Spatial and spectral resolution necessary for remotely sensed vegetation studies

An outline is presented of the required spatial and spectral resolution needed for accurate vegetation discrimination and mapping studies as well as for determination of state of health (i.e., detection of stress symptoms) of actively growing vegetation. Good success was achieved in vegetation discrimination and mapping of a heterogeneous forest cover in the ridge and valley portion of the Appalachians using multispectral data acquired with a spatial resolution of 15 m (IFOV). A sensor system delivering 10 to 15 m spatial resolution is needed for both vegetation mapping and detection of stress symptoms. Based on the vegetation discrimination and mapping exercises conducted at the Lost River site, accurate products (vegetation maps) are produced using broad-band spectral data ranging from the .500 to 2.500 micron portion of the spectrum. In order of decreasing utility for vegetation discrimination, the four most valuable TM simulator VNIR bands are: 6 (1.55 to 1.75 microns), 3 (0.63 to 0.69 microns), 5 (1.00 to 1.30 microns) and 4 (0.76 to 0.90 microns).