Predicting and forecasting disasters: A global scan of traditional and local knowledge
Predicting and forecasting disasters: A global scan of traditional and local knowledge
For centuries, people worldwide have predicted disasters based on observations and experimentation, interpreting animal behavior, plant responses, weather patterns, and celestial phenomena. Despite its significance, such traditional and local knowledge remains under-researched and undocumented, limiting its potential for integration in disaster risk reduction. This review consolidates traditional and local knowledge from 53 research articles, covering 423 cases categorized into three broad indicators: astronomical, meteorological, and biological. These indicators encompass 33 sub-indicators, such as sky, wind, birds, and soil, providing location- and disaster-specific contexts for prediction. Meteorological disasters (e.g., tropical cyclones, storms, and mass movements) constituted the largest share (33 %), followed by hydrological (e.g., floods and storm surges) and climatological disasters (e.g., droughts) (27 %). While there were disaster-specific variations, animal behavior (mammals, insects, birds, etc.) were the most commonly used predictive indicator (39 %), followed by water-related indicators (12 %), plant phenology (9 %), wind (8 %), and both cloud patterns and temperature (5 % each). Other indicators, including observations of the sun, moon, sky, stars, lightning, and rainfall, collectively constituted the remaining 22 %. There were notable similarities and differences in disaster prediction within and across countries in terms of the indicators used. It is, therefore, important to contextualize and localize prediction patterns rather than generalize them. However, scientific metrics need to be explored to assess their broader applicability. This would be a crucial step in harnessing traditional knowledge for integrating effective prediction methods, which requires increased funding and research efforts.
