蒸散发是水文循环的重要组成部分,获取高时空分辨率的数据能够更加精细化蒸散发的时空变化规律,对于水资源管理、生态水文过程量化具有重要意义。由于单一传感器反演的蒸散发无法同时具有高空间和高时间分辨率,以南京地区为例,首先结合Landsat-8遥感影像数据和气象数据,采用基于能量平衡原理的SEBS模型估算日蒸散量。在此基础上,选取典型区域采用基于增强型时空自适应反射融合模型(ESTARFM)将估算的蒸散发结果与低空间分辨率的MOD16A2蒸散发产品数据进行时空融合降尺度研究,并评价模型的融合精度。结果表明:(1)SEBS模型估算的蒸散发结果与蒸发皿折算后的数据、MOD16A2产品数据的平均相对误差分别为0.14 mm/d和0.22 mm/d。(2)南京地区蒸散量季节差异明显,表现为夏季＞秋季＞冬季;各区在夏季的日平均蒸散量差异也较大,六合区蒸散量最大,秦淮区最小;另外,蒸散量分布受土地利用类型的影响,总体上表现为水域＞林地＞耕地＞草地＞其他,且植被覆盖度较高的区域蒸散量较大。(3)基于ESTARFM模型融合的蒸散发结果与基于Landsat-8遥感影像反演的蒸散发数据在空间分布上具有相似性,二者相关系数为0.74。在全球气候变化的背景下,本研究可为蒸散发数据集时空分辨率的提高提供参考,同时也能够为南京地区水循环过程和水资源管理研究提供数据支撑。
Evapotranspiration is an important component of hydrological cycle. Obtaining high spatio-temporal resolution data can refine the spatio-temporal variation features of evapotranspiration, which is of great significance for the management of water resources and quantification of eco-hydrological processes. Considering the evapotranspiration retrieved by single sensor cannot have both high spatial and high temporal resolution, this paper takes Nanjing as a case area for studying the framework on fusing MOD16A2 evapotranspiration product data with high temporal resolution and the estimated evapotranspiration data from Landsat-8 image with high spatial resolution. First, combined with Landsat-8 remote sensing image data and meteorological data, the SEBS model based on energy balance principle is used to estimate the daily evapotranspiration. Then, the ESTARFM model is applied to perform spatio-temporal fusion downscaling between the estimated evapotranspiration data and MOD16A2 evapotranspiration product data in a selected typical area of 144 square kilometers in Nanjing and the fusion accuracy of the model is evaluated. The results show that: (1) the average relative error between the evapotranspiration result estimated by SEBS model and the conversed evaporating pan data is 0.14 mm/d, and the average relative error between the evapotranspiration result estimated by SEBS model and the MOD16A2 product data is 0.22 mm/d. (2) The seasonal difference of evapotranspiration in Nanjing is obvious, with the largest evapotranspiration occurring in summer and the second in autumn. The reason is that the vegetation coverage is the highest in summer with the largest area of leaves. And the temperature and precipitation are higher than that in autumn, which is facilitate to evapotranspiration. Evapotranspiration in winter is the smallest due to the lowest temperature. The daily average evapotranspiration of different administrative regions in summer is quite different, the evapotranspiration in Liuhe District is the largest and that in Qinhuai District is the smallest. The reason for this phenomenon is the different types of land use. As a whole, the evapotranspiration of water is the largest, which is greater than that of forestland and cultivated land. Evapotranspiration values of grassland and other land use type are relatively small. Areas with higher vegetation coverage have higher evapotranspiration. (3) The evapotranspiration result based on the fusion of the ESTARFM model and estimated evapotranspiration data based on the Landsat-8 remote sensing image are spatially similar, and the correlation coefficient between them is 0.74. Under the background of global climate change, this study can not only provide a framework of improving the temporal and spatial resolution of evapotranspiration data set, but also provide data support for studying water cycle process and water resources management in Nanjing.