耕地质量监测中不同抽样方法的精度比较——以河北省黄骅市为例
发布时间：2018-11-06 12:57
【摘要】：耕地质量监测是及时掌握耕地质量和产能变化的必要手段,科学合理的抽样方法不仅能保证监测精度,还可以减少监测成本。本文以河北省黄骅市为研究区域,结合区域耕地质量的实际状况,运用6种抽样方法抽取耕地单元作为监测样点,并计算和比较不同抽样均值的最大相对误差和相对抽样精度。结果表明,抽样精度随着样本容量的增加而提升,速度由快变慢。与简单随机抽样相比,在抽样精度方面,系统分层抽样和系统等距抽样的优势极大,并随着样本数量的增加优势扩大;等别分层抽样的优势较大并且平稳;乡镇分层抽样的优势不大。网格分层抽样不但没有优势,反而可能造成精度的损失。不同的抽样方法在满足相同的监测精度时所需的样本量差异显著,合理的抽样方法在满足监测精度的同时,可以明显地减少样本量。本研究有助于选择出最合适的抽样方法,以期为耕地质量监测中不同抽样方法的精度比较提供方法指导。
[Abstract]:Cultivated land quality monitoring is a necessary means to grasp the change of cultivated land quality and productivity in time. A scientific and reasonable sampling method can not only guarantee the monitoring accuracy but also reduce the monitoring cost. Taking Huanghua City of Hebei Province as the research area and combining with the actual situation of regional cultivated land quality, this paper uses 6 sampling methods to select cultivated land units as monitoring sample points, and calculates and compares the maximum relative error and relative sampling accuracy of different sampling mean values. The results show that the sampling accuracy increases with the increase of sample size, and the speed changes from fast to slow. Compared with simple random sampling, systematic stratified sampling and systematic equidistant sampling have great advantages in sampling accuracy, and expand with the increase of sample number, and the advantage of isotropic stratified sampling is larger and more stable than that of simple random sampling. The advantage of stratified sampling in villages and towns is not big. Grid stratified sampling not only has no advantage, but also may cause loss of precision. When different sampling methods meet the same monitoring accuracy, there are significant differences in sample size, and reasonable sampling methods can significantly reduce the sample size while satisfying the monitoring accuracy. This study is helpful to select the most suitable sampling method and to provide guidance for the accuracy comparison of different sampling methods in cultivated land quality monitoring.
【作者单位】： 河北农业大学国土资源学院;
【基金】：国土资源部公益性行业科研专项经费课题(201311060) 河北省科技支撑计划项目(132276329)
【分类号】：S159.2
本文编号：2314349
[Abstract]:Cultivated land quality monitoring is a necessary means to grasp the change of cultivated land quality and productivity in time. A scientific and reasonable sampling method can not only guarantee the monitoring accuracy but also reduce the monitoring cost. Taking Huanghua City of Hebei Province as the research area and combining with the actual situation of regional cultivated land quality, this paper uses 6 sampling methods to select cultivated land units as monitoring sample points, and calculates and compares the maximum relative error and relative sampling accuracy of different sampling mean values. The results show that the sampling accuracy increases with the increase of sample size, and the speed changes from fast to slow. Compared with simple random sampling, systematic stratified sampling and systematic equidistant sampling have great advantages in sampling accuracy, and expand with the increase of sample number, and the advantage of isotropic stratified sampling is larger and more stable than that of simple random sampling. The advantage of stratified sampling in villages and towns is not big. Grid stratified sampling not only has no advantage, but also may cause loss of precision. When different sampling methods meet the same monitoring accuracy, there are significant differences in sample size, and reasonable sampling methods can significantly reduce the sample size while satisfying the monitoring accuracy. This study is helpful to select the most suitable sampling method and to provide guidance for the accuracy comparison of different sampling methods in cultivated land quality monitoring.
【作者单位】： 河北农业大学国土资源学院;
【基金】：国土资源部公益性行业科研专项经费课题(201311060) 河北省科技支撑计划项目(132276329)
【分类号】：S159.2
【相似文献】
相关期刊论文 前2条
1 李琳一;袁涛;陈旭;;农业抽样调查中统计理论和抽样方法选取研究[J];上海农业学报;2011年04期
2 ;[J];;年期
相关会议论文 前1条
1 吴全;裴志远;;我国农情遥感监测中的抽样方法与问题[A];2008年中国农业资源与区划学会学术年会论文集[C];2008年
，本文编号：2314349
本文链接：https://www.wllwen.com/kejilunwen/nykj/2314349.html
