Estimating Daily Maximum Temperatures Using Artificial Networks (Case study: Kerman)

Considering the capability of the artificial neural networks in simulating sophisticated processes, it is being used in estimation and computation of climatic parameters. The goal of this research is to estimate the daily maximum temperature in Kerman province. To this aim, daily climatic parameters as input to the neural networks and daily maximum temperature as the output during a statistical period of 24 years (1989-2013) were used, the findings revealed that the output of the multi-layer perceptron neural network, considering the error amount and correlation among data, is more precise and shows lower error and more correlation in relation to the expected output (daily maximum temperature). Also, among other climatic parameters, minimum temperature and the average of the wet temperature indicated the estimation of the daily maximum temperature with lower error and more correlation in comparison to other climatic parameters.