Analysis of temperature anomalies using machine learning, numerical methods and statistical techniques in global and Nepal datasets

Abstract

This study analyzes temperature anomalies to compare Nepal’s local trends with global patterns using Berkeley Earth data. Techniques like regression, Gaussian fitting, interpolation, moving averages, Rossler attractors, and spiral graphs revealed cyclical and chaotic climate behaviors. Results show a strong link between Nepal’s climate trends and global patterns, influenced by its unique geography and monsoon-driven climate. Attractor modeling provided new insights into underlying dynamics. The research highlights the importance of integrating local and global perspectives for understanding climate variability, offering valuable insights for regional adaptation and global climate policy, particularly for vulnerable regions like Nepal.