Title: Time series forecasting of global temperature anomalies using machine learning

Authors: Ravi Patel; Aditya Kumar; Jainath Yadav; Mrityunjay Singh

Addresses: Department of Computer Science, Central University of South Bihar, Gaya, Bihar, India ' Department of Computer Science, Central University of South Bihar, Gaya, Bihar, India ' Department of Computer Science, Central University of South Bihar, Gaya, Bihar, India ' Department of Computer Science, Central University of South Bihar, Gaya, Bihar, India

Abstract: Accurate forecasting of global temperature anomalies is essential for understanding long-term climate variability and guiding climate change policies. This study compares classical machine learning (ML) and deep learning (DL) models for predicting global mean temperature anomalies using data from 1880-2024. The framework incorporates extensive pre-processing with several ML and DL architectures. Results show that ML models, especially SVR and Lasso, achieve higher accuracy and efficiency than DL approaches. Forecasts for 2025-2034 indicate a consistent warming trend, with DL models projecting anomalies surpassing 1.5°C by 2031, while regression-based methods estimate more moderate increases of 1.1-1.3°C.

Keywords: global temperature anomalies; climate forecasting; machine learning; deep learning; time series prediction; climate change.

DOI: 10.1504/IJGW.2026.155837

International Journal of Global Warming, 2026 Vol.39 No.4, pp.388 - 403

Received: 05 Sep 2025
Accepted: 27 Dec 2025
Published online: 24 Aug 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article