An investigation of topographic influence on tornades in Georgia

The latitude and longitude of 967 tornadogenesis points in Georgia have been extracted from the NCDC Storm Reports from 1950-1998. When plotted on a map of the state there is a suggestion visually that some regions of Georgia are more favored for tornadogenesis while others are less favored. Some of these areas approximate broad topographic features of the state. This investigation is an effort to explore the degree to which terrain explains the observed patterns.

Terrain parameters have been calculated for 5km by 5km grids evenly placed across the state. A Digital Elevation Model (DEM) has been used to extract terrain parameters for each grid. The total number of tornadogenesis points for each grid was also calculated. The terrain attributes include: 1) Mean elevation, 2) Range of elevation, 3) Maximum elevation, 4) Minimum elevation, 5) Standard deviation from mean elevation (a measure of surface roughness), 6) The first derivative of standard deviation from mean elevation with respect to distance.

Various linear and non-linear statistical techniques were applied to the data set to investigate the relationship between the terrain parameters and the frequency number of tornadogenesis for each grid. All terrain parameters were found to have statistically insignificant correlations with tornadogenesis, with the exception of the standard deviation of mean elevation. The surface roughness parameter has an inverse correlation with the total number of tornadogenesis points in each grid, indicating smoother terrain is more favorable to tornadogenesis. Application of several non-parametric techniques has quantified this inverse correlation and has revealed statistical significance.

This study suggests that the roughness of terrain may be significant for the genesis of tornadoes. Future research examining distribution in other states and pinpointing exact tracks of tornadoes may reveal more terrain influence.