangle, the inertia weight can be adjusted randomly and adaptively to meet the global or local optimization requirements. By typical mathematical function, the improved PSO algorithm is compared with traditional PSO and genetic algorithm (GA) and is proved to be more efficient and accurate. Taking Panjiakou Reservoir on the mainstream of Luanhe River in China as an example, the multi-objective ecological optimal dispatch of the reservoir has been analysed and calculated with the improved PSO algorithm under different targets years, considering flood control, water supply, and ecological demand. The research results can provide reference for developing rationally Luanhe River water resources, and making scientifically ecological dispatch plan of the Panjiakou reservoir.

Reservoir inflow prediction is subject to high uncertainties in data-driven modelling. In this study, a decomposition scheme is proposed to evaluate the individual and combined contributions of uncertainties from input sets and data-driven models to the total predictive uncertainty. Six variables (i.e., inflow (Q), precipitation (P), relative humidity (H), minimum temperature (Tmin), maximum temperature (Tmax) and precipitation forecast (F)), and three data-driven models (i.e., artificial neural network (ANN), support vector machine (SVM), and adaptive neuro fuzzy inference systems (ANFIS)) are used to produce an ensemble of 10-day inflow forecast for Huanren reservoir in China, and the analysis of variance (ANOVA) method is employed to decompose the uncertainty. The ensemble forecast results show that when the three variables, i.e., Q, P and F, are used only, the predictive accuracy of the data-driven models is very high and the addition of the other three variables, i. e., H, Tmin and Tmax, can slightly improve the predictive accuracy. The decomposition results indicate that the input set is the dominant source of uncertainty, the contribution of the data-driven model is limited and has a strong seasonal variation: larger in winter and summer, smaller in spring and autumn. Most importantly, the interactive contribution of the input set and the data-driven model to the total predictive uncertainty is very high and is more significant than the individual contribution from the model itself, implying that the combined effects of the input set and the data-driven model should be carefully considered in the modelling process.

To reveal the influence process of land use changes on runoff variation trends, this paper takes the Luojiang River of China as the study area, and the Soil and Water Assessment Tool (SWAT) model was constructed to quantitatively analyze the impact of different land uses on runoff formation in the watershed, and used the Cellular Automata-Markov (CA-Markov) model to predict future land use scenarios and runoff change trends. The results show that: (1) the SWAT model can simulate the runoff in the Luojiang River basin, (2) the runoff in the Luojiang River basin has a decreasing trend in recent 10 years, caused by the decrease of rainfall and runoff due to changes in land use, (3) the forecast shows that the land-use changes in the basin will lead to an increase in runoff coefficient in 2025. The increase of the runoff coefficient will bring some adverse effects, and relevant measures should be taken to increase the water storage capacity of urban areas. This study can help plan future management strategies for the study area land coverage and put forward a preventive plan for the possible adverse situation of runoff variation.

Understanding the relation between land-use types and baseflow mean response time (BMRT) is important to explore the response mechanism of baseflow processes in watersheds. BMRT was determined using an instantaneous unit hydrograph. The instantaneous unit hydrograph parameters were estimated by autocorrelation functions. The relative importance of land-use types in determining BMRT dynamics was assessed by hydrological model and partial least-squares regression. Our study suggests greater effects of urban area on BMRT than the effects of forest and agricultural land. This may be because the urban interception impervious area may impede baseflow generation over a short timescale. The effects of agricultural land are greater than those of forest in areas with steeper hillslopes, but lower than those of the forest in areas with more plains, reflecting the varied ability of forest and agricultural lands with different topography to hinder overland flow. Variations of BMRT are strongly linked to land use in the watershed. Overall, our study provides insight into the BMRT and dominant factors of land-use types in watersheds, planning of sustainable water resource use, and ecological protection in watersheds.

