Biomass Distribution Mapping Using Airborne Digital Video Imagery and Spatial Statistics in a Semi-Arid Environment

High resolution airborne digital video image data, biomass measurements and spatial statistics were used to map above-ground biomass for the five major semi-arid plant communities in the Jornada Long Term Ecological Research (LTER) site (southern New Mexico). The two principal objectives were to determine: (1) spatial characteristics of arid shrub versus semi-arid grassland vegetation; and (2) a suitable image spatial resolution and ground sampling interval to map above-ground biomass spatial distribution for these vegetation types. The spatial characteristics of each plant community were established by analyzing digital images at varying pixel sizes using semi-variograms. As pixel size increased from 0·5 m to 16 m, little information on vegetation pattern and abundance was lost in grassland and playa grassland sites. In comparison, the pattern and abundance of vegetation became indistinct in shrubland sites once pixel size exceeded mean shrub diameter. This work illustrates the utility of variograms from remotely sensed data for two applications: (1) determining a suitable scale to examine an ecosystemʹs spatial structure; and (2) providing information on the spatial pattern of vegetation as an indicator of ecosystem condition in the context of a model for desertification.