Constraints of artificial neural networks for rainfall-runoff modelling: trade-offs in hydrological state representation and model evaluation

Vos, N. J. de; Rientjes, T.H.M.

doi:10.5194/hess-9-111-2005

Cited by 160 publications

(44 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ANNs with one hidden layer are commonly used in hydrologic modeling (Dawson and Wilby 2001;De Vos and Rientjes 2005) since these networks are considered to provide enough complexity to accurately simulate the nonlinear-properties of the hydrologic process.…”

Section: Artificial Neural Network (Ann)mentioning

confidence: 99%

Assessment of the impacts of nonstationarity on watershed runoff using artificial neural networks: a case study in Ardebil, Iran

Javan

Lialestani

Ashouri

et al. 2015

Model. Earth Syst. Environ.

View full text Add to dashboard Cite

The purpose of this study is to investigate the impacts of climate change on the runoff of Gharehsoo River Basin in the northwest of Iran. In this research, the outputs of monthly precipitation and temperature data of PRECIS (Providing Regional Climates for Impacts Studies) model, a regional climate model with 50 9 50 km spatial resolution on the basis of B2 scenario, is used for the base and the future (2071-2100) periods. The output results of PRECIS show that the average temperature of the watershed increased up to 2-5°C, for the period spanning from 2070 to 2100. In addition, compare to the base period, we are expecting to receive more precipitation in future for the months of January, February, March, September and December. The artificial neural network (ANN) was applied to quantify the future discharge. The results show that in the future, the discharge of Gharehsoo River watershed decreases for all months. Moreover, the peak discharge in the future period happens 1 month earlier, due to increasing in the temperature and earlier start of snow melting season. Finally, 1 and 2°C increase in temperature lead to 0.05-8.2 % and 0.1-13.4 % decrease of average monthly discharge, respectively.

show abstract

Section: Artificial Neural Network (Ann)mentioning

confidence: 99%

Assessment of the impacts of nonstationarity on watershed runoff using artificial neural networks: a case study in Ardebil, Iran

Javan

Lialestani

Ashouri

et al. 2015

Model. Earth Syst. Environ.

View full text Add to dashboard Cite

show abstract

“…In some studies, artificial intelligence models exhibited better performances than conceptual MWBMs (Hsu et al, 1995;Shamseldin, 1997;Machado et al, 2011;Rezaeianzadeh et al, 2013). However, artificial intelligence models have also been criticized for their lack of explanation capability, overparameterization and over-fitting (Kaastra and Boyd, 1996;Gaume and Gosset, 2003;de Vos and Rientjes, 2005).…”

Section: Introductionmentioning

confidence: 99%

Comparison of performance of twelve monthly water balance models in different climatic catchments of China

Bai

Liu

Liang

et al. 2015

Journal of Hydrology

View full text Add to dashboard Cite

Keywords: Monthly water balance model Model comparison Model selection a b s t r a c tMulti-model comparison can provide useful information for model selection and improvement. In this study, twelve monthly water balance models with different structures and various degree of complexity are compared in 153 catchments with different climatic conditions in China. Generally, the GR5M model has the best performance, followed by the GR2M and WBM model. We investigate the relations between model performance and catchment characteristics and find that the climatic characteristic of a catchment is the most important factor impacting model performance. The models have better performance in wet catchments than in dry catchments. Large differences of model performance exist in dry catchments and model users should pay attention to model selection in dry catchments. In addition, we analyze the model performances among different models and conclude that increasing the model complexity does not guarantee a better model performance. Simple models can achieve comparable or even better performance than complex models. For the monthly simulation of hydrological processes, a two-parameter model is sufficient to achieve a good result. Moreover, by comparing the impacts of evapotranspiration simulation and runoff generation simulation on model performance, we find that evapotranspiration simulation has limited influence on the model performance. We suggest model builders focus on runoff generation process rather than evapotranspiration process to improve the performance of a monthly hydrological model.

show abstract

“…In recent years, ANN are increasingly used for prediction and pattern recognition problems in various fields of water and environmental science and technology such as total ozone forecasting (Bandyopadhyay and Chattopadhyay 2007), sea level prediction (Altunkaynak 2007, Imani et al 2013, rainfallrunoff modeling (De Vos and Rientjes 2005, Kuok et al 2010, Nourani et al 2011), water quality prediction (Emamgholizadeh et al 2013). A comprehensive literature review about the application of ANN in river forecasting was presented by Abrahart et al (2012).…”

Section: Introductionmentioning

confidence: 99%

Successive-station monthly streamflow prediction using different artificial neural network algorithms

Mehr

Kahya

ŞAHİN

et al. 2014

Int. J. Environ. Sci. Technol.

View full text Add to dashboard Cite

In this study, applicability of successive-station prediction models, as a practical alternative to streamflow prediction in poor rain gauge catchments, has been investigated using monthly streamflow records of two successive stations on Ç oruh River, Turkey. For this goal, at the first stage, based on eight different successive-station prediction scenarios, feed-forward back-propagation (FFBP) neural network algorithm has been applied as a brute search tool to find out the best scenario for the river. Then, two other artificial neural network (ANN) techniques, namely generalized regression neural network (GRNN) and radial basis function (RBF) algorithms, were used to generate two new ANN models for the selected scenario. Ultimately, a comparative performance study between the different algorithms has been performed using Nash-Sutcliffe efficiency, squared correlation coefficient, and root-meansquare error measures. The results indicated a promising role of successive-station methodology in monthly streamflow prediction. Performance analysis showed that only 1-month-lagged record of both stations was satisfactory to achieve accurate models with high-efficiency value. It is also found that the RBF network resulted in higher performance than FFBP and GRNN in our study domain.

show abstract

Constraints of artificial neural networks for rainfall-runoff modelling: trade-offs in hydrological state representation and model evaluation

Cited by 160 publications

References 43 publications

Assessment of the impacts of nonstationarity on watershed runoff using artificial neural networks: a case study in Ardebil, Iran

Assessment of the impacts of nonstationarity on watershed runoff using artificial neural networks: a case study in Ardebil, Iran

Comparison of performance of twelve monthly water balance models in different climatic catchments of China

Successive-station monthly streamflow prediction using different artificial neural network algorithms

Contact Info

Product

Resources

About