基于Landsat 8数据的亚热带小流域蒸散发估算及时空特征分析

Estimation and Spatio-Temporal Distribution of Evapotranspiration in Small-Scaled Catchments in Subtropics of China Based on Landsat 8 Data

以亚热带典型小流域——金井流域为研究对象,应用Landsat 8遥感数据基于SEBS模型估算了不同季节4个典型日的蒸散发(ET)值。结果表明,春季(5月12日)、夏季(7月31日)、秋季(9月17日)和冬季(次年1月23日)典型日的日ET均值及90%分位数上界分别为2.69和4.14、2.73和5.41、2.75和5.78、1.33和3.25 mm·d^(-1)。为研究ET在不同土地利用类型中的分布差异性,统计了研究区不同土地利用类型的日ET值,结果显示水体日ET值最高〔(4.21±2.63)mm·d^(-1)〕,稻田〔(2.67±1.49)mm·d^(-1)〕次之,再次是林地和茶园〔(2.06±1.35)mm·d^(-1)〕,公路和居民地最低〔(1.17±1.14)mm·d^(-1)〕。对所选4个典型日的ET和归一化植被指数(NDVI)进行相关分析,表明两者存在显著相关性(r值为0.34~0.63,P<0.05)。将ET遥感估算值与Penman-Monteith公式经验值、涡度相关系统的野外观测值进行对比,3者结果较为一致,表明SEBS模型适用于研究区ET估算。

Based on the Landsat 8 remote-sensing data of the Jinjing Catchment,typical of the subtropics of China, evapotranspiration （E T ） in the catchment on 4 days,typical of the four seasons,was estimated separately,using the SEBS model. Results show that the mean and the upper bound of 90% quantile of ET was estimated to be 2. 69 and 4. 14 mm on the spring day （ May 12,2013）,2. 73 and 5. 41 mm on the summer day （July 31,2013）,2. 75 and 5. 78 mm on the autumn day （September 17,2013）,and 1. 33 and 3. 25 mm on the winter day （January 23,2014）, respectively. To study variation of the distribution of ET as affected by land use,statistics ＂were done of ETs varying ＂with type of land use, exhibiting a decreasing order of water body [（4. 21±2. 63） mm·d^-1〉 paddy field [（2. 67±1.49） mm·d^-1 〉 wood land and tea garden [（2. 06±1. 35）mm·d^-1 ] 〉 highway and residential land [（1. 17±1. 14） mm·d^-1 ]. Pearson correlation analysis shows that ET was significantly related （ r = 0. 34-0. 63, P 〈0. 05） to normalized difference vegetation index （ND- VI） on the four typical days. Besides, remotesensing based estimation of ET was quite consistent with that using the Pen- man-Monteith equation or the eddy correlation method, and the values acquired in field observation, which suggests that the SEBS model is applicable to estimation of ET in the Jinjing Catchment.