Using spectral vegetation indices to estimate rangeland productivity

Abstract

Spectral vegetation indices (VI), that use combinations of photographic red (RED) and near‐infrared (NIR) remotely sensed radiances, were used to develop predictive relations for pholosynthetically active (green) biomass of rangeland plant canopies. Two basic types of soil adjusted VIs thai use multiplicative ratios and linear additive transformations of RED and NIR radiances were evaluated. Results showed that the perpendicular vegetation index (PVI), soil adjusted VI (SAVI), transformed soil adjusted VI (TSAVI), and soil adjusted ratio VI (SAVI₂₎, were all equally related to green standing biomass. Soil moisture measurements significantly improved predictive relations for green biomass with VI. Thus, it is postulated that a thermal sensor, that could provide an independent estimate of soil moisture, might be useful for improving estimates of range productivity during periods where rainfall accumulations are temporally and geographically sporadic. These results indicated that most of the Vis tested should be useful for predictive relations that estimate range productivity, but that soil moisture will have a significant effect on the final utility of these relations.