In situ narrow‐band reflectance characteristics of cover components in sagebrush‐steppe

Abstract

To capitalize on the planned deployment of satellite‐bourne narrow‐band multispectral radiometers (e.g., Hyperspectral Imager) in rangeland monitoring, a study was undertaken to evaluate the in situ, spectral uniqueness of 16 important sagebrush steppe cover components in southeastern Idaho during peak plant growth in June of 1995. Five major plant species from the study area were also sampled multi‐temporally in 1996, during peak green (June), mid‐sensecence (late July), and late‐senescence (late August). In situ, simulated TM data were also used to evaluate the broad‐band NDVI of these same components. Balsamroot, flixweed, cheatgrass, bluebunch wheatgrass, horsebrush, and three‐tip sagebrush differed in many portions of their June 1995 narrow‐band spectral response curves. NDVI values decreased through the forb, grass, and shrub components, respectively. Rates of change in reflectance (and NDVI) over time among the five major plants tended to mirror their phenologies, with forbs changing the most, followed by grasses, and finally, shrubs. Maturing cheatgrass (an annual) was also distinct from bluebunch wheatgrass (a perennial) because of the former's rapid senescence that resulted in greater red reflectance by late July. Soil components were less spectrally distinct from one another, as well as from litter and standing dead shrub. Wet and dry soils were separable, in part because the latter had a more variable narrow‐band response curve. The size of the IFOV examined did not affect sagebrush spectral response, but inversely affected the response of balsamroot. Narrow‐band imagery, due to the increased differences among spectral response curves, appears to have significant potential for distinguishing among components in sagebrush steppe rangelands. These differences may be further accentuated if multi‐temporal data are collected to capitalize on phenological variations among cover components.