Cokriging with airborne MSS imagery

The coregionalization of radiation sensed by an airborne multispectral scanner (MSS) and both the green leaf area index (GLAI) of barley and the biomass of pasture has been investigated and modeled. In both crops there was fairly strong spatial interdependence with the radiation, even though the simple correlation was weak. The auto and cross variograms showed dependence extending to some 15 m. These functions were used with the data to estimate GLAI and biomass by cokriging, and to calculate the estimation variances. The results showed that whereas simple regression would have provided worthless estimates, cokriging with the MSS data both increased the precision of the estimates above that attainable by auto kriging, and allowed the planning of a new survey with substantially less sampling effort. The study also revealed considerable redundancy in the MSS data.