Mapping and Assessing Seagrass Bed Changes in Central Florida's West Coast using Multitemporal Landsat TM Imagery

Document Type

Article

Publication Date

8-2014

Keywords

seagrass change, submerged aquatic vegetation (SAV), image preprocessing, remote sensing, Landsat TM imagery, hurricane, Florida

Digital Object Identifier (DOI)

https://doi.org/10.1016/j.ecss.2014.07.014

Abstract

Some seagrass meadows in coastal shallow waters have displayed large scale changes in seagrass spatial extent and hurricanes and/or tropical storms have been suggested as factors responsible for reduction in coverage. Taking advantage of the incidence of three tropical storms passing near a study site along the central west Florida coast within a two-month period in 2004, we evaluated whether satellite remote sensing techniques (Landsat 5 Thematic Mapper (TM) imagery) are useful for assessing dynamics of seagrass (=submerged aquatic vegetation: SAV) cover/abundance in response to these multiple disturbances. We also examined whether an image preprocessing procedure, which included water column correction, applied to the Landsat TM images could further improve the classification and mapping of detailed SAV coverage. We compared a historical set of Landsat TM images, acquired in Fall 2003 and Fall and late Summer 2005, which were processed to classify %SAV cover into five classes using a maximum likelihood classifier. Importantly, our experimental results demonstrated that the application of the image preprocessing procedures led to an overall accuracy 2–14% improvement in SAV classification due to water column correction compared to that currently reported in the literature when similar Landsat TM data are utilized. Based upon the classification results mapped from the TM images and as well as a similar classification of SAV interpreted from aerial photographs collected before and after the passage of these same storms, SAV coverage over the study areas was found to increase about 6% (integrating SAV losses and gains) by 2005/2006 in comparison to cover levels present prior to the repeated storm activity. We conclude that heavy rains during 2004 along with physical disturbance from gale force winds from the tropical storms/hurricanes did not produce any SAV bed loss at the study site that was sustained for more than one year after multiple storm passage.

Was this content written or created while at USF?

Yes

Citation / Publisher Attribution

Estuarine, Coastal and Shelf Science, v. 149, p. 68-79

Share

COinS