Monthly Flow Data Research Articles

Improving forecasting accuracy of river flow using gene expression programming based on wavelet decomposition and de-noising

Abstract Due to the effects of anthropogenic activities and natural climate change, streamflows of rivers have gradually decreased. In order to maintain reliable water supplies, reservoir operation and water resource management, accurate streamflow forecasts are very important. Based on monthly flow data from five hydrological stations in the middle and lower parts of the Hanjiang River Basin, between 1989 and 2009, we consider an efficient approach of adopting the gene expression programming model based on wavelet decomposition and de-noising (WDDGEP) to forecast river flow. Original flow time series data are initially decomposed into one sub-signal approximation and seven sub-signal details using the dmey wavelet. A wavelet threshold de-noising method is also applied in this study. Data that have been de-noised after decomposition are then adopted as inputs for WDDGEP models. Finally, the forecasted sub-signal results are summed to formulate an ensemble forecast for the original monthly flow series. A comparison of the prediction accuracy between the two models is based on three performance evaluation measures. Results show that the new WDDGEP models can effectively enhance accuracy in forecasting streamflow, and the proposed wavelet-based de-noising of the observed non-stationary time series is an effective measure to improve simulation accuracy.

Estimation in periodic restricted EXPAR(1) models

ABSTRACTThis article is devoted to study the problem of estimation in the periodic restricted exponential autoregressive EXPAR(1) models. The estimation procedure that is used is the least-square method. Simulation studies are carried out in order to check the asymptotic properties. An application to monthly flow data for the Fraser River in British Columbia is included.

A stepwise model to predict monthly streamflow

In this study, a stepwise model empowered with genetic programming is developed to predict the monthly flows of Hurman River in Turkey and Diyalah and Lesser Zab Rivers in Iraq. The model divides the monthly flow data to twelve intervals representing the number of months in a year. The flow of a month, t is considered as a function of the antecedent month’s flow (t−1) and it is predicted by multiplying the antecedent monthly flow by a constant value called K. The optimum value of K is obtained by a stepwise procedure which employs Gene Expression Programming (GEP) and Nonlinear Generalized Reduced Gradient Optimization (NGRGO) as alternative to traditional nonlinear regression technique. The degree of determination and root mean squared error are used to evaluate the performance of the proposed models. The results of the proposed model are compared with the conventional Markovian and Auto Regressive Integrated Moving Average (ARIMA) models based on observed monthly flow data. The comparison results based on five different statistic measures show that the proposed stepwise model performed better than Markovian model and ARIMA model. The R2 values of the proposed model range between 0.81 and 0.92 for the three rivers in this study.

Basic tools for setting environmental flows at the regional scale: application of the ELOHA framework in a Mediterranean river basin

AbstractEcological limits of hydrological alteration (ELOHA) framework aims to set limits of hydrologic alteration based on empirical hydroecological relationships built along flow impairment gradients. These relationships can be specific for different natural flow regime types. The object of our study is testing the ELOHA method in the Ebro River Basin. Seven flow alteration–ecological response curves were obtained in the present study as preliminary guidelines for developing and calibrating environmental flow standards in the Ebro River Basin. For continental Mediterranean‐pluvial flow regime mean monthly flow in March was related to Plecoptera abundance (ELOHAs ±21% from baseline), and high flow magnitude and frequency were closely related to Plecoptera, Ephemeroptera and Trichoptera richness (ELOHAs for high flow magnitude and frequency were +18 and +54% the reference central value respectively). For nivo‐pluvial flow regime, high flow magnitude was found indicative of macroinvertebrate biodiversity and Coleoptera abundance (ELOHAS −12 and +23% from central reference value respectively). And for pluvio‐oceanic flow regime, mean monthly flow in January was indicative of the proportion of Gastropoda, Oligochaeta and Diptera taxa (GOLD index) (ELOHA +17% from baseline). These hydrological limits are endorsed by recent international reference works. Flow alteration–ecological relationships presented here were constructed from real monthly flow data records and macroinvertebrate samples at family level. ELOHA values obtained in this study pretend to be a starting point for implementing an environmental flow regime program in the Ebro River Basin. Copyright © 2016 John Wiley & Sons, Ltd.

