摘要
针对传统抽样调查工作面临着野外调查工作量大、资料时效性较低且难以满足人们对数据现势性的高要求等一系列缺点,以新疆棉花种植主棉区沙湾县、玛纳斯县、呼图壁县为研究区,结合遥感技术提出了一种基于PPS分层抽样的空间抽样设计方案,并将该方案用于研究区棉花种植面积的估算。结果显示,PPS抽样与分层抽样结合后极大地提高了PPS抽样反推总体的方法优势。分配样本时分别采用按每层辅助变量之和的期望的算术平方根与该层待抽样单位总数之积、每层辅助变量之和进行比例分配的2种分配方法,其对应的反推总体的估计量变异系数分别为0.008、0.009,相对误差分别为0.016、0.017,分层后的样本变异程度极低,为反推结果的高精度打下了基础。2种样本分配方式的棉花种植面积提取精度均高于94%。该方法不仅精度高,而且在实际操作中简单方便。
The method of traditional sampling survey work has two major drawbacks. One is the large workload ,and the other is the outdated basic data. It is difficult to meet the people5 s high requirements of data updating. In order to solve those problems in a conditional sampling survey, a novel crop area extraction method was proposed in this paper using remote sensing and PPS stratified sampling technology. The algorithm had been tested over a study area in counties of Mnasi,Hutubi and Shawan which are all cotton planting base counties. The results showed that the combination of PPS sampling and stratified sampling greatly improved the overall method advantage of PPS sampling. Two proportional distribution methods for sample distribution were adapted, which were the product of the sum of the desired arithmentic square root of each layer of auxiliary variables and the total number of units in each layer, the sum of auxiliary variables per layer. The variation coefficients of population of extrapolated population total were 0. 008 and 0.009 by two proportional distribution methods,and the relative errors with two distribution methods were 0.016 and 0. 017. The degree of variation of samples was very low after stratification, which layed the foundation for the high precision of the result. The total area accuracy of cotton planting reached more than 94% by the two sample distribution method. This method not only has high precision, but also is simpie and convenient in practical operation.
出处
《河南农业科学》
CSCD
北大核心
2016年第11期155-159,共5页
Journal of Henan Agricultural Sciences
基金
国家统计局新疆棉花种植面积遥感调查项目
国家自然基金项目(41371358)
国家"863"计划项目(2014AA06A511)
国家科技重大专项(14CNIC-032079-32-02)
关键词
遥感
空间抽样
样本分配
外推总体
remote sensing
spatial sampling
distribution of samples
extrapolated population
