Potential of Gini Index to Predict Extreme Events in the Contiguous United States

The precipitation pattern along with the climate change projection across the Contiguous United States (CONUS), has exhibited an increase in extreme events like droughts and floods over the last decade. The occurrence of these events caused not only severe damage to human life but also pivoted significant economic damage in this region. This study considers Q95 (the 5-percentile flow) as an extreme flood condition and Q5(the 95-percentile flow) as an extreme hydrological drought condition for 1250 USGS discharge gauges across CONUS. We considered the daily concentration of these quantiles in terms of the Gini Index to describe how uniformly these extremes are distributed throughout each pentad from 1995-2020. Here, we present observational evidence from existing extreme events databases and insurance data, investigating the regional patterns of discharge extremes and excessive rainfall and drought impacts. The result demonstrates that the Gini Index can capture many high impact events. However, it also limits missing some moderate to low events. Especially in important geographic regions like Texas and California, where there were numerous reported incidents, the Gini Index could not capture some impactful events as the study could not find enough continuous stations in these large regions. The spatial trend shows better predictability in the northeast region(e.g., Mid-Atlantic and New England Flood in 2005). The applicability of this method is vast, as the framework can be used under different future climate and discharge projection scenarios to project a baseline forecast for the policymakers and stakeholders.