PRELIMINARY SPECTRAL ASSESSMENT OF LAGUNA MADRE WATER FEATURES USING AISA+ HYPERSPECTRAL DATA

摘要

Two hundred forty four-band (.39-.97 mum), 1.5 meter resolution, AISA+ hyperspectral data were acquired by Indiana State University and the USDA-ARS, Weslaco, TX in early March 2005. These data were acquired from areas in South Padre Island (SPI), Lower Laguna Madre (LLM) and terrestrial fresh water in the Lower Rio Grande Valley. Preliminary analysis of hyperspectral scenes acquired from the hypersaline LLM was conducted focusing on seagrass and brown tide algal blooms. Insights into basic patterns of seagrass density, water depth, and lagoon bottom feature conditions were provided in a LLM study area near SPI through analysis of unsupervised and supervised classification results that used hyperspectral data subsets in original and principal component forms. Texas brown tide algal blooms and spectral variation patterns within them were identified and analyzed in a LLM study area. Hyperspectral data analysis indicated that important LLM seagrass and associated features are possible to identify. This research supports future use of AISA+ data to acquire many types of water, seagrass, and associated parameter information. Development of data most useful for monitoring and modeling requires the integration of spectral data acquired nearly concurrently with detailed data from field sampling.