CLIMATE CHANGE ANALYSIS USING EARTH’S SURFACE TEMPERATURE: HIERARCHICAL TIME SERIES

Abstract

The concept of climate change encompasses the profound impacts of global warming on Earth's weather systems, with contemporary changes far exceeding historical variations and predominantly driven by human-induced factors such as elevated levels of atmospheric greenhouse gases from activities like fossil fuel combustion and agriculture. Efforts by organizations like the United Nations are actively combating these changes. Within Earth's climatic framework, land surface temperature plays a pivotal role, influencing crucial processes like energy and water exchange between the surface and atmosphere, thereby affecting vegetation growth patterns. Accurate comprehension of global and regional land surface temperatures, coupled with factors like vegetation and soil moisture, aids in evaluating land surface-atmosphere interactions and serves as a valuable metric for surface conditions.

This paper employs hierarchical time series forecasting to analyse and project land surface temperatures for major cities across countries. Hierarchical forecasting is essential when dealing with time series data aggregated hierarchically, ensuring a coherent approach to forecasting across different levels of granularity. By employing hierarchical time series forecasting, this research addresses the challenge of aggregating data to specific levels, ensuring consistency in temperature projections from city to country levels.