基于数据驱动和支持向量机的土壤水分含量反演研究

Inversion of Soil Moisture Content Inversion Based on Data Driven and Support Vector Machine

利用MODIS遥感数据和支持向量机回归算法,反演出研究区土壤含水量。利用分类结果与MODIS L1B日数据,采用Vswi方法反演高植被覆盖区土壤水分,并利用MODIS复合产品数据对低植被覆盖度采用表观热惯量模型。实测数据表明,反演精度大大提高。

Using MODIS remote sensing data and support vector regression algorithm,the soil moisture content of the whole research area was regressive.Using the classification results and MODIS L1 Bdaily data,the soil moisture in high vegetation coverage area was retrieved by Vswi method,and the apparent thermal inertia model was used for low vegetation coverage using MODIS composite product data.The measured data show that the accuracy of inversion is greatly improved.