Drought projection is of critical significance for designing long-term drought adaptation strategies to cope with changing climate. This study presents an application of multi-class models for characterizing future droughts in the Yellow River basin (YRB). Using meteorological observations and simulations of three CMIP5 climatic models as input into VIC hydrologic model, the standardized Palmer drought index-based joint drought index (SPDI-JDI) is calculated to investigate spatiotemporal patterns and possible changes of future drought in the YRB, through an approach of bivariate modeling and analysis. It is concluded that moderate drought will be alleviated with decreased duration and severity, but extreme drought risk is likely to increase for future climate scenarios (2021–2050) relative to baseline period (1961–2000). From the perspective of bivariate analyses, the concurrence probability of drought events with extreme duration and severity might rise in the future, while increasing variability and heterogeneity of duration, severity and bivariate joint/concurrent return period would also increase the stochastic occurrence of extreme drought over time and space. These findings can provide insights into projection of climate change impacts on drought trends for the next 30 years, and help to preparations for strict control of socio-economic water consumption as well as eco-environmental protection and restoration in the YRB.

The objective of the curve-fitting method is to determine the optimal distribution by parameter estimation. The selection of the parameter estimation methods and the determination of the parameter ...

Recently, the booming electricity demand and intermittent energy has sharply increased the peak shaving pressure in China. However, for a majority of regional power grids in China, the installed capacity of flexible energy (like hydropower and pumped-storage) is small and the thermal plants are asked to respond the sudden load change at peak periods. The existing method based on specialist experience and historical information may generate inferior solutions and fail to reduce the peak pressure. Thus, a practical mixed integer linear programming (MILP) model is developed for unit commitment of thermal plants with peak shaving operation aspect in regional power grid lack of flexible hydropower energy, where the goal is chosen to minimize the peak-valley difference of the residual load series obtained by subtracting all the thermal generation from the original load curve while satisfying the necessary physical constraints. The MILP model is used to the thermal system in the China’s largest regional power grid, East China Power Grid. The simulations show that the MILP model can effectively smooth the residual load curve by gathering power generation of thermal plants at peak periods. Therefore, an alternative tool is provided to alleviate the peak pressure of thermal-dominant regional power grid in China.

The Latent Heat Flux (LE) is an important component of surface water heat transfer and hydrological cycle, and monitoring it is of great value for water resource management and crop water demand estimation. The Heihe River Basin has complex topography, which ensures better variable control in LE analysis. In this paper, the time series analysis and statistics of LE under different underlying surface conditions in summer were carried out by using the eddy correlation observation data in the Heihe River Basin, and the regression factors were analyzed. The results show that when the underlying surface types are greatly different, there are obvious differences in the daily distribution of LE, the daily variation trend of LE and the influencing factors. The range of diurnal distribution of LE in dune, Gobi and desert from −50 W/m2 to 100 W/m2. The diurnal LE distribution of vegetable fields, cornfields and wetlands were about 55% concentrated between −50 W/m2 and 100 W/m2. Temperature and carbon dioxide concentration (CO2) are the dominant factors affecting latent heat flux. Further analysis of temperature and CO2 is carried out by stepwise regression analysis, and multiple regression models are established. In terms of correlation and confidence, the results are better than the single factor fitting, which can better reflect the synergistic effect of temperature and CO2 on LE.

Based on the mine water produced by mining, in order to improve the ecological environment, the optimal allocation of mine water resources is studied. In order to reduce the uncertainty of the calculation results of ecological water demand, the wolf colony algorithm neural network model is used for long-term rainfall forecast. Combined with the forecast annual rainfall, the ecological water demand is classified and calculated. The results show that the ecological water demand based on rainfall forecast can reduce the allocation of water resources in wet years to ecological aspects, so that the surplus water resources can be used in industries, irrigation and other aspects that can create economic benefits, and improve the utilization efficiency of water resources. The ecological allocation model of mine water based on long-term rainfall forecast can reduce the uncertainty of regional water resources allocation based on rainfall forecast, which has good guiding significance and practical value for the optimal allocation of water resources in arid and water shortage areas.

