Abstract: 【Objective】 Sample layout is one of the key factors in spatial sampling schemes for estimating crop planting acreage. Optimization of sample layout plays an important role in improving the representativeness of samples versus population and the accuracy of population extrapolation, decreasing the cost of survey sampling. In this study, focusing on the problem that the formulation of samples layout in the spatial sampling scheme for crop acreage estimation was not reasonable (e.g. samples units are  not all independent of each other, when simple random sampling method was used to formulate samples layout; sampling intervals are not able to be reasonably defined, when systematic sampling method was used in the design of samples layout), the author tried to propose a optimal formulation of samples layout to improve the efficiency of the spatial sampling scheme further.【Method】Mengcheng County in Anhui Province, China was chosen as the study area, winter wheat planting acreage as the study object, and square girds as the shape of sampling units. Geostatistics, “3S” techniques (Remote Sensing, Geographic Information Systems and Global Positioning techniques) and traditional sampling methods were used in this study. Firstly, 8 kinds of sampling unit sizes were formulated, and then the study area was split by the sampling units with these 8 kinds of sizes to construct the sampling frame. The winter wheat acreages in all sampling units were calculated based on the spatial distribution data of winter wheat in 2009 and 2010(derived by ALOS AVNIR-2 and Landsat5 TM image, respectively). Secondly, in order to build the Variogram theoretical model of winter wheat planting acreage proportion within the sampling unit (WPS), simple random sampling method was used to draw the initial samples. Spatial correlation and variability of sampling units were analyzed, and spatial correlation threshold was quantitatively determined by the Variogram theoretical model. Thirdly, the equi-spaced pattern (sampling intervals were identical in vertical and horizontal directions, and spatial correlation threshold of samples was chosen as the sampling interval) was used to reasonably formulate the samples layout, following the principle that samples units were independent of each other in the traditional sampling methods. Finally, the population extrapolation accuracy, stability and sampling cost were estimated according to the samples that the spatial layout were reasonably formulated. In order to evaluate the design efficiency of samples layout, relative error, coefficient of variation (CV) and sampling size were selected as the indices, and simple random sampling method as the control treatment. 【Result】The experimental results demonstrate that, the variability of WPS increased with sampling unit size increasing. CV of WPS varies from 32.75% to 43.46% under 8 kinds of sampling unit size levels; There was an intense spatial correlation among all of WPS within a certain range, and the spatial correlation was mainly dominated by structural factors (climate, topography, soil type). Spatial correlation thresholds of WPS increase with sampling unit size increasing; The relative error and CV of population extrapolation that samples layout was optimized were obviously less than those of simple random sampling method at the same sample size, while sampling unit size was small (500m×500m-2 000m×2 000m); Compared those of simple random sampling method, although the relative error and CV were not decreased after optimized design of sample layout, there was an obvious decreasing in sample size, when sampling unit size was larger (2 500m×2 500m-4 000m×4 000m).【Conclusion】In this way, this research can provide a solution for improving the efficiency of spatial sampling scheme to estimate crop planting acreage.

Key words: spatial sampling, winter wheat, planting acreage, sample layout, geostatistics, spatial correlation