Application of the Takagi–Sugeno fuzzy system for combination forecasting of river flow in semiarid mountain regions

Because of high spatial heterogeneity and the degree of uncertainty about hydrological processes in large‐scale catchments of semiarid mountain areas, satisfactory forecasting of daily discharge is seldom available using a single model in many practical cases. In this paper the Takagi–Sugeno fuzzy system (TS) and the simple average method (SAM) are applied to combine forecasts of three individual models, namely, the simple linear model (SLM), the seasonally based linear perturbation model (LPM) and the nearest neighbour linear perturbation model (NNLPM) for modelling daily discharge, and the performance of modelling results is compared in five catchments of semiarid areas. It is found that the TS combination model gives good predictions. The results confirm that better prediction accuracy can be obtained by combining the forecasts of different models with the Takagi–Sugeno fuzzy system in semi‐arid mountain areas. Copyright © 2009 John Wiley & Sons, Ltd.     

Real‐time recurrent learning neural network for stream‐flow forecasting

Various types of neural networks have been proposed in previous papers for applications in hydrological events. However, most of these applied neural networks are classified as static neural networks, which are based on batch processes that update action only after the whole training data set has been presented. The time variate characteristics in hydrological processes have not been modelled well. In this paper, we present an alternative approach using an artificial neural network, termed real‐time recurrent learning (RTRL) for stream‐flow forecasting. To define the properties of the RTRL algorithm, we first compare the predictive ability of RTRL with least‐square estimated autoregressive integrated moving average models on several synthetic time‐series. Our results demonstrate that the RTRL network has a learning capacity with high efficiency and is an adequate model for time‐series prediction. We also investigated the RTRL network by using the rainfall–runoff data of the Da‐Chia River in Taiwan. The results show that RTRL can be applied with high accuracy to the study of real‐time stream‐flow forecasting networks. Copyright © 2002 John Wiley & Sons, Ltd.     

A two‐step‐ahead recurrent neural network for stream‐flow forecasting

In many engineering problems, such as flood warning systems, accurate multistep‐ahead prediction is critically important. The main purpose of this study was to derive an algorithm for two‐step‐ahead forecasting based on a real‐time recurrent learning (RTRL) neural network that has been demonstrated as best suited for real‐time application in various problems. To evaluate the properties of the developed two‐step‐ahead RTRL algorithm, we first compared its predictive ability with least‐square estimated autoregressive moving average with exogenous inputs (ARMAX) models on several synthetic time‐series. Our results demonstrate that the developed two‐step‐ahead RTRL network has efficient ability to learn and has comparable accuracy for time‐series prediction as the refitted ARMAX models. We then investigated the two‐step‐ahead RTRL network by using the rainfall–runoff data of the Da‐Chia River in Taiwan. The results show that the developed algorithm can be successfully applied with high accuracy for two‐step‐ahead real‐time stream‐flow forecasting. Copyright © 2003 John Wiley & Sons, Ltd.     

Long‐term runoff forecasting by combining hydrological models and meteorological records

Reservoir operation is generally based on the inflows of the upstream catchment of the reservoir. If the arriving inflows can be forecasted, that can benefit reservoir operation and management. This study attempts to construct a long‐term inflow‐forecasting model by combining a continuous rainfall–runoff model with the long‐term weather outlook from the Central Weather Bureau of Taiwan. The analytical results demonstrate that the continuous rainfall–runoff model has good inflow simulation performance by using 10‐day meteorological and inflow records over a 33‐year period for model calibration and verification. The long‐term inflow forecasting during the dry season was further conducted by combining the continuous rainfall–runoff model and the long‐term weather outlook, which was found to have good performance. Copyright © 2005 John Wiley & Sons, Ltd.     

