Salinity mapping for Florida bay using Landsat TM images and in-situ observations

Authors:
Publisher website
Google Scholar
Abstract
The salinity of Florida Bay has been increasing for decades due to anthropogenic interventions and natural disturbances of natural sheetflow through the Everglades and adjacent natural systems. Yet little data has been gathered and geospatially analyzed. This study developed a set of multivariate hydro-spatial statistical models to predict the dry and wet season surface salinity concentrations in Florida Bay. The modeling was carried out using in-situ salinity observations from USGS for 2004-2006, and processed Landsat TM 4-5 images for the closest 21 days. To improve the accuracy, the study area was clustered in 3 sub-regions leading to 6 cases (3 sub-regions * 2 seasons). The statistical model produced R2 values ranging from 0.53 to 0.83, which are acceptable considering the large size and geographic complexity of the study domain. The two salinity maps created for the Florida Bay area provide a visual representation of the temporal and spatial variation between seasons.

This publication has 3 references indexed in Scilit: