Search

This paper reports the development of a neurocomputingbased model for estimating the potential evapotranspiration over Gangetic West Bengal, India during the summer monsoon months of June, July and August. An artificial neural network is implemented in the form of multilayer perceptron to generate the model. Three weather variables, surface temperature, vapour pressure and rainfall are used as the independent variables in generating the model. The performance of the model is judged statistically against nonlinear regression in the form of asymptotic regression. The study reveals that an artificial neural network is more efficient than the regression approach to estimate the potential evapotranspiration in the summer monsoon months. Furthermore, it is established that the artificial neural network and nonlinear regression have almost equal efficiency in the previously mentioned estimation in the month of June. However, in July and August the higher values of correlation and Willmotts indices, and lower values of estimation error, indicate that the artificial neural network is more reliable than the nonlinear regression approach. Since evapotranspiration is one of the basic components of the hydrological cycle and is essential for estimating irrigation water requirement, an efficient estimation procedure may help in agrometeorological modelling and irrigation scheduling in the summer monsoon months, which are of high importance for agriculture in the study zone. Copyright © 2009 Royal Meteorological Society [ABSTRACT FROM AUTHOR]

Water resources management and agriculture planning models require a statistical synthesis of some rainfall features, in particular those representing dry atmospheric conditions. The bigger the basin, the more important these features become, as is the case of the La Plata Basin LPB.This paper focuses on the precipitation variability in the large LPB in South America, analysing the number of months per year with low rainfall and the sequences of months with low rainfall, their theoretical distributions and stability, which are needed as input for the models mentioned above.Long time series are used to analyse the lowfrequency variability and the relative importance of decadal variability. Changes are evident in the number of months per year with low rainfall, with a decrease of about 20 in the period after 1970.Theoretical distribution models binomial and geometric are fitted to these empirical distributions, and the regional variability of the fitting parameters is shown. In practically the entire region, the goodnessoffit of the two theoretical models considered is statistically satisfactory.The temporal variability of the parameters of the theoretical binomial p and geometric α distributions is analysed, in excluding subperiods of 10 and 5 years, respectively. The results show lowfrequency variability overlapped on a decadal variability, with low homogeneous regional behaviour.The distribution models have proven to be efficient for frequency adjustments of the rainfall properties studied. These results are an acceptable and necessary input to decision models in LPB. They also make it possible to infer effects of climate change. Copyright © 2008 Royal Meteorological Society [ABSTRACT FROM AUTHOR]