Forecast Modeling and Trend Analysis of Temperature in India

Abstract

Since historical times, the climate of the earth has been changing. In terms of diversity and population, India is a vast country. Any significant change in its climatic conditions can have an impact on the country’s entire economic and cultural structure. Temperature is the most important variable when focussing on climate change study. This study aims to recognise the changing trend of temperature in India. For this purpose, temperature data from 1901 to 2020 has been analysed. To find out the variability in temperature, various statistical tools were applied. The Mann-Kendall method has been used along with Sen’s slope estimator for determining the trend and slope magnitude. ARIMA modelling is applied to the data series to forecast the temperature trend. The study has shown that variability in temperature is increasing in recent years and the temperature trend is increasing in all the months, especially in the cold months of November, December and January. The average annual temperature of India in the year has increased significantly in the past 120 years. The trend analysis of temperature is tested significantly statistically at a 95% confidence level.