面向高分卫星遥感共性产品真实性检验的无人机空港布局

Unmanned aerial vehicle airports for verification of common products from China High-resolution Earth Observation System(CHEOS)

针对高分卫星遥感共性产品真实性检验对无人机数据的需求,本文以已有野外台站为依托,开展无人机空港布局研究。首先,综合考虑影响卫星产品空间变异的因素,构建了地理背景覆盖率的物理模型;其次,基于地理背景覆盖率改进覆盖模型提出最大地理背景覆盖模型,并利用简单随机抽样中的样本容量确定方法确定合理的空港数量,分别利用本文提出的最大地理背景覆盖模型和已有最大面积覆盖模型进行空港布局的研究;最后,对两种模型下的空港布局结果进行对比,并利用MODIS EVI数据产品对布局结果进行验证。结果表明,在空港数有限的条件下(n=60),最大面积覆盖模型选择的空港,其服务范围内的地理背景覆盖率为26.66%,能代表70.37%的中国陆地区域,而最大地理背景覆盖模型选择的空港,其服务范围内的地理背景覆盖率为38.32%,能代表73.36%的中国陆地区域。最大地理背景覆盖模型比最大面积覆盖模型能获得更优的空港布局结果,可用于支撑中国区域面向高分卫星遥感共性产品真实性检验的无人机观测网络建设。

Aiming at the requirement to use unmanned aerial vehicle(UAV) data to verify common products from China High-resolution Earth Observation System(CHEOS), this research examines the distribution of UAV airports using the existing ecological observation stations. Firstly, to comprehensively consider the factors affecting the satellite products, a physical model of the geographical background coverage was proposed. Secondly, a maximal geographic background model is designed based on the improved the coverage model with geographic backgrounds, and the number of UAV airports was determined using the sample size determination method of random sampling. Then, the maximal geographic background covering model and a maximal area covering model were used to study the distribution of UAV airports. Finally, the results of the airport layout under the two models were compared, and MODIS EVI(Enhanced Vegetation Index) data products were used to verify the results. The results showed that for a limited number of airports(n = 60), the geographical background coverage of the service areas of the airports selected by the maximal area coverage model was 26.66%, which represents 70.37% of China’s land area. The geographical background coverage of the service areas of the airports selected by the maximal geographical background covering model was 38.32%, which represents 73.36% of China’s land area. The layout of the UAV airports selected by the maximal geographic covering model is better than that selected by the maximal area covering model. The results can lay a foundation for the design of an unmanned aerial vehicle observation network for the verification of common products from CHEOS.