Utilizing Remotely Sensed Data & Elementary Analytical Techniques in Post-Katrina Mississippi to Examine Storm Damage Modeling

Hurricane Katrina's passage through south Mississippi on August 29, 2005, which damaged or destroyed thousands of hectares of forestland, was followed by massive salvage, cleanup, and assessment efforts. An initial assessment by the Mississippi Forestry Commission estimated that over $1 billion in raw wood material was downed by the storm, with county-level damage percentages ranging from 50 percent to 60 percent across Mississippi's three coastal counties. Remotely sensed data were used to provide a more complete picture of the damage inflicted by Katrina. Moderate (56-29-m resolution) and high (1-0.3-m resolution) resolution data were acquired from spaceborne and airborne platforms in natural color and MultiSpectral (MS) formats. Transformed data such as Normalized Difference Vegetation Index (NDVI) and Normalized Difference Moisture Index (NDMI), along with damage estimates obtained by interpreting aerial photography, were also used as variables in a linear modeling process. This continuous damage prediction process demonstrated the effect of incorporating forest condition thematic information, prestorm moderate resolution imagery with transforms, and poststorm moderate resolution imagery with transforms. The resulting models, all of which used a large number of regressors, had overall fit values of R²_adj=0.708 and RMSE=0.130 with all variable types used, R²_adj=0.492 and RMSE=0.172 with all variables except the forest condition data, and R²_adj=0.599 and RMSE=0.153 with all variables except the poststorm imagery data.