Resource management of forested wetlands: Hurricane impact and recovery mapped by combining Landsat TM and NOAA AVHRR data

Abstract

A temporal suite of NOAA Advanced Very High Resolution Radiometer (AVHRR) images, transformed into a vegetation biomass indicator, was combined with a single-date classification of Landsat Thematic Mapper (TM) to map the association between forest type and hurricane effects. Hurricane effects to the forested wetland included an abrupt decrease and subsequent increase in biomass. The decrease was associated with hurricane impact and the increase with an abnormal bloom in vegetation in the impacted areas. Impact severity was estimated by differencing the biomass maps before and immediately (3 days) after the hurricane. Recovery magnitude was estimated by differencing the biomass maps from immediately (3 days) after and shortly (1.5 months) after the hurricane. Regions of dominantly hardwoods suffering high to moderate impacts and of dominantly cypress-tupelos suffering low impacts identified in this study corroborated findings of earlier studies. Conversely, areas not reported in previous studies as affected were identified, and these areas showed a reverse relationship, i.e., highly impacted cypresstupelo and low or moderately impacted hardwoods. Additionally, generated proportions of hardwood, cypress-tupelo, and open (mixed) forests per each 1-km pixel (impact and recovery maps) suggest that regions containing higher percentages of cypress-tupelos were more likely to have sustained higher impacts. Visual examination of the impact map revealed a spatial covariation between increased impact magnitudes and river corridors dominated by open forest. This spatial association was corroborated by examining changes in the percentage of open forest per 1-km impact pixel; the percentage of open forest peaked at moderate to high impacts. The distribution of recovery supported the impact spatial distribution; however, the magnitudes of the two indicators of hurricane effects were not always spatially dependent. Converse to univariate statistics describing a11 forested area within the basin, higher recoveries tended to be related to higher percentages of hardwoods. Lower recoveries, on the other hand, tended to be related to forests with nearly equal percentages of hardwoods and cypress-tupelo.