国土资源遥感
國土資源遙感
국토자원요감
REMOTE SENSING FOR LAND & RESOURCES
2015年
1期
172-177
,共6页
Landsat TM%SUTM%E-DisTrad%地表温度分解%北京
Landsat TM%SUTM%E-DisTrad%地錶溫度分解%北京
Landsat TM%SUTM%E-DisTrad%지표온도분해%북경
Landsat TM%SUTM%E-DisTrad%LST decomposition%Beijing
如何综合可见光波段信息提高地表温度的空间分辨率一直是热红外遥感应用研究的重要方向。以北京市Landsat TM图像为数据源,对比分析了SUTM和E-DisTrad模型地表温度分解的空间特征差异性和适用范围。结果表明:在植被覆盖较低、地表温度较高的中心城区,SUTM模型的地表温度分解效果更佳,最小均方根误差和平均绝对误差分别为1.522 K和1.191 K;在植被覆盖较高、地表温度较低的郊区,E-DisTrad模型的地表温度分解效果更好,最小均方根误差和平均绝对误差分别为1.768 K和1.173 K。2种模型都能有效地提高地表温度的空间分辨率,但是在植被覆盖不同的地区分解结果呈现一定的差异性。
如何綜閤可見光波段信息提高地錶溫度的空間分辨率一直是熱紅外遙感應用研究的重要方嚮。以北京市Landsat TM圖像為數據源,對比分析瞭SUTM和E-DisTrad模型地錶溫度分解的空間特徵差異性和適用範圍。結果錶明:在植被覆蓋較低、地錶溫度較高的中心城區,SUTM模型的地錶溫度分解效果更佳,最小均方根誤差和平均絕對誤差分彆為1.522 K和1.191 K;在植被覆蓋較高、地錶溫度較低的郊區,E-DisTrad模型的地錶溫度分解效果更好,最小均方根誤差和平均絕對誤差分彆為1.768 K和1.173 K。2種模型都能有效地提高地錶溫度的空間分辨率,但是在植被覆蓋不同的地區分解結果呈現一定的差異性。
여하종합가견광파단신식제고지표온도적공간분변솔일직시열홍외요감응용연구적중요방향。이북경시Landsat TM도상위수거원,대비분석료SUTM화E-DisTrad모형지표온도분해적공간특정차이성화괄용범위。결과표명:재식피복개교저、지표온도교고적중심성구,SUTM모형적지표온도분해효과경가,최소균방근오차화평균절대오차분별위1.522 K화1.191 K;재식피복개교고、지표온도교저적교구,E-DisTrad모형적지표온도분해효과경호,최소균방근오차화평균절대오차분별위1.768 K화1.173 K。2충모형도능유효지제고지표온도적공간분변솔,단시재식피복개불동적지구분해결과정현일정적차이성。
Land surface temperature ( LST) is a vital parameter controlling the energy and water balance between atmosphere and land surface. LST image with high spatial resolution compatible with visible bands of Landsat TM is very important for the application of the LST image to many studies such as environmental monitoring. This paper examines the accuracy and applicability of two widely-used models for decomposition of LST images:SUTM and E-Distrad. Landsat TM data acquired in Beijing were used for the study. LST retrieved by the mono -window algorithm ( MWA) was used to compare the LST decomposition images by the two models. The results achieved by the authors indicate that SUTM is more applicable than E-Distrad in the regions with low vegetation cover and high LST such as downtown, while the latter is better than the former in the high vegetation cover and relatively cold areas such as water bodies. The RMSE and MAE are 1. 522 K and 1. 191 K respectively for SUTM and 1. 768 K and 1. 173 K for E-Distrad. It is thus concluded that both models are applicable for decomposition of LST images for high spatial resolution, but the results of decomposition are different in areas of different vegetation covers.