Spatial interpolation of weather variables using artificial neural networks

Crop growth simulation models use weather data such as temperature, solar radiation, and rainfall to simulate crop development and yield. The crop models are often needed for locations with missing or incomplete observed weather data. An accurate estimation of these weather variables has thus become necessary. Artificial neural network (ANN) models could be used to accurately estimate these weather variables. In this study, ANNbased methods were developed to estimate daily maximum and minimum air temperature and total solar radiation for locations in Georgia. Observed weather data from 1996 to 1998 were used for model development, and data from 1999 to 2000 were used for final ANN model evaluation. In the ANN model development, the preferred number of input weather stations and the input variables for estimating each weather variable were determined. The ANNs provided higher accuracy than the traditional average, inverse distance, and multi-linear regression methods.