Application of periodic autoregressive process to the modeling of the Garonne river flows

Accurate forecasting of river flows is one of the most important applications in hydrology, especially for the management of reservoir systems. To capture the seasonal variations in river flow statistics, this paper develops a robust modeling approach to identify and to estimate periodic autoregressive (PAR) model in the presence of additive outliers. Since the least squares estimators are not robust in the presence of outliers, we suggest a robust estimation based on residual autocovariances. A genetic algorithm with Bayes information criterion is used to identify the optimal PAR model. The method is applied to average monthly and quarter-monthly flow data (1959–2010) for the Garonne river in the southwest of France. Results show that the accuracy of forecasts is improved in the robust model with respect to the unrobust model for the quarter-monthly flows. By reducing the number of parameters to be estimated, the principle of parsimony favors the choice of the robust approach.

Evapotranspiration from Remote Sensing to Improve the Swat Model in Eastern Amazonia

ABSTRACT In this study, we estimated the evapotranspiration from orbital images - MODIS (Moderate Resolution Imaging Spectroradiometer) for assimilation in the hydrological modeling of the SWAT (Soil Water Assessment Tools) model. The data used include the period between October 2003 and December 2006 of the sub-basin of the Lajeado River, located in the Tocantins-Araguaia River basin in Tocantins state. Overall, the results of the use of heat flows estimated by remote sensors in the SWAT model can be considered satisfactory. The values of the COE (coefficient of efficiency of Nash-Sutcliffe) ranged from -0.40 to 0.91 in the comparison with the daily flow data and from 0.17 to 0.77 with the monthly flow data, with the assimilation of evapotranspiration from orbital images. These results indicate benefit to the model adjustment due to improvement in the data assimilated of approximately 0.91 in the COE on daily scale and 0.60 in the CEO on monthly scale.

Comparison of performance of statistical models in forecasting monthly streamflow of Kizil River, China

This paper presents the application of autoregressive integrated moving average (ARIMA), seasonal ARIMA (SARIMA), and Jordan-Elman artificial neural networks (ANN) models in forecasting the monthly streamflow of the Kizil River in Xinjiang, China. Two different types of monthly streamflow data (original and deseasonalized data) were used to develop time series and Jordan-Elman ANN models using previous flow conditions as predictors. The one-month-ahead forecasting performances of all models for the testing period (1998-2005) were compared using the average monthly flow data from the Kalabeili gaging station on the Kizil River. The Jordan-Elman ANN models, using previous flow conditions as inputs, resulted in no significant improvement over time series models in one-month-ahead forecasting. The results suggest that the simple time series models (ARIMA and SARIMA) can be used in one-month-ahead streamflow forecasting at the study site with a simple and explicit model structure and a model performance similar to the Jordan-Elman ANN models.

The Impact of Dynamic Environmental Flow Releases on Hydropower Production in the Zambezi River Basin

AbstractIncorporation of environmental flow releases from reservoirs has proven to be challenging due to fear of losses to existing water uses. Moreover environmental flow requirements (EFR) have not often been operationalized. This study compares the possibility of implementing dynamic EFR based on natural flows lagged against an upstream unregulated gauging point with static EFR. It simulates different scenarios with a high flow release in the wet season and analyses its impacts on hydropower production. This method accounts fully for the natural variability of environmental flows, implying less pressure on existing water uses during relatively dry years. Joint operation of two cascading dams vs. individual operation for EFR was also explored. These approaches were tested for the Zambezi River basin in Southern Africa using a water resources model, WAFLEX. Historic data on reservoir water levels, releases and power generation of the hydropower schemes were synthesized. Combining these yielded a validated series of monthly flow data for a 28 year period (1982‐2010). The results show that Kariba and Cahora Bassa reservoirs face a reduction in power produced when they would annually release an environmental flow. However, the dynamic EFR method entails smaller hydropower losses. Joint environmental flow operations will reduce overall basin power production more than if Cahora Bassa alone would release an environmental flow. However, such joint operation would be more beneficial to the ecosystem.

