Prediction of the likely impact of climate change on monthly mean maximum and minimum temperature in the Chaliyar river basin, India, using ANN-based models

Abstract

In this work, an approach based on Artificial Neural Networks (ANN) has been employed to assess the likely impact of climate change on mean monthly maximum and minimum temperature (T max and T min) in the Chaliyar river basin, Kerala, India. ANN is trained to downscale temperature from the General Circulation Model (GCM) from a coarser resolution to the required resolution of the river basin. The work aims to estimate the GCMs’ output to the scales compatible with that employed in a hydrologic model of the river basin. In order to satiate this purpose, predictor variables were obtained from the National Centre for Environmental Prediction and National Centre for Atmospheric Research (NCEP/NCAR) reanalysis data; this was utilized for training the ANN using a feed-forward network with a back-propagation algorithm. These models were validated further and used to downscale CGCM3 GCM simulations for the scenarios outlined in the IPCC Special Report on Emission Scenarios (SRES). Results showed that both T max and T min are increasing consistently in all the scenarios. T max exhibited an average increase of maximum 3 °C during the dry season (December–May) and 1 °C during the wet season (June–November) by the year 2100, while T min showed an average increase of 2.5 °C in the dry season and 0.5 °C in the wet season.