The Effects of Tidal Inundation on the Reflectance Characteristics of Coastal Marsh Vegetation

Hyperspectral and sophisticated multispectral imagery are increasingly used to obtain detailed information on species composition, condition, biomass, and other characteristics of coastal marshes. However, how differing levels of tidal inundation affect the reflectance characteristics of emergent marsh vegetation remains not well documented. In 19941996, general reflectance spectra were collected for three common brackish marsh species with different canopy architectures (Schoenoplectus americanus, Spartina patens, and Spartina cynosuroides) in field experiments in which changing levels of tidal inundation were simulated in specially nonreflected enclosures. Reprocessing the data with software not then available has allowed greater delineation of variations in the data, especially in the red-near-infrared boundary, that previously could not be differentiated. Spectra for all three species showed significant reductions in near-infrared reflectance (900-1100 nm) with progressive substrate inundation, the sharpest decline occurring for S. americanus, for which the narrow, tapering culms allowed substantial submergence of total aboveground biomass (and greatly decreased leaf area index [LAI]) after water depths on substrates reached 15 cm. Increasing inundation produced a particularly significant change in the 700-825-nm band of the spectral curve. Maximum peak percent reflectance shifted as much as 40 nm in these wavelengths, with two maxima over 100 nm apart in wavelength appearing in both the S. americanus and S. patens curves at water depths >30 cm. Simulations of Landsat Thematic Mapper bands show that the Normalized Difference Vegetation Index (NDVI) is highly correlated with LAI and suggest that NDVI-based estimates for marsh biomass can be strongly influenced by the effects of marsh canopy submergence on LAI.