| 基于总体抽样下子总体追加抽样设计的棉花种植面积遥感估算方法研究 |
| |
| 引用本文： | 胡永森,王力,施开分,周巍,王长耀.基于总体抽样下子总体追加抽样设计的棉花种植面积遥感估算方法研究[J].河南农业科学,2017,46(1). |
| |
| 作者姓名： | 胡永森 王力 施开分 周巍 王长耀 |
| |
| 作者单位： | 1. 东华理工大学测绘工程学院,江西南昌330013;中国科学院遥感与数字地球研究所/遥感科学国家重点实验室,北京100101;流域生态与地理环境监测国家测绘地理信息局重点实验室,江西南昌330013;2. 中国科学院遥感与数字地球研究所/遥感科学国家重点实验室,北京,100101;3. 国家统计局农村社会经济调查司,北京,100826 |
| |
| 基金项目： | 国家统计局新疆棉花种植面积遥感调查项目，国家自然科学基金项目，国家863计划项目，国家科技重大专项 |
| |
| 摘 要： | 针对遥感与抽样相结合的农作物种植面积估算方法中,子总体的作物种植面积估算需要新的抽样体系设计和样本野外调查等繁重工作,缺少基于总体下对子总体的作物种植面积估算的方法,以沙湾县、玛纳斯县和呼图壁县为总研究区,提出了基于遥感和PPS抽样相结合的总体抽样下子总体追加抽样设计的子总体研究区呼图壁县棉花种植面积估算方法,并以新疆建设兵团统计局公布数据为真值对其估算结果进行对比分析。结果显示,基于总体抽样下子总体追加抽样设计后,子总体研究区估计量的变异系数为0.023 3,远低于0.05,而追加抽样设计前变异系数为0.122 3,说明样本在该方法下的代表性得到极大提高。以新疆建设兵团统计局公布数据为真值进行对比发现,子总体研究区棉花种植面积提取精度达到94.2%,能够有效提取子总体中的棉花种植面积,同时避免了重新建立子总体研究区抽样体系所需要的人力、物力、财力等资源的消耗。
|
| 关 键 词： | 空间抽样 棉花种植面积 子总体 遥感 |
Study on Remote Sensing Estimation of Cotton Planting Area Based on Additional Subpopulation Sampling under Overall Spatial Sampling Design |
| |
| Abstract: | In the method of crop acreage estimation based on the combination of remote sensing and sampling,the estimation of the parameters of the subpopulation requires a new sampling system design and the large workload of a field survey,the crop planting area remote sensing measurement method based on additional subpopulation sampling under overall spatial sampling design was lack.This paper took the Shawan county,Manasi county and Hutubi county as the research areas,and put forward a cotton planting area remote sensing measurement method based on additional subpopulation sampling under overall spatial sampling design with remote sensing and probability proportionate to size sampling.Finally,the data of cotton growing area released by Bureau of Statistics of Xinjiang Construction Corps were compared with the results of the sampling estimation.The results showed that the coefficient of variation of estimator of subpopulation in the study area was 0.023 3 after the additional sampling design,far below 0.05;but before the additional sampling design,which was 0.122 3,illustrating that the representative of the sample under the method had been enhanced greatly.Compared with the data released by Bureau of Statistics of Xinjiang Construction Corps,cotton planting area extraction accuracy of subpopulation in the study area reached 94.2％.This method could extract the cotton growing area of subpopulation effectively;at the same time,it could avoid the waste of human,material and financial resources and other resources needed to reestablish the sub population in the study area sampling system. |
| |
| Keywords: | spatial sampling cotton planting area subpopulation remote sensing |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
