中国农业科学院农业信息研究所 中国农学会
Agricultural Information Institute of CAAS China Association of Agricultural Science Societies
En
Ch
中国农业科学院农业信息研究所 中国农学会 主办

AI检索

订阅

返回顶部
当前位置: 首页 > 文章 > 基于多元线性回归模型的土壤养分空间预测——以陕西省蓝田县农耕区为例 土壤通报 2017 (5) 1102-1113
Position: Home > Articles > Prediction of Spatial Distribution of Soil Nutrients Based on Multiple Linear Regression Model——a Case Study in Lantian County of Shaanxi Province Chinese Journal of Soil Science 2017 (5) 1102-1113

基于多元线性回归模型的土壤养分空间预测——以陕西省蓝田县农耕区为例

作  者：
杨煜岑;杨联安;王晶;王安乐;黄安;张彬;向莹;王智宇
单  位：
西北大学城市与环境学院
关键词：
空间预测;多元线性回归;影响因子;土壤养分
摘  要：
以陕西省蓝田县农耕区为例,基于土壤类型、地形、植被、人类活动等土壤养分影响因子,采用多元线性回归预测方法,对土壤养分分布进行空间预测。通过数据标准化和分类统计均值法将所有影响因子统一为相对度量值,根据影响因子与土壤养分含量的相关性显著程度,筛选合适的因子构建回归模型,最后使用随机抽取的验证样本集,计算了预测结果误差。结果表明:多元线性回归预测结果虽然与Kriging法预测结果宏观上具有基本一致的空间分布趋势,但是它克服了传统插值方法的不合理突变问题,更加精细地描述了土壤养分空间分布特征;其平均误差和均方根误差小于Kriging插值法,可作为集成多元影响因子对土壤养分空间分布预测的有效方法。分析发现,研究区内洪积土、水稻土、塿土、立茬土等土壤类型分布区,以及地势较低、坡度在0~8°之间、坡向为东坡、东南坡,等水热组合条件良好的区域养分含量较高。该研究可以为土壤资源合理利用、土地的科学管理以及测土施肥技术提供重要依据。
译  名：
Prediction of Spatial Distribution of Soil Nutrients Based on Multiple Linear Regression Model——a Case Study in Lantian County of Shaanxi Province
作  者：
YANG Yu-cen;YANG Lian-an;WANG Jing;WANG An-le;HUANG An;ZHANG Bin;XIANG Ying;WANG Zhi-yu;College of Urban and Environmental Science, Northwest University;Lantian County Agricultural Technology Popularization Center;School of Geography and Resources Science of Neijing Normal University;
单  位：
College of Urban and Environmental Science, Northwest University%Lantian County Agricultural Technology Popularization Center%School of Geography and Resources Science of Neijing Normal University
关键词：
Spatial prediction;;Multiple linear regression;;Soil factor;;Soil nutrient
摘  要：
Taking Lantian agricultural cropping area as an example, based on the factors affecting the formation of soil nutrition including soil type, terrain, plants, and as well as human activities, and adopting multiple linear regression prediction, we obtained the spatial distribution of soil nutrients. All the influencing factors were unified as the relative measurement by the method of data standard and classified statistics, and appropriating factors were selected by the level of significant correlation degree to construct the model. Finally, we calculated the Mean-Percent Error(MPE) and Root Mean Square Error(RMSE) for 500 randomly selected sampling points. The results indicated that the two methods exhibited similar accuracy and spatial patterns at a large scale. The prediction performance of multiple linear regression was better than that of Kriging. The former described more accurately the spatial distribution of soil nutrients. This suggested that the multiple linear regression was an efficient method for predicting the spatial distribution of soil nutrients. Areas with high values of soil nutrient contents in the investigation region were mainly associated with soil types such as diluvial soil, paddy soil, and Lou soil and better hydro-thermal conditions such as lowly terrain, the slope gradient of 0 ~8°, the aspect of east and southeast. This study would provide an important support for soil resource management, soil monitoring and fertilization.

相似文章

导出引用
计量
文章访问数: 4
HTML全文浏览量: 0
PDF下载量: 0

所属期刊

推荐期刊

友情链接Friendly link
农业信息研究所Agricultural information institute 农业智能知识服务平台Agricultural intelligent knowledge service platform 农业学术服务平台Agricultural Knowledge Service System 中国知网CNKI 维普网VIP 万方WANFANG DATA
联系我们Contact us

电话Tell：010-82109652-808

邮箱E-mail：agrijournal@caas.cn

服务Service
关于我们About us 意见建议Opinions and suggest
版权说明Copyright
单位：中国农业科学院农业信息研究所 中国农学会 © 2017中国农业科学院农业信息研究所 ICP备案号：京ICP备09089781号-46