Geospatial Assessment of Recovery Rates Following a Tornado Disaster

Abstract

Remote sensing has proven to be instrumental in monitoring land alterations from natural disasters and anthropogenic processes. Additionally, geospatial analyses of tornado disasters have provided damage assessments, whereas hazard research has limited recovery evaluation to economic and migration perspectives. This study examines recovery from the 1999 Moore, OK, tornado disaster across three consecutive years, utilizing medium-resolution imagery and a series of image-processing algorithms. Spectral enhancements including normalized difference vegetation index, soil-adjusted vegetation index, urban index (UI), and two new indices, i.e., ShortWave Radiation Index (SWIRI) and Coupled Vegetative Urban Index (CVUI), were utilized in conjunction with a recovery index and statistical thresholds to assess recovery. Classification accuracy assessments prove that geospatial techniques and medium-resolution imagery can capture the rate of recovery with the most effective results noted with SWIRI using the 1.5 standard deviation threshold. Computed annual and Fujita Scale recovery rates indicate that 1) the most severely damaged areas associated with an F5 rating were the slowest to recover whereas the lesser damaged areas (F1-F3) were the quickest to rebuild and 2) complete recovery was never attained, even three years after the event, regardless of the F-scale damage zones. Recovery appears to be a significant and direct function of the level of damage sustained. With these results, decision makers and other policyholders could implement more resilient approaches in reconstructing the more severely damaged areas.