论文部分内容阅读
为了获取城市尺度组分温度,实现城市水热平衡的高精度反演,探索了一种多波段热红外遥感影像的城市尺度组分温度反演算法。算法选取了植被、土壤和不透水表面等3种组分,并且针对ASTER数据,利用线性混合像元分解方法获取像元平均比辐射率,以MODIS近红外数据估算大气水汽含量和大气透过率,采用牛顿迭代法获取大气平均温度,并用最小二乘原理获取地表组分温度。最后,应用长沙市区的实验影像进行了实验研究,通过纯净像元上组分温度反演结果与分裂窗算法反演结果的对比分析,以及组分温度反演结果与实测数据的对比分析,对算法的精度进行了验证,结果表明:(1)纯净像元上,组分温度反演结果与分裂窗算法反演结果具有较好的相关性,植被组分相关性最高,达0.9796,2种结果平均绝对偏差值为0.36℃;(2)组分温度反演结果与实测组分温度绝对偏差范围为0.2～1.4℃,植被组分温度与实测值偏差相对较小,不透水表面组分温度与实测值偏差相对最大。
In order to obtain the temperature of urban scale components and realize the high-precision inversion of urban water-heat balance, an urban temperature-based component temperature inversion algorithm for multi-band thermal infrared remote sensing imagery is explored. The algorithm selected three components of vegetation, soil and impervious surface. For ASTER data, the average emissivity of pixels was obtained by linear mixed pixel decomposition method. The atmospheric water vapor content and atmospheric transmittance were estimated by MODIS near infrared data , Using Newton’s iterative method to obtain the average temperature of the atmosphere, and using least squares principle to obtain the temperature of the surface composition. Finally, the experimental images of Changsha urban area were used to do the experimental research. By comparing the inversion results of component temperature and the split window algorithm on the pure pixel and the contrastive analysis of the component temperature inversion results with the measured data, The accuracy of the algorithm is verified. The results show that: (1) On pure pixels, the results of component temperature inversion have good correlation with the inversion results of split-window algorithm, the correlation of vegetation components is the highest, reaching 0.9796, 2 The average absolute deviation of the results was 0.36 ℃. (2) The absolute deviation of component temperature from the measured temperature was 0.2-1.4 ℃. The deviation of vegetation temperature from the measured value was relatively small. The deviation of temperature from the measured value is the largest.