Abstract:
In order to improve the current crop acreage sampling survey systems, the experiment on spatial sampling methods for estimating winter wheat sown acreage was conducted combining the “3S” techniques (Remote Sensing, Geographic Information Systems and Global Positioning techniques) with the traditional sampling methods. Taken Shandong province as study area and winter wheat sown area as the study object, 8 kinds of sample size from 74 to 333 and 4 sets of spatial sampling methods (simple random sampling method, stratified sampling with the winter wheat planting regionalization as stratification symbol, stratified sampling with cultivated land types as stratification symbol, stratified sampling with winter wheat sown area as stratification symbol in each county in study area) were designed in this experiment. The results demonstrated that: the efficiency of stratified sampling method which the winter wheat sown area in each county was selected as the stratum symbol is the highest among 4 sampling methods under the condition that the relative errors of 4 sampling methods were nearly equivalent. Inferring population values and estimating sampling error under 8 kinds of sample size level based on sample observations, and the relative errors between population total truth-values and estimates decreased with the sample size increased, however, variation coefficient (CV) of population total estimator was still higher.