摘要
将吉林省榆树市弓棚镇作为研究区域,利用采样密度和插值方法(BP神经网络方法和Kriging方法),研究其对农田土壤碱解氮空间变异性的影响。研究结果表明:Kriging模型插值精度随采样密度的减少,呈显著下降趋势,采样密度对BP神经网络插值精度影响相对较小。当采样密度较大时,插值精度表现为Kriging模型显著高于BP神经网络模型;当采样密度较小时,BP神经网络模型的插值精度优于Kriging模型插值精度。研究为精准农业土壤养分插值方法的选取、制定优化采样策略提供科学依据。
This article takes the Village of Yushu Gongpeng Town in Jilin Province as the study area to research the effects of sampling density and spatial interpolation(the Kriging method and BP neural network method) on alkali-hydrolyzable nitrogen Spatial Variability in Farmland Soils. The results of the study show that the interpolation precision of the Kriging model drops obviously along with the reduction of the sampling density, but the sampling density has less effect on the BP neural network model's interpolation precision. In the larger sampling densities, the interpolation precision of the Kriging model is apparently higher than the BP neural network model; in the less sampling densities, the interpolation precision of the BP neural network model is better than that of the Kriging model.This research provides the scientific basis for the selection of the accurate anthropogenic soil nutrient interpolation method and the formulation of optimized sampling strategy.
出处
《土壤通报》
CAS
CSCD
北大核心
2014年第4期789-794,共6页
Chinese Journal of Soil Science
基金
吉林省科技厅科技支撑计划项目(20080207)
国家农业科技成果转化资金项目(2009GB2B100095)
吉林师范大学研究生创新科研计划项目(2013036)资助
关键词
采样密度
插值方法
空间变异性
土壤碱解氮
Sampling density
Interpolation method
Spatial variability
Soil alkali-hydrolyzable nitrogen