江汉平原耕地土壤氮素空间尺度套合与变异规律研究

 【目的】为克服尺度效应对氮素空间变异规律认识的影响，更加充分的利用多尺度空间信息。【方法】利用地统计学方法，以江汉平原后湖农场流塘分场为样区进行多尺度嵌套取样，探讨了全氮在不同采样尺度上的结构差异，提出了多尺度套合模型，研究了区域氮素的空间分布现状。【结果】各尺度上样点数据均符合正态分布，且以球状模型拟合效果最佳；全氮在不同尺度上都存在空间变异性，但其特征参数不同，尺度差异明显；采用多尺度套合模型进行kriging估值，可以有效提高拟合精度，且不局限于套合尺度的距离范围（本研究为100 m和50 m），部分区域内估计误差仅为普通克立格方法误差的1/2，精度提高的程度与套合尺度的设置有关。【结论】应用多尺度套合法研究土壤养分的空间变异，可有效提高拟合结果的可靠性和变量描述的准确性，可为精准农业方案的制定与实施提供科学依据。

Spatial Variability of Soil Nitrogen in Different Scales and Nested Simulation

Along with the development of research on soil spatial variation and precision agricultural practice, scale effect of soil nutrition performs obviously, which is promoted bottleneck in research and practicce on accurate agriculture. In this research, it is sampled by nested scenario in Houhu farm in Jianghan Plain; the variation of soil total nitrogen (TN) is simulated by using geostatistics in scale of 400, 100 and 50 meters separately; effect of scale on TN variation is analyzed; the simulating method under multi-scale data is put forward; and the spatial distribution of TN is described in this region. The research result shows that, data in three scale(400m, 100m, 50m) are all agree with normal distribution; spherical model fit best for semi-variogram; in different scale, TN varies under different driving factors, which leads to the appearance of scale effect; the estimation, through multi-scale simulation, is more accurate than traditional method, which isn’t limited by nested distance(100m, 50m in this research); furthermore, the standard error is only 1/3-1/2 of the original method in local area; the degree that the precision improved relates to the concert of scale distance. The simulation of soil nutrition by nested kriging method, is more reliable and concrete, which can offer scientific basis for formulation and implementation of the precision agricultural scheme in Jianghan Plain in China