Performance of Neural Networks in Daily Streamflow Forecasting

Birikundavyi, S.; Labib, Richard; Trung, H T; Rousselle, Jean

doi:10.1061/(asce)1084-0699(2002)7:5(392)

Cited by 135 publications

(66 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Neural technologies continue to make enormous strides in their struggle to become established as recognized tools that offer efficient and effective solutions for modelling and analysing the behaviour of complex dynamical systems. Time series forecasting has been a particular focus of interest and superior performing models have been reported in a diverse set of fields that include rainfall-runoff modelling (ASCE, 2000a, b;Dawson and Wilby, 2001;Birikundavy et al, 2002;Campolo et al, 2003;Huang et al, 2004;Riad et al, 2004;Hettiarachchi et al, 2005;Senthil Kumar et al, 2005) and sediment prediction (Abrahart and White, 2001;Nagy et al, 2002;Yitian and Gu, 2003;Kisi, 2004;Bhattacharya et al, 2005;Kisi, 2005). Moreover, for flood forecasting purposes, neural solutions offer practical advantages related to operational costs and socio-economic resources that would be of interest in developing countries, e.g.…”

Section: Introductionmentioning

confidence: 99%

Neural network modelling of non-linear hydrological relationships

Abrahart

See

2007

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

Abstract. Two recent studies have suggested that neural network modelling offers no worthwhile improvements in comparison to the application of weighted linear transfer functions for capturing the non-linear nature of hydrological relationships. The potential of an artificial neural network to perform simple non-linear hydrological transformations under controlled conditions is examined in this paper. Eight neural network models were developed: four full or partial emulations of a recognised non-linear hydrological rainfallrunoff model; four solutions developed on an identical set of inputs and a calculated runoff coefficient output. The use of different input combinations enabled the competencies of solutions developed on a reduced number of parameters to be assessed. The selected hydrological model had a limited number of inputs and contained no temporal component. The modelling process was based on a set of random inputs that had a uniform distribution and spanned a modest range of possibilities. The initial cloning operations permitted a direct comparison to be performed with the equation-based relationship. It also provided more general information about the power of a neural network to replicate mathematical equations and model modest non-linear relationships. The second group of experiments explored a different relationship that is of hydrological interest; the target surface contained a stronger set of non-linear properties and was more challenging. Linear modelling comparisons were performed against traditional least squares multiple linear regression solutions developed on identical datasets. The reported results demonstrate that neural networks are capable of modelling nonlinear hydrological processes and are therefore appropriate tools for hydrological modelling.

show abstract

Section: Introductionmentioning

confidence: 99%

Neural network modelling of non-linear hydrological relationships

Abrahart

See

2007

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

show abstract

“…The Nash-Sutcliffe model evaluation statistic is widely used to validate various models [45][46][47]. The Nash-Sutcliffe model efficiency statistic is defined in Equation (5).…”

Section: Water Balancementioning

confidence: 99%

Agricultural Water Conservation in the High Plains Aquifer and Arikaree River Basin

Prior¹,

Oad²,

Kinzli³

2013

JWARP

View full text Add to dashboard Cite

Yuma County is the top crop producing County in Colorado that is dependent on groundwater supplies from the High Plains aquifer for irrigation. The Arikaree River, a tributary of the Republican River in eastern Colorado, is supplied with water from the High Plains aquifer. The Arikaree River alluvium is also a habitat for many terrestrial invertebrates and the threatened Hybognathus hankinsoni (Brassy Minnow). The constant demand on the High Plains aquifer has created declining water levels at the linear rate of 0.183 m/year with the deepest pool in the Arikaree River drying up in 8 to 12 years. In addition to the demands for habitats, the surrounding irrigated agricultural lands require water for crop production. These challenges are currently confronting farmers in eastern Colorado and this research presents possible alternatives to meet these demands. This research presents a combination water balance model, water conservation model, and water conservation survey results from farmers in eastern Colorado to identify alternatives to extend the life of the Arikaree River. The first alternative was to examine the reduction in irrigation water from removing the 18 alluvial irrigation wells that could extend the Arikaree River pools from drying up for 30 years. The other scenario found that water conservation practices with participation of 43%, 57%, and 62% of farmers would extend the drying time to 20, 30, and 40 years, respectively. The final alternative studied was the required participation in conservation practices to stop the decline of the High Plains Aquifer. The analysis found that 77% participation of farmers in all conservation alternatives or reducing pumping by 62.9% would be necessary to stabilize the High Plains Aquifer.

show abstract

“…Time series modeling using ANN has been a particular focus of interest and better performing models have been reported in a diverse set of fields that include rainfall-runoff modeling [8][9][10][11][12] and groundwater level prediction [13][14][15][16]. It is also reported that ANN models are not very satisfied in precision for forecasting because it considered only few aspects of time series property [17].…”

Section: Introductionmentioning

confidence: 99%

Time Series Modeling of River Flow Using Wavelet Neural Networks

Krishna¹,

Rao²,

Nayak³

2011

JWARP

View full text Add to dashboard Cite

A new hybrid model which combines wavelets and Artificial Neural Network (ANN) called wavelet neural network (WNN) model was proposed in the current study and applied for time series modeling of river flow. The time series of daily river flow of the Malaprabha River basin (Karnataka state, India) were analyzed by the WNN model. The observed time series are decomposed into sub-series using discrete wavelet transform and then appropriate sub-series is used as inputs to the neural network for forecasting hydrological variables. The hybrid model (WNN) was compared with the standard ANN and AR models. The WNN model was able to provide a good fit with the observed data, especially the peak values during the testing period. The benchmark results from WNN model applications showed that the hybrid model produced better results in estimating the hydrograph properties than the latter models (ANN and AR).

show abstract

Performance of Neural Networks in Daily Streamflow Forecasting

Cited by 135 publications

References 16 publications

Neural network modelling of non-linear hydrological relationships

Neural network modelling of non-linear hydrological relationships

Agricultural Water Conservation in the High Plains Aquifer and Arikaree River Basin

Time Series Modeling of River Flow Using Wavelet Neural Networks

Contact Info

Product

Resources

About