Development of a new index to assess river regime impacts after dam construction

Construction of dams affects river flow regimes and changes downstream ecosystems in many ways. The impacts of dams depend on local climate, dam purpose and operation policy. This study analysed the impacts of dam construction on river flow regimes and developed a new method to evaluate the effects on river flow. By using three main flow characteristics of river regime (magnitude, timing and intra-annual alteration), a new combined river impact (RI) index to assess construction impacts based on monthly flow data was developed. A preliminary evaluation of the method was made using the river Ebro as a case. A classification was developed where the impact was assessed using a scale from low to drastic. The method was further tested to quantify the effect of dams in 15 rivers located in different regions. The method showed good potential to classify different dam types of variable sizes constructed for different purposes. The RI index method also proved suitable for assessing the environmental impacts of dam construction and could be extended by adding other characteristics of flow regime to evaluate the effects of climate change or changes in land use on river flow regimes.

A genetic programming approach to river flow modeling

This paper proposes the application of genetic programming (GP) to forecast monthly river flow. The river flow models were improved by the monthly rainfall and flow data from three stations for Kizilirmak River, Turkey. The coefficient of determination (R2) and root mean square error (RMSE) values were used for evaluating the accuracy of the developed models. The most appropriate GP model was determined as model having monthly flow data of Yamula and Bulakbasi stations according to the model performance criteria for testing data set. The models obtained using the GP were compared with multiple linear regression (MLR) techniques in river flow forecasting. The comparison results revealed that the suggested GP model performs quite well compared to MLR models. It was shown that the suggested GP model with R2 = 0.96 and RMSE = 8.02 m3/s for testing period could be used in planning and management of water resources.

Ecological Significance of Instream Hydrological Statistical Parameters

AbstractHydrological parameters are significant both in hydrological events and ecological conservation. The implications of hydrological statistical parameters in river ecology were investigated by considering three basic parameters, as follows: (1) mean value, (2) coefficient of variation, and (3) extreme ratio. Four typical rivers selected as case studies include the Yellow and Yiluo Rivers in north China, and the Yangtze and Oujiang Rivers in south China. Hydrological statistical parameters were calculated using monthly flow data collected from the key hydrometric stations on each of the four rivers. The results revealed that the three hydrological statistical parameters are all crucial ecological indicators that reflect the ecological conditions of river ecosystems. The mean value of the flow may be the key determinant of habitat range and aquatic biomass. The coefficient of variation may reflect the stability or vulnerability of the ecological conditions of rivers, and the extreme ratio may reflect ...

Assessment of ecological flow rate by flow duration and environmental management class in the Geum River, Korea

This study attempted to analyze flow duration in a basin using a method to estimate environmental flow developed by the International Water Management Institute, and simulate the effects of runoff characteristics unique to a river and flow variability due to basin developments on aquatic ecosystems. To do so, KModSim, a simulation model for basin-wide water distribution, was used to assess flow duration in the Geum River basin, one of the four major river basins in Korea, by environmental management class (EMC). Flow duration curves by EMC at Sutong and Gongju sites were derived on the basis of natural flow in the Geum River basin. As a result, they were found to be consistent with the results of previous studies. Time series of mean monthly flow data by EMC were plotted together with those of simulated flow data by reservoir operation scenario; Sutong and Gongju points both showed flow behaviors corresponding almost to “A” in EMC. In addition, the characteristics of habitats by species of fish were identified through monitoring fish habitat at the Sutong site, so that optimal ecological flow rate was estimated. For this purpose, relations between flow discharge and weighted usable area for Coreoleuciscus splendidus and Pseudopungtungia nigra were projected using physical habitat simulation system, and EMCs consistent with flow duration curves (estimated taking in-stream flow) were assessed. The results or findings reported in this study are expected to serve as basic data for making a plan to efficiently monitor and manage aquatic ecosystems in the Geum River basin.

