Narx Neural Networks Models for Prediction of Standardized Precipitation Index in Central Mexico

Some of the effects climate change may be related to a in patterns rainfall intensity or scarcity. Therefore, humanity is facing environmental challenges due an increase occurrence and droughts. The forecast droughts can great help when trying reduce adverse that scarcity water brings, particularly agriculture. When evaluating conditions scarcity, as well identification characterization droughts, use predictive models drought indices could very useful tool. In this research, utility Artificial Neural Networks with exogenous inputs was tested, aim predicting monthly Standardized Precipitation Index 4 regions (Semi-desert, Highlands, Canyons Mountains) north-central México using predictor data from 1979 2014. best model found scaled conjugate gradient backpropagation algorithm optimization method set following architecture: 6-25-1 network. correlation coefficient predicted observed values for test dataset between 0.84 0.95. As result, Network performed successfully at four analyzed regions. developed tested research suggest remarkable prediction abilities study region.