Global Warming In Ghana's Major Cities Based on Statistical Analysis of NASA's POWER Over 3-Decades

Joshua Attih

arXiv preprint · 20 August 2023

In one sentence. Linear trends and XGBoost analyse three decades of POWER temperature data in four Ghanaian cities, finding local warming, with Accra warming most.

Abstract

Global warming's impact on high temperatures in various parts of the world has raised concerns. This study investigates long-term temperature trends in four major Ghanaian cities representing distinct climatic zones. Using NASA's Prediction of Worldwide Energy Resource (POWER) data, statistical analyses assess local climate warming and its implications. Linear regression trend analysis and eXtreme Gradient Boosting (XGBoost) machine learning predict temperature variations. Land Surface Temperature (LST) profile maps generated from the RSLab platform enhance accuracy. Results reveal local warming trends, particularly in industrialized Accra. Demographic factors aren't significant. XGBoost model's low Root Mean Square Error (RMSE) scores demonstrate effectiveness in capturing temperature patterns. Wa unexpectedly has the highest mean temperature. Estimated mean temperatures for mid-2023 are: Accra 27.86°C, Kumasi 27.15°C, Kete-Krachi 29.39°C, and Wa 30.76°C. These findings improve understanding of local climate warming for policymakers and communities, aiding climate change strategies.

Labels

Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.