Isaacs, Rachel E. (2007-08). Spatial patterns of ice storm disturbance in the forested landscape of Ouachita Mountains, Arkansas and Oklahoma. Master's Thesis. Thesis uri icon


  • Large-extent ice storms have received relatively little attention from researchers. This research investigates the effects of abiotic and biotic factors on the spatial patterns of ice storm disturbance on a forested landscape. This investigation provides a landscape-level perspective on the impacts of ice storm disturbance, clarifies the effects on ecosystem dynamics, and will aid future forest management plans. The study was conducted in Ouachita National Forest (ONF) in west-central Arkansas and southeastern Oklahoma and examined approximately 6000 km2 of forest between 150 and 800 m elevation. Normalized Difference Vegeation Index (NDVI) difference values were calculated using two Landsat 7 ETM+ scenes to identify NDVI changes that potentially were associated with ice storm damage to the forests. Forty-six geolocated field sites were used to determine the relationship of NDVI difference to actual forest damage caused by the ice storm by counting the number of downed tree boles intersecting a 100 m transect. These field sites encompassed a broad range of each of the physical variables (i.e. elevation, slope, and aspect), forest type, and degree of damage. The linear regression model determined the relationship between NDVI difference and ice storm damage. Elevation, slope, and aspect were calculated based on individual pixels from the DEM. Categories of forest damage were based on NDVI difference values. A chi-square test of correspondence and Cramer's V test were then used to analyze relationships of damage to abiotic and biotic variables. The strong, negative relationship observed in the linear regression model suggested that NDVI was representative of ice storm damage in the study area. The chi-square test of correspondence indicated the abiotic and biotic variables all had associations with NDVI difference results (p<0.001). The Cramer's V test established that elevation had the strongest influence on the degree of ice storm damage followed closely by slope and aspect. Moderate elevations, moderate slopes, and windward aspects received the highest percentage of major storm damage. Forest type displayed a weak relationship with the extent of damage. The topographic patterns of ice storm damage are similar to patterns found in previous research. Topography influenced spatial patterns of ice storm damage. Elevation, slope, and aspect were all found to be important variables influencing the degree of ice storm damage. Knowledge concerning these spatial patterns is critical for future studies of ecosystem dynamics and forest management practices.

publication date

  • August 2007