Spatial and temporal analysis of natural hazard mortality in Nepal

Abstract

The impacts of natural hazards are typically measured in terms of loss of human lives and economic damage, and recent studies demonstrate that deaths attributed to natural hazards have increased. Using the publicly available DesInventar database, we examined spatial and temporal patterns of natural hazard mortality from 1971 to 2011 at the district and village levels of Nepal and identified natural hazards that contributed most to mortality. Spatial clusters of mortality at the district and village levels were detected using local and global spatial autocorrelation measures (Moran's I). Landslides (41.91%) and floods (32.52%) accounted for approximately three quarters of natural hazard mortalities over the study period. A Global Moran's I test positively confirmed clustering at both the district (0.199, p < .001) and village (0.256, p < .001) levels, whereas a Local Moran's I test further detected clustering in the central and terai regions, where dynamic geologic and geomorphic processes combined with human-environment interaction constitute major risk factors. A better understanding of multihazard mortality patterns across geographic landscapes and time has the potential to aid policy makers, planners, and local officers to more efficiently allocate scarce capital and human resources to reduce mortality.