Increasing pressure is being applied to resource management agencies to inventory the resources under their jurisdiction in order to formulate appropriate management plans. Present inventory procedures are inadequate when considering budget levels and the tremendous areas to be inventoried. It is necessary, then, to develop inventory procedures which provide for detailed ground information over large areas with adequate update capabilities. Inventory methods are now being developed and refined which integrate information from multiple levels of remotely sensed data and ground data in appropriate statistical frameworks.

This paper describes the methods and statistical procedures, an analysis of variance routine (ANOVA), for relating ground information to spectral classes resulting from the automatic classification of LANDSAT MSS data. The procedures were a portion of a research project completed for the FIRESCOPE program of the US Forest Service.

Date of this Version