In the optimum curve-fitting method, due to the dissimilar purposes, the discrepant accuracy and positions of the experience points, the importance of the points should be different. For the limited sample size of the hydrologic sequence, there are sampling errors in the parameter estimation. In order to focus on the important points and reduce the errors effectively, the weight has been introduced in the optimum curve-fitting method. The existing weighted optimum curve-fitting methods are analyzed and studied. The Fuzzy Weighted Optimum Curve-fitting Method (FWOCM), which are the limited nomograph length and the determination of the membership degree function without the premise of a large sample. In order to solve the problems, the improvement of the method should be conducted. A new membership degree function is deducted and demonstrated on the premise that the hydrologic sequence is a large sample. The Monte Carlo statistical test optimum curve-fitting method is used to extend the nomograph to the entire frequency range. The improved FWOCMs are tested by the ideal data and the real data. In order to evaluate the performances of the improved FWOCMs, the selected excellent method and the improved percentage method are introduced to analyze the relative errors. The results show that the extension of the nomograph and the new membership degree function to a certain extent weakens the impact of the shorter hydrologic sequence on the curve-fitting. It indicates that the effect of the improved optimum curve-fitting methods is satisfying and can be used in the engineering practice.

The Palmer drought severity index (PDSI) is physically based with multivariate concepts, but requires complicated calibration and cannot easily be used for multiscale comparison. Standardized drought indices (SDIs), such as the standardized precipitation index (SPI) and standardized precipitation evapotranspiration index (SPEI), are multiscalar and convenient for spatiotemporal comparison, but they are still challenged by their lack of physical basis. In this study, a hybrid multiscalar indicator, the standardized Palmer drought index (SPDI), was used to examine drought properties of two meteorological stations (the Beijing and Guangzhou stations) in China, which have completely different drought climatologies. The results of our case study show that the SPDI is correlated with the well-established drought indices (SPI, SPEI, and PDSI) and presents generally consistent drought/wetness conditions against multiple indicators and literature records. Relative to the PDSI, the SPDI demonstrates invariable statistical characteristics and better comparable drought/wetness frequencies over time and space. Moreover, characteristics of major drought events (drought class, and onset and end times) indicated by the SPDI are generally comparable to those detected by the PDSI. As a physically-based standardized multiscalar drought indicator, the SPDI can be regarded as an effective development of the Palmer drought indices, providing additional choices and tools for practical drought monitoring and assessment. Keywords: Drought, PDSI, Multiscalar index, SPDI, Multifaceted comparison

The river ecosystem has the characteristics of the coexistence of clarity and ambiguity. The starting point of eco-friendly reservoir operation is to fully consider the ecological water requirements of the lower reaches of the reservoir, so it also has the characteristics of clarity and vagueness. The fuzzy theory is an excellent tool to realize the quantification of fuzzy concepts. In this paper, the variable set theory (VS) is introduced into the decision-making field of eco-friendly reservoir scheduling scheme optimization. Taking Ertan Hydropower Station as an example, the scheduling scheme optimization is carried out. To verify the effectiveness of the evaluation method, this paper compares and analyzes the evaluation results of the fuzzy set evaluation method and the composite fuzzy matter-element method. The results show that the variable fuzzy set method has the advantages of rigorous theory, a concise model, and simple calculation, and the decision result is reasonable and reliable. This research can provide new ideas for the decision-making research of engineering.

Accurate prediction of extreme flood peak discharge is essential in developing the best management practices to avoid and reduce flood disaster. In recent years, many techniques have been pronounced as a branch of computer science to model wide range of hydrological process. Nevertheless, exploration of more efficient technique is necessary in terms of accuracy and applicability. In this study, a novel hermite-PPR model with SSO and LS algorithm is proposed for designing annual maximum flood peak discharge forecasting model at Yichang station on Yangtze River in China. The statistical properties of the data series are utilized for identifying an appropriate input vector to the model and then the performance of the proposed models were compared with adaptive neuro-fuzzy inference system (ANFIS), artificial neural network (ANN) and multiple linear regression (MLR) methods in terms of root mean squared error (RMSE), mean absolute relative error (MARE), coefficient of correlation (CC), Nash-Sutcliffe efficiency coefficient (NSEC) and qualified rate (QR). The results indicate that the presented methodology in this research can obtain significant improvement in forecasting accuracy in terms of different evaluation criteria during training and validation phases.

Understanding the response of flood frequency to impact factors could help water resource managers make better decisions. This study applied an integrated approach of a hydrological model and partial least squares (PLS) regression to quantify the influences of rainfall and forest landscape on flood frequency dynamics in the Upper Honganjian watershed (981 km2) in China. The flood events of flood seasons in return periods from two to 100 years, wet seasons in return periods from two to 20 years, and dry seasons in return periods from two to five years show similar dynamics. Our study suggests that rainfall and the forest landscape are pivotal factors triggering flood event alterations in lower return periods, that flood event dynamics in higher return periods are attributed to hydrological regulations of water infrastructures, and that the influence of rainfall on flood events is much greater than that of land use in the dry season. This effective and simple approach could be applied to a variety of other watersheds for which a digital spatial database is available, hydrological data are lacking, and the hydroclimate context is variable.