Changes in potential evapotranspiration and surface runoff in 1981–2010 and the driving factors in Upper Heihe River Basin in Northwest China

Changes in potential evapotranspiration and surface runoff can have profound implications for hydrological processes in arid and semiarid regions. In this study, we investigated the response of hydrological processes to climate change in Upper Heihe River Basin in Northwest China for the period from 1981 to 2010. We used agronomic, climatic and hydrological data to drive the Soil and Water Assessment Tool model for changes in potential evapotranspiration (ET₀) and surface runoff and the driving factors in the study area. The results showed that increasing autumn temperature increased snow melt, resulting in increased surface runoff, especially in September and October. The spatial distribution of annual runoff was different from that of seasonal runoff, with the highest runoff in Yeniugou River, followed by Babaohe River and then the tributaries in the northern of the basin. There was no evaporation paradox at annual and seasonal time scales, and annual ET₀ was driven mainly by wind speed. ET₀ was driven by relative humidity in spring, sunshine hour duration in autumn and both sunshine hour duration and relative humility in summer. Surface runoff was controlled by temperature in spring and winter and by precipitation in summer (flood season). Although surface runoff increased in autumn with increasing temperature, it depended on rainfall in September and on temperature in October and November. Copyright © 2016 John Wiley & Sons, Ltd.     

Modelling stream flow for use in ecological studies in a large,arid zone river,central Australia

Australian arid zone ephemeral rivers are typically unregulated and maintain a high level of biodiversity and ecological health. Understanding the ecosystem functions of these rivers requires an understanding of their hydrology. These rivers are typified by highly variable hydrological regimes and a paucity, often a complete absence, of hydrological data to describe these flow regimes. A daily time‐step, grid‐based, conceptual rainfall–runoff model was developed for the previously uninstrumented Neales River in the arid zone of northern South Australia. Hourly, logged stage data provided a record of stream‐flow events in the river system. In conjunction with opportunistic gaugings of stream‐flow events, these data were used in the calibration of the model. The poorly constrained spatial variability of rainfall distribution and catchment characteristics (e.g. storage depths) limited the accuracy of the model in replicating the absolute magnitudes and volumes of stream‐flow events. In particular, small but ecologically important flow events were poorly modelled. Model performance was improved by the application of catchment‐wide processes replicating quick runoff from high intensity rainfall and improving the area inundated versus discharge relationship in the channel sections of the model. Representing areas of high and low soil moisture storage depths in the hillslope areas of the catchment also improved the model performance. The need for some explicit representation of the spatial variability of catchment characteristics (e.g. channel/floodplain, low storage hillslope and high storage hillslope) to effectively model the range of stream‐flow events makes the development of relatively complex rainfall–runoff models necessary for multisite ecological studies in large, ungauged arid zone catchments. Grid‐based conceptual models provide a good balance between providing the capacity to easily define land types with differing rainfall–runoff responses, flexibility in defining data output points and a parsimonious water‐balance–routing model. Copyright © 2004 John Wiley & Sons, Ltd.     

Effects of the catchment runoff coefficient on the performance of TOPMODEL in rainfall–runoff modelling

Effects of the catchment runoff coefficient on the performance of TOPMODEL in simulating catchment rainfall–runoff relationships are investigated in this paper, with an aim to improve TOPMODEL's simulation efficiency in catchments with a low runoff coefficient. Application of TOPMODEL in the semi‐arid Yihe catchment, with an area of 2623 km² in the Yellow River basin of China, produced a Nash–Sutcliffe model efficiency of about 80%. To investigate how the catchment runoff coefficient affects the performance of TOPMODEL, the whole observed discharge series of the Yihe catchment is multiplied with a larger‐than‐unity scale factor to obtain an amplified discharge series. Then TOPMODEL is used to simulate the amplified discharge series given the original rainfall and evaporation data. For a set of different scale factors, TOPMODEL efficiency is plotted against the corresponding catchment runoff coefficient and it is found that the efficiency of TOPMODEL increases with the increasing catchment runoff coefficient before reaching a peak (e.g. about 90%); after the peak, however, the efficiency of TOPMODEL decreases with the increasing catchment runoff coefficient. Based on this finding, an approach called the discharge amplification method is proposed to enhance the simulation efficiency of TOPMODEL in rainfall–runoff modelling in catchments with a low runoff coefficient. Copyright © 2004 John Wiley & Sons, Ltd.     

