Use of NDVI from Satellite Imagery to Assess Tornado Tracks in Crops.
Abstract
There are numerous tornadoes that occur each year in Canada. However, many of them occur in Canada’s croplands, which means they cannot be classified because the Enhanced Fujita (EF) scale is a damage-based assessment which relies on the tornado hitting damage indicators whose wind loads have been well-documented, which currently does not include crops. Not knowing the true intensities of these tornadoes results in an inaccurate tornado climatology for the Prairie regions of Canada (Alberta, Saskatchewan, and Manitoba). Therefore, the goal of this paper is to better understand Canada’s tornado climatology by developing a method to use a Normalized Difference Vegetation Index (NDVI) from multispectral satellite imagery to assess characteristics of a tornado such as path length, path width, and intensity. To accomplish this, multispectral satellite imagery collected by the Northern Tornadoes Project of all Canadian tornadoes across the Prairie regions of Canada between 2017-2023 was analyzed. The results indicate that the methodology is able to easily identify damage caused by EF4 and EF5 tornadoes, occasionally able to identify damage caused by EF2 and EF3 tornadoes, and unable to identify damage from EF0 and EF1 tornadoes. Additionally, a weak relationship between the observed maximum width of crop damage and the maximum width of the tornado was identified. Finally, it was noted that solely looking at crop damage, it is difficult to determine the full path of a tornado due to fluctuations of tornado intensity and land cover.