Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: A semiautomated approach for monitoring landscape changes in Texas seagrass beds from aerial photography.

Authors
item Fletcher, Reginald
item Pulich, Warren - TX. ST. UNIV.-SAN MARCOS
item Hardegree, Beau - U.S.FISH&WILDLF-CORPUS

Submitted to: Journal of Coastal Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: October 15, 2008
Publication Date: March 1, 2009
Citation: Fletcher, R.S., Pulich, W.M., Hardegree, B. 2009. A semiautomated approach for monitoring landscape changes in Texas seagrass beds from aerial photography. Journal of Coastal Research. 25:500-506.

Interpretive Summary: In support of the Texas Seagrass Monitoring Program, remote sensing research is underway to develop methods for documenting landscape changes in Texas seagrass beds related to human/natural disturbances. This study evaluated a technique that integrated aerial color photography, color transformation, threshold models, and geographic information system tools for mapping changes within a Texas seagrass bed for three site years. Maps developed with this technique had accuracies ranging from 75% to 100%. Temporal changes in the seagrass bed were easily detected in these maps. The method employed in this study has significant potential as an automated tool for monitoring fine scale landscape disturbance indicators in seagrass beds.

Technical Abstract: In support of the Texas Seagrass Monitoring Program, remote sensing research is underway to develop methods for documenting landscape changes in Texas seagrass beds related to human/natural disturbances. A technique is described and evaluated for detecting, assessing, and monitoring 1 m ground feature changes in disturbed bare areas within a Texas, USA, seagrass bed for three site years. Digitized aerial color photo transparencies of the study site were transformed from red (R), green (G), and blue (B) color space to intensity (I), hue (H), and saturation (S) color space. Visual analysis of the I/S imagery and their histograms were used to identify bare areas. Based on these evaluations, threshold models were developed for isolating bare areas from vegetated areas with seagrass and macroalgae. Maps developed with this technique had accuracies ranging from 75% to 100%. Changes in the bare areas between dates were determined with geographic information system tools available in the commercial software. The method employed in this study has significant potential as an automated tool for monitoring fine scale landscape disturbance indicators in seagrass beds.

Last Modified: 10/22/2014
Footer Content Back to Top of Page