摘要
为了探讨地形因子对黑土地理化性质航空高光谱反演的影响,本研究以黑龙江省海伦市典型黑土地为例,开展12种地形因子与土壤有机质含量、全氮含量、全磷含量、全钾含量、pH值、阳离子交换量、全盐量的相关性研究,并对比不同建模方法在加入地形因子前后的预测精度。航空高光谱反演结果表明,海拔、坡度、山谷指数、地形粗糙指数、起伏度与多种土壤理化指标显著相关,在加入地形因子后,支持向量机法和随机森林法的模型预测精度稳定提升,而偏最小二乘法的模型预测精度却大幅下降。随机森林法的模型预测效果整体最好,与纯光谱反射率的反演模型相比,加入地形因子后全氮含量模型预测R2的提升幅度最大(提升了0.062),7种理化指标预测模型R2的平均提升幅度为0.036。总之,地形因子对提升黑土地理化性质的航空高光谱反演精度有积极作用,且适合用于机器学习方法进行反演。
In order to explore the effects of terrain factors on the airborne hyperspectral inversion of physical and chemical properties of black soil,the typical black soil area of Hailun in Heilongjiang province was selected as the research area.The correlation research of 12 terrain factors with soil organic matter content,total nitrogen content,total phosphorus content,total potassium content,pH,cation exchange capacity and total salt content was carried out,and the prediction accuracy of different modeling methods before and after adding terrain factors was compared.The results showed that there was a significant correlation between various physical and chemical properties and the terrain factors such as altitude,slope,valley index,terrain roughness index,fluctuation degree.After adding terrain factors,the model prediction accuracy of support vector machine method and the random forest method was improved steadily,but the prediction accuracy of partial least squares method was greatly decreased.In this study,the modeling effect of random forest method was the best.Compared with the pure spectral inversion model,the model prediction R2 of seven indices was generally improved after adding terrain factors,with an average increase of 0.036,while the increase of total nitrogen content model was the largest(0.062).On the whole,the terrain factor plays an active role in improving the accuracy of airborne hyperspectral inversion of the physical and chemical properties of black soil,and it is suitable to use machine learning method for inversion.
作者
赵宁博
杨佳佳
赵英俊
秦凯
杨越超
崔鑫
ZHAO Ning-bo;YANG Jia-jia;ZHAO Ying-jun;QIN Kai;YANG Yue-chao;CUI Xin(National Key Lab of Remote Sensing Information and Image Analysis Technology,Beijing Research Institute of Uranium Geology,Beijing 100029,China;Shenyang Center of Geological Survey,CGS,Shenyang 110000,China)
出处
《江苏农业学报》
CSCD
北大核心
2020年第6期1444-1451,共8页
Jiangsu Journal of Agricultural Sciences
基金
中国地质调查局东北黑土地1∶250000土地质量地球化学调查项目(DD20160316)
中国地质调查局兴凯湖平原及松辽平原西部土地质量地球化学调查项目(DD20190520)。
关键词
黑土地
理化性质
航空高光谱数据
地形因子
反演
black soil
physical and chemical properties
airborne hyperspectral data
terrain factors
inversion
