无人机热红外反演土壤含水率的方法

以不同生育期夏玉米为对象,讨论无人机热红外反演夏玉米田土壤含水率的精度及反演方法.利用无人机获取试验区的可见光和热红外图像.通过可见光图像提取冠层掩膜并叠加在热红外图像上提取玉米冠层温度,分析冠层温度的变化趋势及与叶面积指数(LAI)的相关性.最后,利用冠气温差的相反数与叶面积指数构建了一个新指标(D_TL),讨论了冠气温差或D_TL指标反演土壤含水率的准确性.结果表明:冠层温度随着土壤含水量的增加而降低,夏玉米LAI在一定程度上可以表征冠层温度;对比4个时期的数据,发现冠气温差反演效果在灌溉后较好(如2次灌后R²分别为0.614 6和0.463 7);与冠气温差相比,D_TL指标可以提高土壤含水量反演的精度,如0~20 cm深度的R²从0.614 6和0.463 7提高到0.661 6和0.485 0.该研究对热红外反演夏玉米田间土壤含水率方法进行了新的尝试.

Inversion of soil moisture content based on UAV thermal infrared image

This study takes summer maize in different growth periods as the research object. The accuracy and inversion methods of the UAV thermal infrared inversion of soil moisture content in summer maize fields were discussed. The visible light and thermal infrared images of the test area were acquired by UAV. The canopy mask was extracted from the visible light image and superimposed on the thermal infrared image to extract the corn canopy temperature. The change trend of the canopy temperature and its correlation with the leaf area index(LAI)were analyzed. Finally, a new index(D_TL)was constructed by using the inverse number of atmospheric temperature difference and leaf area index, and the accuracy of retrieving soil water content by canopy atmospheric temperature difference or the D_TL index was discussed. The experimental results showed that using visible light image to obtain the canopy mask and superposed with thermal infrared image to extract the maize canopy thermal infrared could effectively improve the accuracy of canopy temperature extraction. In the same period, canopy temperature decreases with the increase of soil moisture content(regardless of marginal effect). Summer maize LAI can represent canopy temperature to a certain extent. Comparing the data of the four periods, it is found that the inversion effect of canopy air temperature difference is better after irrigation(for example, R² was 0.614 6 and 0.463 7). Compared with canopy air temperature difference, the D_TL index inversion can improve the accuracy of soil moisture retrieval. For example, R² was improved from 0.614 6 and 0.463 7 to 0.661 6 and 0.485 0 at depth of 0-20 cm. This study is a new attempt to retrieve the soil moisture content of summer maize field by thermal infrared.