Time Series Modelling of Monthly Temperature Deviations Using ARIMA Techniques in Python

P. Ramakrishna Reddy, M. Naresh, S. Asif Alisha, P. Maheswari, A. Srinivasulu, Hema Sekhar Palla, B. Sarojamma, S. Venkatramana Reddy

Abstract: Monthly temperature deviations from the 1951 to 1980 baseline were examined for the period 2019 to 2023 using time series modelling in a Python environment. Several ARIMA specifications, including ARIMA (1,0,0), ARIMA (0,1,1), and ARIMA (1,1,1), were estimated to understand patterns in atmospheric variability and to identify a model that captures temporal dependence effectively. Model adequacy was evaluated through Akaike and Bayesian information criteria, along with autocorrelation and partial autocorrelation diagnostics and Ljung?Box testing. Descriptive statistics, seasonal grouping, heat mapping, and rolling mean analysis supported the empirical assessment of monthly fluctuations. Comparative evaluation using root mean square error indicated that ARIMA (1,0,0) provided the most reliable representation of the observed temperature deviations. The findings contribute to statistical forecasting practices for climatic variables and demonstrate the applicability of computational tools in environmental data analysis.

Keywords: temperature deviation, ARIMA modelling, time series forecasting, atmospheric variability, Python analysis

How to Cite?: P. Ramakrishna Reddy, M. Naresh, S. Asif Alisha, P. Maheswari, A. Srinivasulu, Hema Sekhar Palla, B. Sarojamma, S. Venkatramana Reddy, "Time Series Modelling of Monthly Temperature Deviations Using ARIMA Techniques in Python", Volume 15 Issue 3, March 2026, International Journal of Science and Research (IJSR), Pages: 1526-1532, https://www.ijsr.net/getabstract.php?paperid=SR26318141528, DOI: https://dx.dx.doi.org/10.21275/SR26318141528

Download Citation: APA | MLA | BibTeX | EndNote | RefMan