Reconstruction of a tornado disaster employing remote sensing techniques: a case study of the 1999 Moore, Oklahoma tornado

Remote sensing has demonstrated to be an instrumental tool in monitoring land changes as a result of anthropogenic change or natural disasters. Most disaster studies have focused on large-scale events with few analyzing small-scale disasters such as tornadoes. These studies have only provided a damage assessment perspective with the continued need to assess reconstruction. This study attempts to fill that void by examining recovery from the 1999 Moore, Oklahoma Tornado utilizing Landsat TM and ETM+ imagery. Recovery was assessed for 2000, 2001 and 2002 using spectral enhancements (vegetative and urban indices and a combination of the two), a recovery index and different statistical thresholds. Classification accuracy assessments were performed to determine the precision of recovery and select the best results. This analysis proved that medium resolution imagery could be used in conjunction with geospatial techniques to capture recovery. The new indices, Shortwave Infrared Index (SWIRI) and Coupled Vegetation and Urban Index (CVUI), developed for disaster management, were the most effective at discerning reconstruction using the 1.5 standard deviation threshold. Recovery rates for F-scale damages revealed that the most incredibly damaged areas associated with an F5 rating were the slowest to recover, while the lesser damaged areas associated with F1-F3 ratings were the quickest to rebuild. These findings were consistent for 2000, 2001 and 2002 also exposing that complete recovery was never attained in any of the F-scale damage zones by 2002. This study illustrates the significance the biophysical impact has on recovery as well as the effectiveness of using medium resolution imagery such as Landsat in future research.