Temperature data of Hyderabad from the temperature of three neighboring cities using the ANN and the multiple regression methods

Abstract

The notion of this research was to find the climatological parameter of a location (whose data are notavailable) with the help of climatological parameters of its neighboring locations. In this study, we supposed Hyderabad as such a location; its neighboring cities Karachi, Badin, and Nawabshah form a triangle, and Hyderabad lies within the perimeter of the triangle. The task was to find the temperature distribution of Hyderabad using the temperatures of its neighboring cities. Two different methods, Artificial Neural Networks and multiple regression analysis, have been used to find the temperature of Hyderabad by using the temperature distribution of neighboring cities, Karachi,Nawabshah, and Badin. ANN with one hidden layer of 10 neurons is used, which connect three known temperature of Karachi, Nawabshah, and Badin to the output temperature of Hyderabad. Amultiple regression analysis with three independent variables, i.e., the temperatures of Karachi,Nawabshah, and Badin, are used to find a linear equation of multiple variables for the temperature of Hyderabad.Both ANN and multiple regression analysis produce an excellent result, and it is suggested toimplement the technique where direct measurement of the climatological data system is not present.