Evaluation of LANDSAT multispectral scanner data for mapping vegetated landscapes

Abstract

LANDSAT multispectral scanner data were selected to represent periods of adequate soil water, soil water deficit, and beginning of soil water recharged in Brazos County, Texas to evaluate LANDSAT multispectral scanner data for mapping vegetated soil landscapes. Six test sites were evaluated using LARSYS computer programs developed by the Laboratory for Application of Remote Sensing, Purdue University. Open grassland soils were best separated at a period when soil moisture was being replenished after a deficit period. Woodland soils were separated by LANDSAT data only when adequate moisture was available in late spring. Separation was not possible during a period of soil water deficit. Properties of soils that are important to plant growth were separable n the test sites and extended 6 km by 6 km areas using June and October LANDSAT data. The particle-size (family) and soil moisture regime (suborder) were readily separated. Soils with argillic horizons were also well separated. Further research is necessary, however, to show conclusively the ability to separate soils with argillic horizons because the majority of the soil taxa under vegetation had argillic horizons.