Chen Yutong, Fok Hok Sum, Ma Zhongtian, Tenzer Robert
School of Geodesy and Geomatics, Wuhan University, Wuhan 430079, China.
Key Laboratory of Geospace Environment and Geodesy, Ministry of Education, Wuhan University, Wuhan 430079, China.
Sensors (Basel). 2019 Aug 1;19(15):3386. doi: 10.3390/s19153386.
Total basin discharge is a critical component for the understanding of surface water exchange at the land-ocean interface. A continuous decline in the number of global hydrological stations over the past fifteen years has promoted the estimation of total basin discharge using remote sensing. Previous remotely sensed total basin discharge of the Yangtze River basin, expressed in terms of runoff, was estimated via the water balance equation, using a combination of remote sensing and modeled data products of various qualities. Nevertheless, the modeled data products are presented with large uncertainties and the seasonal error characteristics of the remotely sensed total basin discharge have rarely been investigated. In this study, we conducted total basin discharge estimation of the Yangtze River Basin, based purely on remotely sensed data. This estimation considered the period between January 2003 and December 2012 at a monthly temporal scale and was based on precipitation data collected from the Tropical Rainfall Measuring Mission (TRMM) satellite, evapotranspiration data collected from the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite, and terrestrial water storage data collected from the Gravity Recovery and Climate Experiment (GRACE) satellite. A seasonal accuracy assessment was performed to detect poor performances and highlight any deficiencies in the modeled data products derived from the discharge estimation. Comparison of our estimated runoff results based purely on remotely sensed data, and the most accurate results of a previous study against the observed runoff revealed a Pearson correlation coefficient (PCC) of 0.89 and 0.74, and a root-mean-square error (RMSE) of 11.69 mm/month and 14.30 mm/month, respectively. We identified some deficiencies in capturing the maximum and the minimum of runoff rates during both summer and winter, due to an underestimation and overestimation of evapotranspiration, respectively.
流域总径流量是理解陆地 - 海洋界面地表水交换的关键组成部分。在过去十五年中，全球水文站数量持续减少，这推动了利用遥感技术估算流域总径流量。此前长江流域的遥感总径流量是以径流表示的，它是通过水量平衡方程，结合不同质量的遥感和模型数据产品估算得出的。然而，模型数据产品存在很大的不确定性，而且遥感总径流量的季节误差特征很少被研究。在本研究中，我们完全基于遥感数据对长江流域的总径流量进行了估算。该估算考虑了2003年1月至2012年12月的月度时间尺度，数据基于从热带降雨测量任务（TRMM）卫星收集的降水数据、从中分辨率成像光谱仪（MODIS）卫星收集的蒸发散数据以及从重力恢复与气候实验（GRACE）卫星收集的陆地水储量数据。进行了季节精度评估，以检测模型数据产品在流量估算中的不佳表现并突出任何不足之处。将我们完全基于遥感数据估算的径流结果，与之前一项研究针对观测径流的最准确结果进行比较，结果显示皮尔逊相关系数（PCC）分别为0.89和0.74，均方根误差（RMSE）分别为11.69毫米/月和14.30毫米/月。我们发现，由于蒸发散分别被低估和高估，在捕捉夏季和冬季径流率的最大值和最小值方面存在一些不足。
