Texas A&M University, College Station, TX, USA.
Texas A&M University, College Station, TX, USA.
Sci Total Environ. 2020 Nov 15;743:140702. doi: 10.1016/j.scitotenv.2020.140702. Epub 2020 Jul 9.
Water resource development opens up opportunities for improving smallholder farmer livelihoods in sub-Saharan Africa; however, implementation of water resource interventions to ensure sustainability hinges on the availability of sufficient quantity and quality data for monitoring, analysis and planning. Such data is often acquired through instrumentation of water resources (e.g. stream flow monitoring) or the use of hydrological models. In sub-Saharan Africa, data scarcity has limited the ability to monitor and make appropriate decisions for water resource allocation and use. Data derived from remote sensing has been considered a viable option to fill this gap; however, there is limited research in the region that evaluate the quality of the remotely sensed based datasets. This study evaluated actual evapotranspiration (AET) estimates derived from Advanced Very High Resolution Radiometer (AVHRR AET) images and Moderate Resolution Imaging Spectrometer (MOD16 AET) images using estimates from a grid-based Soil and Water Assessment Tool (SWAT). The SWAT model was set up for the entire country of Ethiopia, and calibrated and validated using observed streamflow at several meso-scale watersheds in which satisfactory model performance was obtained. AET estimates from the calibrated and validated SWAT model were then used to evaluate remotely sensed based AET for three landscapes. The AVHRR AET better agreed with the SWAT-simulated AET than the MOD16 AET, although the AVHRR AET overestimated the SWAT-simulated AET in all of the landscapes. Both remotely sensed AET products showed better agreement with the SWAT-simulated AET over agriculture dominated landscapes compared to grassland and forest dominated landscapes. The findings of the study suggest that remotely sensed based AET may help to fine-tune hydrological models in agricultural landscapes in data-scarce regions to improve studies on the impacts of water management interventions aiming to ensure environmental sustainability while enhancing agricultural production, and household income and nutrition.
水资源开发为改善撒哈拉以南非洲小农的生计提供了机会；然而，为确保水资源干预措施的可持续性，需要有足够数量和质量的数据来进行监测、分析和规划。这些数据通常是通过水资源仪器（例如流量监测）或水文模型获得的。在撒哈拉以南非洲，数据的缺乏限制了对水资源分配和利用进行监测和做出适当决策的能力。遥感数据已被认为是填补这一空白的可行选择；然而，该地区评估基于遥感的数据集质量的研究有限。本研究评估了高级甚高分辨率辐射计（AVHRR AET）图像和中等分辨率成像光谱仪（MOD16 AET）图像从实际蒸散量（AET）估算得出的实际蒸散量（AET）估算值，使用基于网格的土壤和水评估工具（SWAT）的估算值。SWAT 模型在整个埃塞俄比亚进行了设置，并使用几个中尺度流域的观测流量进行了校准和验证，其中获得了令人满意的模型性能。然后，使用经过校准和验证的 SWAT 模型的 AET 估算值来评估三种景观的基于遥感的 AET。AVHRR AET 比 MOD16 AET 更符合 SWAT 模拟的 AET，尽管在所有景观中，AVHRR AET 都高估了 SWAT 模拟的 AET。与草原和森林为主的景观相比，这两种遥感 AET 产品在农业为主的景观中与 SWAT 模拟的 AET 更吻合。研究结果表明，基于遥感的 AET 可以帮助在数据稀缺地区的农业景观中微调水文模型，以改善旨在确保环境可持续性同时提高农业生产、家庭收入和营养的水资源管理干预措施的影响研究。
