【摘要】 近年来，使用无人机获取地表温度研究城市热环境的案例逐渐增多，但尚未形成统一的观测方法，甚至是否需要进行温度反演也未形成共识，严重影响了热环境分析结果的准确性。为解决此问题，对广州市某高校内的10种典型城市下垫面进行了无人机热红外遥感观测。通过遥感观测数据与接触测量数据对比，分析了热红外影像的光斑效应和观测高度对地表温度反演结果的影响。结果显示，在5种观测高度下，热红外影像测点取值直径为2/4/10 pixels时地表温度反演结果一致性较高，光斑效应对反演结果影响较小；所有下垫面的反演温度平均误差均小于9%,人工下垫面的反演精度均高于自然下垫面。当观测高度低于100 m时，反演温度平均误差小于5%,绝对误差小于2℃。更多还原

【Abstract】 In recent years, the number of cases using UAVs to obtain the land surface temperature for thermal environment research has gradually increased. However, no unified observation method has been formed, and there is not even a consensus on whether temperature inversion is needed, which seriously affects the accuracy of thermal environment analysis results. To solve this problem, UAVs thermal infrared remote sensing was conducted on ten urban typical underlying surfaces in a university in Guangzhou. And the impact of thermal infrared camera spot-size effect and observation height on the surface temperature retrieval results was analyzed by comparing with the contact measurement method. The results indicated that the retrieval results with the measurement spot in 2/4/10 pixels diameter are highly consistent under the five observation heights. The spot-size effect has little influence on the retrieval results. The Mean Bias Error of all the retrieved temperatures of the measured underlying surfaces at all observation heights are less than 9%. The accuracy of UAVs retrieval temperature for the artificial underlying surfaces is higher than that for the natural underlying surfaces. The Mean Bias Error is less than 5% and the Mean Absolute Error is less than 2 ℃ when the observation height is lower than 100 m.更多还原