Date of Award
2015
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Atmospheric Science
Committee Chair
Robert Griffin
Committee Member
Andrew Molthan
Committee Member
Udaysankar S. Nair
Subject(s)
Satellite meteorology, Severe storms, Hail
Abstract
The central United States is primarily covered in agricultural lands with a growing season that peaks during the same time as the region's climatological maximum for severe weather. These severe thunderstorms can bring large hail that can cause extensive areas of crop damage, which can be difficult to survey from the ground. Satellite remote sensing can help with the identification of these damaged areas. This study examined three techniques for identifying damage using satellite imagery that could be used in the development of a near-real time algorithm formulated for the detection of damage to agriculture caused by hail. The three techniques: a short term Normalized Difference Vegetation Index (NDVI) change product, a modified Vegetation Health Index (mVHI) that incorporates both NDVI and land surface temperature (LST), and a feature detection technique based on NDVI and LST anomalies were tested on a single training case and five case studies. Skill scores were computed for each of the techniques during the training case and each case study. Among the best-performing case studies, the probability of detection (POD) for the techniques ranged from 0.527 - 0.742. Greater skill was noted for environments that occurred later in the growing season over areas where the land cover was consistently one or two types of uniform vegetation. The techniques struggled in environments where the land cover was not able to provide uniform vegetation, resulting in POD of 0.067 - 0.223. The feature detection technique was selected to be used for the near-real-time algorithm, based on the consistent performance throughout the entire growing season.
Recommended Citation
Bell, Jordan R., "The development of a near-real time hail damage swath identification algorithm for vegetation" (2015). Theses. 108.
https://louis.uah.edu/uah-theses/108