基于分层抽样的中国水稻种植面积遥感调查方法研究  

Paddy rice area estimation using a stratified sampling method with remote sensing in China

作　　者：焦险峰[1] 杨邦杰[1] 裴志远[1] 

机构地区：[1]农业部规划设计研究院农业资源监测站

出　　处：《农业工程学报》2006年第5期105-110,共6页Transactions of the Chinese Society of Agricultural Engineering

基　　金：国家自然科学基金资助项目"大尺度农作物种植面积空间统计方法与可靠性研究"(30471009)

摘　　要：及时准确的统计水稻种植面积对国家和区域的粮食生产、贸易及粮食安全预警有重要意义。传统的按行政单元逐级上报和农业产量抽样调查方法在数据获取过程中受人为因素的干扰,难以避免的出现诸如错报、漏报、空报等问题。遥感技术具有及时、准确、客观的特点,对于农作物种植面积监测具有其他方法不可替代的优势。但是,一般的作物面积遥感监测是全覆盖或典型地区调查。在大尺度农作物遥感调查时,全面普查(卫星遥感数据全覆盖)的方法在时间和经费方面是不可行的,以典型调查代替总体的方法缺乏科学依据。科学合理的抽样方法是可运行的大尺度作物面积监测的关键因素。研究在背景数据库的支持下,以土地利用数据库为辅助变量,设计了基于分层抽样的中国水稻种植面积遥感监测方法。以全国稻田面积为总体,采用1∶5万比例尺标准地形图幅为分层抽样的抽样单元。以遥感与地面调查相结合的方法监测样本的当年和上一年水稻种植面积,在给定精度条件下估算水稻种植面积年际变化率。结合上年统计部门发布的水稻种植面积统计数据,推算当年水稻种植面积。该项研究为农业部全国水稻遥感监测提供了可行的大尺度水稻遥感监测的运行方案。It is essential for rice production and trade management to collect objective and timely paddy rice planting at a regional and national level. Because of rapid economic development and crop rotation, the distribution of paddy rice planting area has changed dramatically and the area of paddy rice planting has decreased. In this study, an operational method was developed to monitor the inter-annual changes in paddy rice planting areas using stratified sampling method and multi-source remote sensing data in China. With the support of the background database, the population is defined as the total rice growing area, the sampling unit is designed with a 1 ： 50000 scale topographic map. The information for the sampling unit was collected by multi-source remote sensing survey and ground truth data collection. The total paddy rice planting area was estimated at a precision and confidence level. The method is being applied in main crop area estimation system in the Chinese Ministry of Agriculture.

关 键 词：水稻 种植面积估算 分层抽样 抽样单元 遥感监测 

分 类 号：S127[农业科学—农业基础科学] TP274.2[自动化与计算机技术—检测技术与自动化装置]