Multi-sensor examination of hail damage swaths for near real-time applications and assessment

Abstract

Severe thunderstorms occurred across portions of South Dakota, Nebraska, Iowa, and Missouri on 18 August 2011, resulting in several swaths of large hail that caused significant damage to crops toward the end of the growing season. Changes with time in the normalized difference vegetation index (NDVI) and truecolor imagery from the Moderate Resolution Imaging Spectroradiometer instrument aboard the National Aeronautics and Space Administration's Terra satellite identify the damaged area, with corroboration by available radar data. Higher resolution satellite imagery from the public and private sector further identify both damage and NDVI changes within individual fields of planted crops that are susceptible to hail damage. Similar imagery can be developed for near real-time applications seeking to identify crop-damage areas during the growing season.

Evaluation of approaches to identifying hail damage to crop vegetation using satellite imagery

“…Molthan et al performed a similar analysis as Gallo et al (2012) but also incorporated observations with higher spatial resolutions from addi-tional earth-observing instruments, including Landsat-7 Enhanced Thematic Mapper Plus (30-m spatial resolution), Terra Advanced Spaceborne Thermal Emission and Radiometer (15 m), Système Pour l'Observation de la Terre-4 (SPOT-4, 20 m), and SPOT-5 (10 m). Molthan et al (2013) obtained similar results to those of Gallo et al (2012) with respect to the correlations between NDVI change and maximum estimated size of hail (MESH) size. Molthan et al (2013) used the higher spatial resolution sensors to confirm fieldscale damage and NDVI reductions in areas that corresponded to the MESH hail-fall streaks and coarserresolution MODIS observations.…”

“…Molthan et al (2013) obtained similar results to those of Gallo et al (2012) with respect to the correlations between NDVI change and maximum estimated size of hail (MESH) size. Molthan et al (2013) used the higher spatial resolution sensors to confirm fieldscale damage and NDVI reductions in areas that corresponded to the MESH hail-fall streaks and coarserresolution MODIS observations. This study builds on previous work by comparing damage identified by NDVI change (thresholding) and an image-processing technique (anomaly detection) to identify areas of hail damage in Aqua MODIS imagery.…”

“…Previous studies focused on identifying areas of damage by identifying NDVI change from a pre-event composite and post-event image (Parker et al 2005;Jedlovec et al 2006;Gallo et al 2012;Molthan et al 2013). In this study, an NDVI change product (threshold hereafter) is developed from the methods described by Molthan et al (2013) and Gallo et al (2012).…”

“…When the threshold for radar-estimated hail size was increased to >2.54 cm (1 in) and >5.08 cm (2 in), the percentage of pixels with a decrease in NDVI increased to 87% and 92%, respectively. Molthan et al (2013) examined hail swaths produced from severe thunderstorms that occurred on 18 August 2011 across extreme southeastern South Dakota, eastern Nebraska, southwestern Iowa, and northwestern Missouri. Molthan et al performed a similar analysis as Gallo et al (2012) but also incorporated observations with higher spatial resolutions from addi-tional earth-observing instruments, including Landsat-7 Enhanced Thematic Mapper Plus (30-m spatial resolution), Terra Advanced Spaceborne Thermal Emission and Radiometer (15 m), Système Pour l'Observation de la Terre-4 (SPOT-4, 20 m), and SPOT-5 (10 m).…”

“…Such an algorithm will assist in mapping hail damage following severe thunderstorms. The NDVI change threshold is based on the difference between a preevent NDVI composite and post-event single-day NDVI image, as used previously by Gallo et al (2012) and Molthan et al (2013). The anomaly detection technique uses single-day NDVI and LST anomalies to identify potential areas of damage.…”