Semi‐distributed flood runoff model at the subcontinental scale for southwestern Iran

Rainfall–runoff modelling, as a surface hydrological process, on large‐scale data‐poor basins is currently a major topic of investigation that requires the model parameters be identified by using basin physical characteristics rather than calibration. This paper describes the application of the TOPMODEL framework accompanied by a kinematic wave model to the Karun River sub‐basins in southwestern Iran with just one conceptual parameter for calibration. ISLSCP1, HYDRO1K and Reynolds data sets are presented in a geographical information system and used as data sources for meteorological information, hydrological features and soil characteristics of the study area respectively. The results show that although the model developed can adequately predict flood runoff in the catchment with only one calibrated parameter, it is suggested that the effect of surface reservoirs be considered in the proposed model. Copyright © 2007 John Wiley & Sons, Ltd.     

Effects of land‐use changes on hydrological processes in the middle basin of the Heihe River,northwest China

The effects of land‐use changes on the runoff process in the midstream plain of this arid inland river basin are a key factor in the rational allocation of water resources to the middle and lower reaches. The question is whether and by how much increasingly heavy land use impacts the hydrological processes in such an arid inland river basin. The catchment of the Heihe River, one of the largest inland rivers in the arid region of northwest China, was chosen to investigate the hydrological responses to land‐use change. Flow duration curves were used to detect trends and variations in runoff between the upper and lower reaches. Relationships among precipitation, upstream runoff, and hydrological variables were identified to distinguish the effects of climatic changes and upstream runoff changes on middle and downstream runoff processes. The quantitative relation between midstream cultivated land use and various parameters of downstream runoff processes were analysed using the four periods of land‐use data since 1956. The Volterra numerical function relation of the hydrological non‐linear system response was utilized to develop a multifactor hydrological response simulation model based on the three factors of precipitation, upstream runoff, and cultivated land area. The results showed that, since 1967, the medium‐ and high‐coverage natural grassland area in the midstream region has decreased by 80·1%, and the downstream runoff has declined by 27·32% due to the continuous expansion of the cultivated land area. The contribution of cultivated land expansion to the impact on the annual total runoff is 14–31%, on the annual, spring and winter base flow it is 44–75%, and on spring and winter discharge it is 23–64%. Once the water conservation plan dominated by land‐use structural adjustments is implemented over the next 5 years, the mean annual discharge in the lower reach could increase by 8·98% and the spring discharge by 26·28%. This will significantly alleviate the imbalance between water supply and demand in both its quantity and temporal distribution in the middle and lower reaches. Copyright © 2006 John Wiley & Sons, Ltd.     

A systematic approach to the input determination for neural network rainfall–runoff models

Input determination has a great influence on the performance of artificial neural network (ANN) rainfall–runoff models. To improve the performance of ANN models, a systematic approach to the input determination for ANN models is proposed. In the proposed approach, the irrelevant inputs are removed. Then an adequate ANN model, which only includes highly relevant inputs, is constructed. Unlike the trial‐and‐error procedure, the proposed approach is more systematic and avoids unnecessary trials. To demonstrate the effectiveness of the proposed approach, an application to actual typhoon events is presented. The results show that the proposed ANN model, which is constructed by the proposed approach, has advantages over those obtained by the trial‐and‐error procedure. The proposed ANN model has a simpler architecture, needs less training time, and performs better. The proposed ANN model is recommended as an alternative to existing rainfall–runoff ANN models. Copyright © 2007 John Wiley & Sons, Ltd.     

