在获取农作物种植面积方面遥感空间技术有着巨大的优势，设计空间分层抽样中高效的分层指标是非常关键的。在传统的分层指标设计中，由于忽视了遥感识别结果中的分类误差，一定程度上影响了抽样效率。因此本文选择了北京市通州、大兴区为研究区域，以一定的抽样方案为基础，提出了基于遥感分类误差面积的分层标志——误差校正面积，从全局空间自相关性、辅助变量与目标变量的相关性两个方面进行相关性分析，并多次抽样进行研究区冬小麦总量面积估算和精度评价，将误差校正面积与传统分层指标（面积规模）进行比较和评价。实验表明：一定抽样单元尺寸下，误差校正面积指标在全局空间自相关性、与目标真值相关性、抽样误差、推断稳定性等方面均优于面积规模，其中推断精度能稳定提高近1%；CV值能稳定减少近0.8%；空间自相关指数能稳定减少近0.5%；相关系数大于0.7且能提高至87.15%，说明该指标可在一定程度上提高抽样精度、外推稳定性，可提高空间抽样的抽样效率，验证了其作为遥感空间抽样面积估算分层指标的有效性及相比面积规模指标的优势。

Remote sensing has strong superiority in crop planting area acquisition. It is critical to design an efficient hierarchical indicator in a stratified sampling investigation. Traditional method neglects the classifying error of the classification results remote sensing, which reduces the sampling efficiency. This paper searched in Tongzhou and Daxing, and based on a certain sampling inspection plan, point out a new hierarchical indicator that can correct the error. This paper will analyze the Moran's I and the correlation between the target and auxiliary variables, and to estimate the total winter wheat area and accuracy evaluation by Multiple sampling. Then comparing this indicator with the area scale to testify the validity of this indicator. Experiments showed that: With the sizes of sampling unit, this indicator is better than the area scale in the correlation, Moran's I, stability and accuracy. Nearly 1% can be enhanced stability in accuracy. This indicator can reduce CV value nearly 0.8%. And Moran's I can be decreased 0.5% approximately. The correlation is greater than 0.7 and can be improved to 87.15%. Shows that the indicator can improve the accuracy of sampling, and the stability of inference in a certain extent. The effectiveness and advantages of the indicator has been proved by the experimental results.