spring minimum temperature modeling of zab baisn with the neural network artifical

today one of the most important problems in agriclimatology in effecting of climate on products. one of the climate elements in freezing which according to the continuity , extend and amount of it on agricultural activities ,transportation and invironment has a lot of affections. least temperature forcasts as one of the important climatic elements to forcasting late coldness and freezing in spring is very important. one of the new ways in forcasting climate elements is artifical neural networks(anns). for this purpose from a 27 periodical time(1986-2012) synoptical station in zab baisn inmatlab software for aducating and testing the topic has been used. the best model for artifical neural network to forcasting least temperature is presptron with 3layers and 5 nerons in interal layer, 4nerons in hidden layer and one neron in output layer and marqwart-lonberg algoritm.