Expansion of agricultural oasis in the Heihe River Basin of China: Patterns,reasons and policy implications

The Heihe River Basin (HRB) is the second largest inland river basin in the arid region of northwestern China. An agricultural oasis is a typical landscape in arid regions providing precious fertile soil, living space and ecological services. The agricultural oasis change has been one of the key issues in sustainable development in recent decades. In this paper, we examined the changes in the agricultural oasis in HRB and analyzed the socio-economic and climatic driving forces behind them. It was found that the agricultural oasis in HRB expanded by 25.11% and 14.82% during the periods of 1986–2000 and 2000–2011, respectively. Most of the newly added agricultural oases in HRB were converted from grassland (40.94%) and unused land (40.22%). The expansion in the agricultural oasis mainly occurred in the middle reaches of HRB, particularly in the counties of Shandan, Minle, Jinta and Jiuquan city. Changes in the rural labor force, annual temperature and precipitation have significant positive effects on agricultural oasis changes, while the ratio of irrigated agricultural oases has significant negative effects on agricultural oasis changes. The agricultural oasis expansion in HRB is the combined effect of human activity and climate change.     

Green‐Ampt Curve‐Number mixed procedure as an empirical tool for rainfall–runoff modelling in small and ungauged basins

The Soil Conservation Service Curve Number (SCS‐CN) method is a popular rainfall–runoff model that is widely used to estimate direct runoff from small and ungauged basins. The SCS‐CN is a simple and valuable approach to quantify the total streamflow volume generated by storm rainfall, but its use is not appropriate for estimating the sub‐daily incremental rainfall excess. To overcome this drawback, we propose to include the Green‐Ampt (GA) infiltration model into a mixed procedure, which is referred to as Curve Number for Green‐Ampt (CN4GA), aiming to distribute in time the information provided by the SCS‐CN method. For a given storm, the computed SCS‐CN total net rainfall amount is employed to calibrate the soil hydraulic conductivity parameter of the GA model. The proposed procedure is evaluated by analysing 100 rainfall–runoff events that were observed in four small catchments of varying size. CN4GA appears to provide encouraging results for predicting the net rainfall peak and duration values and has shown, at least for the test cases considered in this study, better agreement with the observed hydrographs than the classic SCS‐CN method. Copyright © 2012 John Wiley & Sons, Ltd.     

Error analysis for the evaluation of model performance: rainfall–runoff event time series data

This paper provides a procedure for evaluating model performance where model predictions and observations are given as time series data. The procedure focuses on the analysis of error time series by graphing them, summarizing them, and predicting their variability through available information (recalibration). We analysed two rainfall–runoff events from the R‐5 data set, and evaluated 12 distinct model simulation scenarios for these events, of which 10 were conducted with the quasi‐physically‐based rainfall–runoff model (QPBRRM) and two with the integrated hydrology model (InHM). The QPBRRM simulation scenarios differ in their representation of saturated hydraulic conductivity. Two InHM simulation scenarios differ with respect to the inclusion of the roads at R‐5. The two models, QPBRRM and InHM, differ strongly in the complexity and number of processes included. For all model simulations we found that errors could be predicted fairly well to very well, based on model output, or based on smooth functions of lagged rainfall data. The errors remaining after recalibration are much more alike in terms of variability than those without recalibration. In this paper, recalibration is not meant to fix models, but merely as a diagnostic tool that exhibits the magnitude and direction of model errors and indicates whether these model errors are related to model inputs such as rainfall. Copyright © 2004 John Wiley & Sons, Ltd.     

