INLAND WETLAND CHANGE DETECTION IN THE EVERGLADES WATER CONSERVATION AREA 2A USING A TIME-SERIES OF NORMALIZED REMOTELY-SENSED DATA

Recent and historical satellite remote sensor data were used to invent ory aquatic macrophyte (especially cattail and sawgrass) changes withi n the Florida Everglades Water Conservation Area 2A using Landsat Mult ispectral Scanner (MSS) data (1973, 1976, and 1982) and SPOT High Reso lution Visible (HRV) multispectral data (1987 and 1991). The method re quired a single base year of remotely sensed data with adequate ground reference information (1991). Historical remotely sensed data were '' normalized'' to the base year's radiometric characteristics. Statistic al clusters extracted from each date of imagery were found in relative ly consistent regions of multispectral feature space (using red and ne ar-infrared bands) and labeled using a ''core cluster'' approach. Wetl and classification maps of each year were analyzed using ''post-classi fication comparison'' change detection techniques to produce maps of ( 1) cattail change and (2) change in the ''sawgrass/cattaai mixture'' c lass. The amount of hectares in each wetland class was tabulated by ye ar. The spa tial distribution of the wetland was then overlaid onto a soil porewater phosphorus statistical surface obtained through in situ investigation. The cattail and cattail/sawgrass mixture classes appea r to be spatially associated with the distribution of relatively high concentrations of porewater phosphorus in Water Conservation Area 2A.