Knowledge-Based System to Assist in the Design of Soil Survey Schemes

摘要

Soil survey information with quantified accuracy requires a statistical samplingstrategy as part of the survey scheme. In designing such schemes several decisions are made, each affecting the survey costs and/or the survey results, for example how many samples to take where, and what to measure how. This study investigates the possibilities of a knowledge-based system to assist in the design of survey schemes, incorporating pedological and statistical knowledge, with models predicting costs and accuracy from the available prior information, and algorithms to find the sampling strategy that gives maximum accuracy for a given budget.