Classification of rainfall variability by using artificial neural networks

In this paper, the usefulness of artificial neural networks (ANNs) as a sui table tool for the study of the medium and long-term climatic variability i s examined. A method for classifying the inherent variability of climatic d ata, as represented by the rainfall regime, is investigated. The rainfall r ecorded at a climatological station in Cyprus over a long time period has b een used in this paper as the input for various ANN and cluster analysis mo dels. The analysed rainfall data cover the time span 1917-1995. Using these values, two different procedures were followed for structuring the input v ectors for training the ANN models: (a) each 1-year subset consisting of th e 12 monthly elements, and (b) each 2-year subset consisting of the 24 mont hly elements. Several ANN models with a varying number of output nodes have been trained, using an unsupervised learning paradigm, namely, the Kohonen 's self-organizing feature maps algorithm. For both the 1- and 2-year subse ts, 16 classes were empirically considered as the optimum for computing the prototype classes of weather variability for this meteorological parameter . The classification established by using the ANN methodology is subsequent ly compared with the classification generated by using cluster analysis, ba sed on the agglomerative hierarchical clustering algorithm. To validate the classification results, the rainfall distributions for the more recent yea rs 1996, 1997 and 1998 were utilized. The respective 1- and 2-year distribu tions for these years were assigned to particular classes for both the ANN and cluster analysis procedures. Compared with cluster analysis, the ANN mo dels were more capable of detecting even minor characteristics in the rainf all waveshapes investigated, and they also performed a more realistic categ orization of the available data. It is suggested that the proposed ANN meth odology can be applied to more climatological parameters, and with longer c ycles. Copyright (C) 2001 Royal Meteorological Society.