A combined generalized regression neural network wavelet model for monthly streamflow prediction

The ability of a combined model, Wavelet-Generalized Regression Neural Network (WGRNN), is investigated in the current study for the prediction of monthly streamflows. The WGRNN model is obtained by combining two methods, Discrete Wavelet Transform (DWT) and Generalized Regression Neural Network (GRNN), for one-month-ahead streamflow forecasting. The monthly flow data of two stations, the Gerdelli Station on the Canakdere River and the Isakoy Station on the Goksudere River, in the Eastern Black Sea region of Turkey are used in the study. The forecasts of the WGRNN model are tested using the Root Mean Square Error (RMSE), Variance Account For (VAF) and correlation coefficient (R) statistics and the results are compared with those of the single GRNN and Feed Forward Neural Network (FFNN). The comparison results revealed that the WGRNN performs better than the GRNN and FFNN models in monthly streamflow prediction. For the Gerdelli and Isakoy stations, it is found that the WGRNN models with RMSE = 5.31 m3/s, VAF = 52.3%, R = 0.728 and RMSE = 3.36 m3/s, VAF = 55.1%, R = 0.742 in the test period are superior in forecasting monthly streamflows than the best accurate GRNN models with RMSE = 6.39 m3/s, VAF = 30.1%, R = 0.553 and RMSE = 4.19 m3/s, VAF = 30.1%, R = 0.549, respectively.

ASSESSMENT OF USING ANN FOR MODELLING THE RAINFALL-RUNOFF RELATION: AN APPLICATION FOR MIDDLE PART OF EUPHRATES BASIN

Artificial intelligence methods such as Artificial Neural Network (ANN), Genetic Algorithm (GA) and Fuzzy Logic (FL) provide prosperous results have recently been used in the modelling of rainfall-flow relations and they are becoming more popular in hydraulic engineering practices. In this paper the relations between the average monthly flow data from the flow observation station numbered as 2122 and the monthly total rainfall data from the rainfall observation station numbered as 17099 located in the central Euphrates river basin are investigated by using the feed-forward back-propagation neural network (FFBPNN) method from ANN solutions and afterwards the results are compared using Multi linear Regression (MLR) method. New flow values are estimated by this procedure that uses the flow and rainfall data as input. This paper concludes that FFBPNN method provides better results compared to the results from MLR method.

Flow regime patterns and their controlling factors in the Ebro basin (Spain)

Natural intra-annual flow fluctuations vary between rivers, being a determining factor for aquatic insects, fish and riparian communities which are adapted to the habitat conditions and different flows throughout the seasons. Moreover, restoration of seasonal flow patterns plays an important role in achieving good ecological status of rivers, through the preservation and/or recovery of components and processes of natural river ecosystems. In this work we: (a) classify fluvial segments in the Ebro basin (North-Eastern Spain) according to the intra-annual variability of flows under natural conditions using statistical cluster analysis of monthly mean flow data; (b) characterise the resulting flow typologies according to several ecologically important hydrological variables; (c) analyse the relationships between flow regimes of fluvial segments and physical variables from their catchments; and finally (d) predict the most probable natural flow regime using logistic models based on the most determinant physical characteristics. Fifteen natural flow typologies were described in the Ebro basin, which were characterised in terms of flow fluctuation through the year as well as timing, flow ratio and duration of the maximum and minimum flows. Precipitation, biogeography and geology of catchments showed the highest correlations with flow regimes. Basin size, mean elevation and slope were also correlated. The logistic model we developed had a prediction success of 72% in the Ebro basin. The definition of the natural hydrological conditions (to which the biological communities are tailored), even when flow data are not available, is an important support in the management of river ecosystems. It is especially suitable for setting goals in aquatic ecosystem conservation or restoration projects.

