土壤硝态氮含量检测和分布图生成中的最优样本量

土壤参数的时空变异是实施精细农业时要考虑的重要因素，在土壤检测的栅格采样中有必要确定最优样本量。本研究的试验田是一块生长中的玉米地块，试验区的面积为4.2 m×4.2 m，该试验区被假定为土壤采样中的一个栅格，该栅格又被细分为49个0.6 m×0.6 m的子栅格。采样时，所分析的土壤参数为土壤硝态氮含量，从播种到收获共进行了7次采样。通过对土壤样本土壤硝态氮时空变异的分析，揭示了样本量和土壤硝态氮含量预测误差之间的相关关系。土壤硝态氮含量呈非正态分布，通过对玉米各个生长期获得的数据分析表明：含量水平的预测误差随深度的增加而增大；当从一个栅格只采集1个土壤样本时(样本量为1)，预测误差基本在50％左右(显著水平：α≈0.10)，而当从一个栅格采集5个土壤样本时(样本量为5)，预测误差将降至25％左右。另一方面，当要求预测误差低于30％时，对于普通生长条件下的土壤需要从1个栅格至少采集3个样本，而对于追肥后的土壤则至少需采集15个样本。

Optimum sample size of nitrate nitrogen in soil sensing and mapping

Temporal and spatial variability of soil parameters is very important for precision agriculture. It is necessary to determine how many samples should be collected for representing the feature over one grid space in soil sensing. Test area located in a growing cornfield was 4.2 m×4.2 m with 0.6 m×0.6 m sub-grid spacing. The soil parameter analyzed was nitrate nitrogen. Seven tests were conducted. The first sampling was carried out a month later after seeding, and the final one was made ten days later after harvesting. The variation of nitrate nitrogen was analyzed in order to clarify the relationship between sample size and estimation error of the nitrate nitrogen level. Data obtained provide non-normal distribution. Analyses of the data from various stages of growth show that estimation errors increased with depth. When only one sample is collected from a grid to estimate the level of nitrate nitrogen, the estimation errors were almost 50% (given α≈0.10). When collecting five samples from one grid (sample size was 5), estimation error would decrease to 25%. Keeping estimation error below 30% with significant level 0.1 required about 3 of sample size in common growing stages, but at least 15 of sample size after top dressing.