单 位：
西北农林科技大学旱区农业水土工程教育部重点实验室;西北农林科技大学水土保持研究所;西北农林科技大学水利与建筑工程学院
关键词：
玉米;土壤含水率;覆盖度;热红外遥感;可见光;冠层温度
摘 要：
为提高基于冠层温度信息反演土壤含水率的精度,以不同水分处理的拔节期大田玉米为研究对象,采用无人机热红外和可见光相机获取试验区遥感图像,通过不同图像分类方法剔除土壤背景,提取玉米植被覆盖度(Corn vegetation coverage,Vc)及冠层温度(Canopy temperature,Tc),并计算冠-气温差(Tca)和冠-气温差与覆盖度的比值,分析这两种指数与土壤含水率(Soil moisture content,Smc)之间的关系。结果表明,不同分类方法提取的玉米覆盖度以及冠层温度均存在差异,由灰度分割法、RGRI指数法、GBRI指数法3种分类方法剔除土壤背景后计算的冠-气温差、冠-气温差与覆盖度之比与土壤含水率均呈线性关系,并且冠-气温差、冠-气温差与覆盖度之比两种指数反演0～30 cm玉米根域深度的土壤含水率效果较好;其中,未剔除土壤背景的冠-气温差反演土壤含水率效果较差,GBRI指数分类法剔除土壤背景的冠-气温差反演土壤含水率效果有所提高(0～10 cm、10～20 cm、20～30 cm深度的R2由0. 255、0. 360、0. 131提高至0. 425、0. 538、0. 258);而冠-气温差与覆盖度的比值反演土壤含水率相比于冠-气温差精度明显提高,0～10 cm、10～20 cm、20～30 cm深度建模集R2高达0. 488、0. 600、0. 290,P <0. 001,验证集R2达0. 714、0. 773、0. 446,表明冠-气温差与覆盖度之比是反演玉米根域深度土壤含水率效果更优的指标。
译 名：
Influence of Coverage on Soil Moisture Content of Field Corn Inversed from Thermal Infrared Remote Sensing of UAV
作 者：
ZHANG Zhitao;XU Chonghao;TAN Chengxuan;BIAN Jiang;HAN Wenting;Key Laboratory of Agricultural Soil and Water Engineering in Arid Areas,Ministry of Education,Northwest A&F University;College of Water Resources and Architectural Engineering,Northwest A&F University;Institute of Soil and Water Conservation,Northwest A&F University;
关键词：
corn;;soil moisture content;;coverage;;thermal infrared remote sensing;;visible light;;canopy temperature
摘 要：
In order to improve the accuracy of retrieving soil moisture content based on canopy temperature information,taking the different moisture treatment of the jointing field corn as the research object,and the UAV thermal infrared and visible light camera were used to obtain the remote sensing images of the experimental area. Different image classification methods were applied to remove the soil background and extract corn coverage and canopy temperature, then the indices such as crowntemperature difference and the ratio of crown-temperature to coverage were calculated, and the relationship between the two indices and soil moisture content was analyzed subsequently. The results showed that there were differences in corn coverage extracted by different classification methods,and there were also differences in corn canopy temperature extracted by different classification methods. The crown-temperature difference, crown-temperature difference to coverage ratio calculated by three classification methods(Grayscale segmentation,RGRI index,GBRI index) had a linear relationship with soil moisture content,and it was better to invert the soil moisture content of 0 ～ 30 cm corn root depth by the two indices; the crown-temperature difference without removing the soil background held the worst effect on soil moisture content,while removing soil background by GBRI index classification enjoyed the better effect on the soil moisture content(R2 was improved from 0. 255,0. 360 and 0. 131 to 0. 425,0. 538 and 0. 258 at depth of 0 ～ 10 cm,10 ～ 20 cm and 20 ～ 30 cm); the ratio of crown-temperature difference to coverage inversion of soil moisture content performed much better than that of crown-temperature difference. At the depth of 0 ～ 10 cm,10 ～ 20 cm and 20 ～ 30 cm,R2 was 0. 488,0. 600 and 0. 290 in the model set,P < 0. 001,and R2 was 0. 714,0. 773 and 0. 446 in the verification set,indicating that the ratio of crown-temperature difference to coverage was a new indicator for reversing the effect of deep soil moisture in the corn root zone. This study provided a new method for inversion of the soil moisture content of corn in the field by thermal infrared remote sensing.
