Remote sensing surveys design in regional agricultural inventories

Abstract In this paper, we consider the methodology problems of remote sensing surveys design in regional agricultural inventories. The strategy of samples, based on the combined use of multispectral aerospace data and ground truth data obtained on test sites in the region under supervision, is used. The strategy of samples includes: selection of areas, which are statistically homogenous with certain agricultural parameters under research, identification of representative test sites grid; remote sensing from aerospace platforms and ground truth data acquisition on test sites as well. The ground measurements of biometrical parameters of certain agricultural crops under research are taken at test sites, maps of anomalies are compiled, spectrometrical and other optico-physical characteristics of vegetation canopies and soils are defined. The derived data are used in automatic interactive imagery processing at the training stages of the procedures of classification and application of thematical remote sensing data processing results to the entire region.