Abstract

In field water balance studies, one of the major difficulties is the separation of evapo-transpiration into plant transpiration and soil evaporation. In this paper, the radial basis function (RBF) neural network was implemented using C language to estimate daily soil water evaporation from average relative air humidity, air temperature, wind speed and soil water content in a cactus field study. The RBF neural network learned rapidly and converged after about 1000 training iterations. The optimum number of hidden neurons was found to be six. The RBF neural network achieved good agreement between predicted and measured values. The average absolute percent error and the root mean squared error was 21·0% and 0·17 mm for the RBF neural networkvs. 30·1% and 0·28 mm for the multiple linear regression (MLR). The RBF neural network technique appears to be an improvement over the MLR technique for estimating soil evaporation.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call