SPATIAL PREDICTION OF VEGETATION QUANTITIES USING GROUND AND IMAGE DATA

A major challenge in Earth system studies is mapping vegetation quanti ties over large regions. Aspatial regression is typically the empirica l method applied to remotely sensed and ground data for the spatial pr ediction of vegetation variables. Geostatistical methods, such as cokr iging and stochastic simulation, have rarely been used for this purpos e. A synthetic example was constructed from imaging spectrometer data to allow an objective comparison between regression, cokriging and a n ew stochastic simulation method. A range of linear relations between g round sample data and image data was represented in the example. The l owest root-mean-square-error was achieved with cokriging until the cor relation coefficient (r) between direct and ancillary data exceeded 0. 89, at which point regression was the more accurate predictor. Probabi lity-field simulation gave a range of possible realizations, overall l ess accurate than those from regression but more faithful to the histo gram and spatial pattern of the variable to be predicted. The strength of the relation between ground measurements and image data was shown to be a critical factor in choosing a spatial prediction method.