Hydrological cycle simulation of an irrigation district based on a SWAT model

A distributed hydrological cycle model of an irrigation district was developed based on the SWAT model (Soil and Water Assessment Tool) and an artificial-natural composite of the hydrological cycle was considered. The SWAT model was modified in the aspects of the extraction of ditches, distributed subbasins and hydrologic response units, and the calculation method of the crop’s actual ET. The applicability of the model was validated in the Fenhe irrigation district. The information on hydrology, weather and water use from 1996 to 2001 in the Fenhe irrigation district was used to simulate and analyze the water balance. The sensitive parameters were estimated by E n s , RE and R 2 . The model was further validated with the monthly flow data from 2002 to 2006. The results showed that the simulated results of two monitoring points meet the estimated requirements. The RE value of the average annual runoff at the Erba Station varied from −7.34% to 19.13% except a low RE value (−30.70%) in 2006. The RE of the average annual runoff at the Yitang station was from −17.21% to 9.86% with an exceptional RE value (−21.13%) in 2003. From the monthly simulated results, the R 2 of the monthly runoff at the Erba and Yitang stations was 0.81 and 0.77, respectively. The E n s of the monthly runoff was 0.72 and 0.65, respectively. A modified SWAT model was applicable for water balance simulation at the Fenhe irrigation district.

EXTENSION OF STREAM FLOW SERIES USING ARTIFICIAL NEURAL NETWORKS

ABSTRACT Sufficient length of record of flow is very important for accurate estimation of probable maximum flood of a river at the location of any hydraulic structure for its economic design or planning of reservoir operation policy. Many a times the length of record of flow is insufficient or not available at all. Normally under such situations hydrologists either extend the short-term data somehow or generate the data with the properties of historical observed data. The most accepted method to extend short- term data with the help of long-term data at a neighboring station is Langbein's log deviation. This method consists of developing a correlation between the streamflows at short-term and long-term gauging stations and to estimate missing flow series the procedure requires cumbersome mathematical computations. This paper demonstrates the application of artificial neural networks (ANNs) in estimating the missing mean monthly streamflow of Narmada river at Dindori G & D site (M.P.), India using the monthly flow data obtained from the Cental Water Commission. The performance of the artificial neural network is compared with that of the Langbein's log deviation method. The results suggest that the ANN approach may provide a superior alternative to Langbein's log deviation method for estimating the missing flow series or forecasting flows with less computational effort and time.

Identification of periodic autoregressive moving average models and their application to the modeling of river flows

The generation of synthetic river flow samples that can reproduce the essential statistical features of historical river flows is useful for the planning, design, and operation of water resource systems. Most river flow series are periodically stationary; that is, their mean and covariance functions are periodic with respect to time. This article develops model identification and simulation techniques based on a periodic autoregressive moving average (PARMA) model to capture the seasonal variations in river flow statistics. The innovations algorithm is used to obtain parameter estimates. An application to monthly flow data for the Fraser River in British Columbia is included. A careful statistical analysis of the PARMA model residuals, including a truncated Pareto model for the extreme tails, produces a realistic simulation of these river flows.

Application of linear stochastic models to monthly flow data of Kelkit Stream

This paper presents a methodology on modeling of historical data for monthly flows from Kelkit Stream. The watershed is located on the north Anatolia and the stream is formed by joining together of small streams. For the modeling purpose, linear stochastic models known as either Box–Jenkins or ARIMA (autoregressive-moving average) were used to simulate monthly data. Diagnostic checks were done for all the models selected from the autocorrelation function (ACF) and partial autocorrelation function (PACF). The models that have the minimum Schwarz Bayesian Criterion (SBC) among the selected models fulfilled all the diagnostic checks were assumed as the best model for monthly data. For five years, the predicted data using the best models is compared to the observed data. The basic statistical properties of the observed and predicted data were compared. The results show that generated data preserve the basic statistical properties of the original series.

River Flow Modeling Using Artificial Neural Networks

This paper demonstrates the application of artificial neural networks (ANNs) in predicting mean monthly streamflow. The estimation of flow is very important for reservoir operation policy; therefore, in this study, based on the monthly flow data obtained from the Turkey State of Water Works, ANNs have been used to predict river flow. Autoregressive (AR) models have also been applied to the same data. The performance of the neural network is compared with that of the statistical method.