Land Cover Influence on Detection of Hail Swaths Using GOES Advanced Baseline Imager

In rural areas, verification of damage caused by severe hail events can be difficult because of the lack of observer reports. Satellite data provides an alternative means to detect damage from hail swaths. By using Advanced Baseline Imager (ABI) data from Geostationary Operational Environmental Satellites (GOES), we calculated the normalized difference vegetation index (NDVI) change of the damaged area by finding the difference of NDVI measured before the event to NDVI measured roughly seven days after the event. The observed hail swaths can cover tens to hundreds of kilometers as well as pass over a variety of vegetation. Generally, the ABI-based NDVI change was found to indicate areas of hail damage. However, we also found that NDVI change needs to be analyzed by the individual vegetation type when examining hail swaths that have many different types of vegetation. The error in estimated areal coverage of hail damage was greater without analysis of individual vegetation types due to variations in NDVI change associated with vegetation type.The analysis also found that certain vegetation had statistically insignificant differences with NDVI change in areas where there was no hail (maximum expected size of hail (MESH) equal to zero) compared to areas with MESH values above the National Weather Service criteria for severe hail of 25.4 mm. Because of these responses, it is important to analyze the ABI NDVI change process based on vegetation type so that damage from hail swaths is correctly identified.