Weighting alternative estimates when using multi-source auxiliary data forforest inventory

Five auxiliary data sources (Landsat TM, IRS-IC, digitized aerial photograp hs, visual photo-interpretation and old forest compartment information) app lying three study areas and three estimators, two-phase sampling with strat ification, the k nearest neighbors and regression estimator, were examined. Auxiliary data were given for a high number of sample plots, which are her e called first phase sample plots. The plots were distributed using a syste matic grid over the study areas. Some of the plots were then measured in th e field for the necessary ground truth. Each auxiliary data source in combi nation with field sample information was applied to produce a specific esti mator for five forest stand characteristics: mean diameter, mean height, ag e, basal area, and volume of the growing stock. When five auxiliary data so urces were used, each stand characteristic and each first phase sample plot were supplied with five alternative estimates with three alternative estim ators. Mean square errors were then calculated for each alternative estimat or using the cross validation method. The final estimates were produced by weighting alternative estimates inversely according to the mean square erro rs related to the corresponding estimator. The result was better than the f inal estimate: of any of the single estimators. The improvement over the be st single estimate, as measured in mean square error, was 16.9 % on average for all five forest stand characteristics. The improvement was fairly equa l for all five forest stand characteristics. Only minor differences among t he accuracies of the three alternative estimators were recorded.