Abstract
This paper asks if meteorological data can help predict instances of drought and flood in the ethnographic record, with the goal of being able to use such data to more easily evaluate the effect of climate/weather on cultural beliefs, practices, and attitudes. The authors develop several indices of droughts, floods, and extreme weather and test them by applying data from weather stations in close proximity to SCCS societies and comparing the predictions to ethnographic accounts of natural hazards. Weather data included daily precipitation, minimum temperature, and maximum temperature dating back as far as the 1950s. The authors found that precipitation - evaporation (P-E) measures and Palmer Drought Severity Index (PDSI) overpredicted droughts, whereas a combined drought measure underpredicted droughts, and that the flood indices generally were not very useful in predicting flood events. The authors suggest using P-E measures and PDSI to hindcast the presence of droughts, and the combined drought measure to hindcast the absence of droughts. They suggest that the overprediction of both floods and droughts is a result of extreme weather not necessarily having a severe impact on crop yields.