Hydrological time series forecasting is one of the most important applications in modern hydrology, especially for the effective reservoir management. In this research, an artificial neural network (ANN) model coupled with the ensemble empirical mode decomposition (EEMD) is presented for forecasting medium and long-term runoff time series. First, the original runoff time series is decomposed into a finite and often small number of intrinsic mode functions (IMFs) and a residual series using EEMD technique for attaining deeper insight into the data characteristics. Then all IMF components and residue are predicted, respectively, through appropriate ANN models. Finally, the forecasted results of the modeled IMFs and residual series are summed to formulate an ensemble forecast for the original annual runoff series. Two annual reservoir runoff time series from Biuliuhe and Mopanshan in China, are investigated using the developed model based on four performance evaluation measures (RMSE, MAPE, R and NSEC). The results obtained in this work indicate that EEMD can effectively enhance forecasting accuracy and the proposed EEMD-ANN model can attain significant improvement over ANN approach in medium and long-term runoff time series forecasting.

Hydrological time series forecasting is one of the most important applications in modern hydrology, especially for effective reservoir management. In this research, the auto-regressive integrated moving average (ARIMA) model coupled with the ensemble empirical mode decomposition (EEMD) is presented for forecasting annual runoff time series. First, the original annual runoff time series is decomposed into a finite and often small number of intrinsic mode functions (IMFs) and one residual series using EEMD technique for a deep insight into the data characteristics. Then each IMF component and residue is forecasted, respectively, through an appropriate ARIMA model. Finally, the forecasted results of the modeled IMFs and residual series are summed to formulate an ensemble forecast for the original annual runoff series. Three annual runoff series from Biuliuhe reservoir, Dahuofang reservoir and Mopanshan reservoir, in China, are investigated using developed model based on the four standard statistical performance evaluation measures (RMSE, MAPE, R and NSEC). The results obtained in this work indicate that EEMD can effectively enhance forecasting accuracy and that the proposed EEMD-ARIMA model can significantly improve ARIMA time series approaches for annual runoff time series forecasting.

Forecasting of river discharge is crucial in hydrology and hydraulic engineering owing to its use in the design and management of water resource projects. The problem is customarily settled with data-driven models. In this research, a novel hybrid model which combines continuity equation and fuzzy pattern-recognition concept with artificial neural network (ANN), is presented for downstream river discharge forecasting in a river network. Time-varying water storage in a river station and fuzzy feature of river flow are considered accordingly. To verify the proposed model, traditional ANN model, fuzzy pattern-recognition neural network model, and hydrological modeling network model have been employed as the benchmark models. The root mean squared error, Nash–Sutcliffe efficiency coefficient and accuracy are adopted as evaluation criteria. The proposed hybrid model is applied to compute downstream river discharge in the Yellow River, Georgia, USA. Results indicate that the proposed hybrid model delivers better performance, which can effectively improve forecasting capability at the studied station. It is, therefore, proposed as a novel model for downstream river discharge forecasting because of its highly nonlinear, fuzzy and non-stationary properties.

There has been an increasing need for the proper evaluation of river water quality in order to safeguard public health and to protect the valuable fresh water resources. In order to overcome the own limitations of the traditional evaluations which can only use a point value instead of an interval for grading standards, on the basis of the fuzzy binary comparison method (FBCM) and the theory of variable fuzzy sets (VFS), an integrated variable fuzzy evaluation model (VFEM) is proposed for the assessment of river water quality in this paper. This model possesses the preciseness of the algorithm and operability in practice, can well solve the grading standards which are interval form. In order to explore and compare the present method with other traditional methods, two cases studies in the Three Gorges and Tseng-Wen River are made. The results show that the proposed VFEM method can convey water cleanliness to certain degree by using the eigenvector of level H, which is much stricter in the superior level, and that it can improve the veracity for the assessment of water quality.