Modelling the rainfall–runoff process of the Mara River basin using the Soil and Water Assessment Tool

The delicate balance between human utilization and sustaining its pristine biodiversity in the Mara River basin (MRB) is being threatened because of the expansion of agriculture, deforestation, human settlement, erosion and sedimentation and extreme flow events. This study assessed the applicability of the Soil and Water Assessment Tool (SWAT) model for long‐term rainfall–runoff simulation in MRB. The possibilities of combining/extending gage rainfall data with satellite rainfall estimates were investigated. Monthly satellite rainfall estimates not only overestimated but also lacked the variability of observed rainfall to substitute gage rainfall in model simulation. Uncertainties related to the quality and availability of input data were addressed. Sensitivity and uncertainty analysis was reported for alternative model components and hydrologic parameters used in SWAT. Mean sensitivity indices of SWAT parameters in MRB varied with and without observed discharge data. The manual assessment of individual parameters indicated heterogeneous response among sub‐basins of MRB. SWAT was calibrated and validated with 10 years of discharge data at Bomet (Nyangores River), Mulot (Amala River) and Mara Mines (Mara River) stations. Model performance varied from satisfactory at Mara Mines to fair at Bomet and weak at Mulot. The (Nash–Sutcliff efficiency, coefficient of determination) results of calibration and validation at Mara Mines were (0.68, 0.69) and (0.43, 0.44), respectively. Two years of moving time window and flow frequency analysis showed that SWAT performance in MRB heavily relied on quality and abundance of discharge data. Given the 5.5% area contribution of Amala sub‐basin as well as uncertainty and scarcity of input data, SWAT has the potential to simulate the rainfall runoff process in the MRB. Copyright © 2012 John Wiley & Sons, Ltd.     

Conceptual and numerical models for groundwater flow in an arid inland river basin

The Heihe River Basin (HRB) is an inland watershed in northwest China with a total area of approximately 130,000 km², stretching from the Qilian Mountains in the south to the oases and agricultural fields in the middle and further to the Gobi desert in the north bordering Mongolia. As part of a major ecohydrological research initiative to provide a stronger scientific underpinning for sustainable water management in arid ecosystems, a regional‐scale integrated ecological and hydrological model is being developed, incorporating the knowledge based on the results of environmental isotope tracer analysis and the multiscale observation datasets. The first step in the model development effort is to construct and calibrate a groundwater flow model for the middle and lower HRB where the oases and vegetation along the Heihe river corridor are highly dependent on groundwater. In this study, the software tool ‘Arc Hydro Groundwater’ is used to build and visualize a hydrogeological data model for the HRB that links all relevant spatiotemporal hydrogeological data in a unified geodatabase within the ArcGIS environment. From the conceptual model, a regional‐scale groundwater flow model has been developed using MODFLOW‐2005. Critical considerations in developing the flow model include the representation of mountainous terrains and fluvial valleys by individual model layers, treatment of aquifer heterogeneities across multiple scales and selection of proper observation data and boundary conditions for model calibration. This paper discusses these issues in the context of the Heihe River Basin, but the results and insights from this study will have important implications for other large, regional groundwater modelling studies, especially in arid and semiarid inland river basins. Copyright © 2014 John Wiley & Sons, Ltd.     

Comparative analysis of a geomorphology‐based instantaneous unit hydrograph in small mountainous watersheds

A conceptual insytnataneous unit hydrograph (IUH) based on geomorphologival association of linear reservoirs (GR) previously developed by the authors has been compared with other IUH models: a distributed GR variation (GR(v)), the Nash IUH, the Chutha and Dooge IUH, and the Troutman and Karlinger IUH for the analysis of direct runoff hydrographs recorded in three experimental watershed of the north of Spain. The comparison was made through a calibration‐validation process in which a leave‐one‐out cross‐validation method was applied. The results indicate the satisfactory performance of all the models, with the advantage for the GR model of the dependence on only one parameter, which can be identified from the watershed and event characteristics. This property makes its use easier than that of other models. Copyright © 2011 John Wiley & Sons, Ltd.     

Predicting catchment flow in a semi‐arid region via an artificial neural network technique

A model of rainfall–runoff relationships is an essential tool in the process of evaluation of water resources projects. In this paper, we applied an artificial neural network (ANN) based model for flow prediction using the data for a catchment in a semi‐arid region in Morocco. Use of this method for non‐linear modelling has been demonstrated in several scientific fields such as biology, geology, chemistry and physics. The performance of the developed neural network‐based model was compared against multiple linear regression‐based model using the same observed data. It was found that the neural network model consistently gives superior predictions. Based on the results of this study, artificial neural network modelling appears to be a promising technique for the prediction of flow for catchments in semi‐arid regions. Accordingly, the neural network method can be applied to various hydrological systems where other models may be inappropriate. Copyright © 2004 John Wiley & Sons, Ltd.     

