Monitoring vegetation response to episodic disturbance events by using multitemporal vegetation indices

Journal of Coastal Research
By: , and 

Links

Abstract

Normalized Difference Vegetation Index (NDVI) derived from MODerate-resolution Imaging Spectroradiometer (MODIS) satellite imagery and land/water assessments from Landsat Thematic Mapper (TM) imagery were used to quantify the extent and severity of damage and subsequent recovery after Hurricanes Katrina and Rita of 2005 within the vegetation communities of Louisiana's coastal wetlands. Field data on species composition and total live cover were collected from 232 unique plots during multiple time periods to corroborate changes in NDVI values over time. Aprehurricane 5-year baseline time series clearly identified NDVI values by habitat type, suggesting the sensitivity of NDVI to assess and monitor phenological changes in coastal wetland habitats. Monthly data from March 2005 to November 2006 were compared to the baseline average to create a departure from average statistic. Departures suggest that over 33% (4,714 km2) of the prestorm, coastal wetlands experienced a substantial decline in the density and vigor of vegetation by October 2005 (poststorm), mostly in the east and west regions, where landfalls of Hurricanes Katrina and Rita occurred. The percentage of area of persistent vegetation damage due to long-lasting formation of new open water was 91.8% in the east and 81.0% and 29.0% in the central and west regions, respectively. Although below average NDVI values were observed in most marsh communities through November 2006, recovery of vegetation was evident. Results indicated that impacts and recovery from large episodic disturbance events that influence multiple habitat types can be accurately determined using NDVI, especially when integrated with assessments of physical landscape changes and field verifications.

Study Area

Publication type Article
Publication Subtype Journal Article
Title Monitoring vegetation response to episodic disturbance events by using multitemporal vegetation indices
Series title Journal of Coastal Research
DOI 10.2112/SI63-011.1
Issue 63
Year Published 2013
Language English
Publisher Coastal Education and Research Foundation
Contributing office(s) National Wetlands Research Center
Description 13 p.
First page 118
Last page 130
Country United States
State Louisiana
Google Analytic Metrics Metrics page
Additional publication details