Exploring how extreme weather events, natural disasters and climate change are reported in online news articles for countries with differing climate baselines

Abstract

Online news article analysis has been a field of research that has been undergoing particular attention by the computer and data science disciplines. Research on extreme weather and natural disaster news coverage has been confined to new machine learning and natural language processing techniques to detect different categories, and events. This research has been lacking in analyzing the long term trends in online news articles as they relate to real world conditions. This study conducts a quantitative and qualitative analysis of news coverage about extreme weather events and natural disasters between two countries, Colombia and South Africa, with similar economic and development conditions but differing climate baselines. Overall this study finds that precipitation levels alone are not sufficient, and can be misleading, in identifying and understanding what and how extreme weather events and natural disasters occur. Instead a quantitative and qualitative analysis of the news coverage of such events provides a more nuanced and comprehensive timeline of these extreme weather events and natural disasters, for both high and low precipitation baseline climates.