Multilevel modeling of NPP change and impacts of water resources in the Lower Heihe River Basin

Net primary productivity (NPP) lays the foundation for provision of various ecosystem services, and understanding the impacts of potential influencing factors on NPP is of great significance to formulating appropriate management measures to guarantee the sustainable provision of essential ecosystem services. This study analyzed the impacts of potential influencing factors on NPP in the lower Heihe River Basin, a typical arid and semi-arid region in China. First, NPP was estimated with the C-FIX model, and then the multilevel model was used to analyze the impacts of potential influencing factors on NPP during 2000–2008. Finally decomposition analysis was used to further analyze the contribution of influencing factors to NPP change during 2000–2008. The average NPP increased by approximately 9.07% during 2000–2008, and results of the multilevel model indicate that both the socioeconomic variables and demographic variables are useful in explaining NPP change. In particular, coefficients of rainfall and evapotranspiration which represent the water availability reached 0.0456 and 0.2956, respectively. Results of decomposition analysis suggested that the water availability played an important role in increasing NPP, with a contribution rate of 44.17%, and it is necessary to carry out some policies that can promote the water use efficiency to increase NPP under the background of climate change and intensified human activities. There are some uncertainties in the results of this study, but these results still can provide valuable reference information for the water resource management to increase the ecosystem service supply in the lower Heihe River Basin.     

Recognition method for mid‐ to long‐term runoff forecasting factors based on global sensitivity analysis in the Nenjiang River Basin

Mid‐ to long‐term runoff forecasting is important to China. Forecasting based on physical causes has become the trend of this field, and recognition of key factors is central to recent development. Here, global sensitivity analysis based on back‐propagation arithmetic was used to calculate the sensitivity of up to 24 factors that affect runoff in the Nenjiang River Basin. The following five indices were found to be key factors for mid‐ to long‐term runoff forecasting during flood season: Tibetan Plateau B, index of the strength of the East Asian trough, index of the area of the northern hemisphere polar vortex, zonal circulation index over the Eurasian continent and index of the strength of the subtropical high over the western Pacific. The hydrological climate of the study area and the rainfall–runoff laws were then analysed in conjunction with its geographical position and topographic condition. The rationality of the results can be demonstrated from the positive analysis point of view. The results of this study provide a general method for selection of mid‐ to long‐term runoff forecasting factors based on physical causes. Copyright © 2012 John Wiley & Sons, Ltd.     

Effect of rainfall uncertainty on the performance of physically based rainfall–runoff models

This paper analyses the effect of rain data uncertainty on the performance of two hydrological models with different spatial structures: a semidistributed and a fully distributed model. The study is performed on a small catchment of 19.6 km² located in the north‐west of Spain, where the arrival of low pressure fronts from the Atlantic Ocean causes highly variable rainfall events. The rainfall fields in this catchment during a series of storm events are estimated using rainfall point measurements. The uncertainty of the estimated fields is quantified using a conditional simulation technique. Discharge and rain data, including the uncertainty of the estimated rainfall fields, are then used to calibrate and validate both hydrological models following the generalized likelihood uncertainty estimation (GLUE) methodology. In the storm events analysed, the two models show similar performance. In all cases, results show that the calibrated distribution of the input parameters narrows when the rain uncertainty is included in the analysis. Otherwise, when rain uncertainty is not considered, the calibration of the input parameters must account for all uncertainty in the rainfall–runoff transformation process. Also, in both models, the uncertainty of the predicted discharges increase in similar magnitude when the uncertainty of rainfall input increase.