NDVI Change Analysis and Damage Mapping of the Vilonia, Arkansas Tornado, 27 April 2014. (2nd January 2017)

• Record Type: Journal Article
• Authors: Burow, Daniel; Rundquist, Bradley; Atkinson, Christopher
• Is Part Of: Papers in applied geography. Volume 3:Number 1(2017)
• Page Start: 85
• Page End: 100
• Publication Date: 2017-01-02
• Subjects: Landsat -- tornadoes -- NDVI; Applied human geography -- Periodicals; Geography -- Periodicals; 910
• DOI: 10.1080/23754931.2016.1264989
• Languages: English
• ISSNs: 2375-4931

Abstract

ABSTRACT: Normalized difference vegetation index (NDVI) change analysis is performed on Landsat-7 ETM+ imagery of an area in north-central Arkansas that was hit by an EF-4 tornado on 27 April 2014. The NDVI values of seventy-seven pixels containing National Weather Service damage reports from the tornado are analyzed and organized by the severity of the report on the Enhanced Fujita Scale. Decreases in NDVI are found to be common in pixels containing ≥ EF-2 damage. Furthermore, a strong correlation ( ) is observed between pretornado NDVI and NDVI change (and therefore posttornado NDVI) for a given pixel when only considering pixels containing EF-3 or EF-4 damage reports. Using the linear equation that describes this correlation, a simple model is derived with the purpose of capturing pixels that fit this pattern and therefore might have sustained ≥ EF-3 damage. The model correctly classifies seventy of seventy-five (93 percent; two pixels with negative pretornado NDVI were excluded) reports as either ≤ EF-2 or ≥ EF-3 intensity. If this correlation exists for other tornadoes, a similar model can be used to map the most intense damage for analysis, research, or other purposes.