Use of Landsat data to track historical water quality changes in Florida Keys marine environments

Satellite remote sensing has shown the advantage of water quality assessment at synoptic scales in coastal regions, yet modern sensors such as SeaWiFS or MODIS did not start until the late 1990s. For non-interrupted observations, only the Landsat series have the potential to detect major water quality events since the 1980s. However, such ability is hindered by the unknown data quality or consistency through time. Here, using the Florida Keys as a case study, we demonstrate an approach to identify historical water quality events through improved atmospheric correction of Landsat data and cross-validation with concurrent MODIS data. After aggregation of the Landsat-5 Thematic Mapper (TM) 30-m pixels to 240-m pixels (to increase the signal-to-noise ratio), a MODIS-like atmospheric correction approach using the Landsat shortwave-infrared (SWIR) bands was developed and applied to the entire Landsat-5 TM data series between 1985 and 2010. Remote sensing reflectance (Res) anomalies from Landsat (2 standard deviations from a pixel-specific monthly climatology) were found to detect MODIS Res anomalies with over 90% accuracy for all three bands for the same period of 2002-2010. Extending this analysis for the entire Landsat-5 time-series revealed Res anomaly events in the 1980s and 1990s, some of which are corroborated by known ecosystem changes due in part to changes in local freshwater flow. Indeed, TM Res anomalies were shown to be useful in detecting shifts in seagrass density, turbidity increases, black water events, and phytoplankton blooms. These findings have large implications for ongoing and future water quality assessment in the Florida Keys as well as in many other coastal regions. (C) 2013 Elsevier Inc. All rights reserved.