Ethiopian Institute of Water Resources, Addis Ababa University, Ethiopia.
Department of Earth and Environment, Institute of Environment, Florida International University, Miami, USA.
Sci Total Environ. 2022 Feb 1;806(Pt 1):150443. doi: 10.1016/j.scitotenv.2021.150443. Epub 2021 Sep 21.
Understanding the suitability of Satellite Rainfall Estimates (SREs) in simulating high flows and Actual Evapotranspiration (AET) is crucial for developing flood monitoring systems. Therefore, this study aims to assess i) the suitability of SREs in simulating both high flows and AET for different levels of model complexity, and ii) the effect of streamflow calibration on simulating AET for different rainfall inputs in Melkakunitre catchment, Upper Awash Basin, Ethiopia. Three state-of-the-art SREs (TRMM 3B42v7, IMERG v06B, and TAMSAT v3) were used and their usefulness in simulating high flows (Q5), daily streamflow, and wet season flows (from June to September) was assessed using the HBV-light model for the period 2003-2015. The model was set up for two levels of complexity: with and without considering the effect of orography on rainfall and temperature. Moreover, the water balance derived AET was compared against three remotely sensed AET products, MOD 16A2, GLEAM v3, and SSEBob, so as to examine the effect of streamflow calibration on AET simulation. Results show that rainfall inputs and model complexity have a strong impact on simulating streamflow and AET. For all rainfall forcing datasets, the performance of the hydrological model improves when we consider the effects of orography on rainfall and temperature. The IMERG v06B and TAMSAT v3 products showed the highest and least performances in simulating all the three flow conditions, respectively. Moreover, the MODIS-AET is the best remotely sensed AET product in reproducing the water balance-derived AET for all rainfall inputs except TAMSAT v3. The HBV-light model parameters calibrated with streamflow provided better results for simulating AET as well. On average, the usefulness of the IMERG v06B product for simulating high flows and AET is outstanding and can be thus used for developing flood monitoring and management systems in the study catchment.
了解卫星降水估计值（SRE）在模拟高流量和实际蒸散量（AET）方面的适用性对于开发洪水监测系统至关重要。因此，本研究旨在评估：i）SRE 在模拟不同模型复杂度水平下的高流量和 AET 的适用性，以及 ii）在埃塞俄比亚上 Awash 流域 Melkakunitre 集水区，考虑到不同降雨输入时，对流量进行校准对模拟 AET 的影响。本研究使用了三种最先进的 SRE（TRMM 3B42v7、IMERG v06B 和 TAMSAT v3），并使用 HBV-light 模型评估了它们在模拟 2003-2015 年期间的高流量（Q5）、日流量和湿季流量（6 月至 9 月）的适用性。模型的建立分为两个复杂程度：考虑和不考虑地形对降雨和温度的影响。此外，将水量平衡得出的 AET 与三种遥感 AET 产品 MOD 16A2、GLEAM v3 和 SSEBob 进行了比较，以检验对流量进行校准对 AET 模拟的影响。结果表明，降雨输入和模型复杂度对模拟流量和 AET 有很大影响。对于所有降雨输入数据集，当我们考虑地形对降雨和温度的影响时，水文模型的性能都会得到提高。在模拟所有三种水流条件方面，IMERG v06B 和 TAMSAT v3 产品的性能最高和最低。此外，除了 TAMSAT v3 之外，MODIS-AET 是所有降雨输入中复制水量平衡得出的 AET 的最佳遥感 AET 产品。使用流量校准的 HBV-light 模型参数在模拟 AET 方面也能取得更好的效果。总体而言，IMERG v06B 产品在模拟高流量和 AET 方面非常有用，因此可用于开发研究流域的洪水监测和管理系统。
