抽样单元空间相关性和变异性对农作物面积空间抽样效率的影响机理研究

Aiming at the issue that the sampling efficiency is still not high, due to the lack of considering spatial correlation and variability of sampling units in the spatial sampling scheme for crop acreage estimation. In this study, 3 national sampling survey counties(DeHui Couty in Jilin Province, PuYang County in Henan Province and LiYang County in Jiangsu Province)are selected as the study area, and main food crops(winter wheat, rice and maize)in China are chosen as the study objects to investigate and understand michanisims of impacts of spatial correlation and variability on spatial sampling efficiencies.To do that, Geostatistics and landscape ecology theory are introduced, spacial analysis,"3S" technology (Remote Sensing, Geography Information Systems, Global Positioning Systems)and the traditional sampling methods are used to study the characteristics of spatial corrlation and variability of sampling units. The influnces of sampling unit scale, crop spatial distribution and farmland landscape fragmentation on spatial correlation and variability are analyzed; The efficiencies of traditional sampling methods and spatial sampling methods that considering spatial correlation and variability of sampling units are evaluated; The impact of spatial correlation and variability on spatial sampling efficiencies is discussed; Finally,the method on optimizing spatial sampling scheme for crop acreage estimation is proposed. This study will play an important role to verify the influence rules of spatial correlation and variability on spatial sampling efficiencies,improve the sampling survey productivity and evaluate the application suitability of spatial sampling methods in the different regions.

针对目前国内外农作物面积空间抽样方法研究中因对抽样单元空间相关性和变异性考虑不足，从而限制了空间抽样效率的提高这一问题，本项目以我国主要粮食作物（小麦、水稻、玉米）为研究对象，选取3个国家级抽样调查县（吉林省德惠市、河南省濮阳市和江苏省溧阳市）为研究区，拟引入地统计学及景观生态学理论，通过空间分析技术、"3S"技术与传统抽样方法相结合，研究农作物面积抽样单元空间相关性和变异性特征，分析抽样单元尺度、农作物空间分布、耕地景观破碎度等因素对空间相关性和变异性的影响；评价传统抽样与考虑空间相关性和变异性的空间抽样方法在估计农作物面积时的效率；研究空间相关性和变异性对农作物面积空间抽样效率的影响，提出农作物面积空间抽样优化设计方法。预期成果对探明空间相关性和变异性对农作物面积空间抽样效率的影响机理，提高农作物面积抽样调查效率，评价空间抽样方法在不同地区应用的适宜性具有重要学术意义和实用价值。

针对以往国内外农作物面积空间抽样研究中因未考虑抽样单元空间相关性和异质性，导致空间抽样方案的制定合理性不足、效率不高的问题，利用全局空间自相关指数Moran’s I 和变异函数定量评价了抽样单元内农作物面积空间相关性和异质性，分析了抽样单元尺度、农作物类型、农作物种植强度、耕地地块破碎度、地表坡度等对抽样单元内农作物面积占比空间相关性和变异性的影响，阐明了抽样单元内农作物面积占比的空间相关性和异质性特征及其影响因素。选取抽样外推相对误差RE、变异系数CV及样本容量n为评价指标，比较了传统抽样与考虑空间相关性和变异性的空间抽样的效率高低，分析了空间相关性和异质性对农作物面积抽样调查效率的影响机理，提出了空间相关性和变异性存在下的农作物面积空间抽样方法，实现了农作物面积空间抽样方案的科学合理设计，改善了区域农作物面积抽样调查效率，为空间抽样方法在不同农业类型区的应用选择提供了参考依据。通过四年的试验研究，完成了项目的预期任务。已发表论文11篇，其中4篇SCI收录、6篇EI收录，主编专著2部。培养硕士研究生2名。