Abstract
The Normalized Difference Vegetation Index (NDVI) is one of the most widely used vegetation indexes to measure and monitor plant growth, vegetation cover, and biomass production. A variety of different multispectral satellite systems could provide NDVI time series to examine the long-term vegetation process patterns. The effects of hurricanes on forested ecosystems can range from very minor defoliation of only a few trees to catastrophic blow-down of whole forest. During 2005, two hurricanes hit South Florida, Hurricane Katrina on Aug 25th and Hurricane Wilma on Oct. 24th. Several studies have evaluated the impact of hurricanes on mangroves, but relatively few have analyzed the rates and patterns of recovery from disturbances of mangrove forest and other vegetation using a NDVI time series. This thesis uses NDVI time series from imagery 2001 to 2010 with an 8-day interval derived from the MODIS Terra to detect the recovery rate of mangrove after the two major hurricanes in South Florida. The slope obtained from Ordinary Least Square (OLS) is used to analyze the recovery rates in this thesis. The slope values from OLS were compared between pre-hurricane and post-hurricane periods to show that the most rapid recovery rates were happening in the first years after the 2005 hurricanes events. This research showed that it took around 2 to 3 years for vegetation to recover to normal level in terms of phenological rhythms.