A review of retrieval of land surface evapotranspiration based on remotely sensed surface temperature versus vegetation index triangular/trapezoidal characteristic space

Land surface Evapotranspiration (ET) is an important component of surface water cycle and energy balance, and its accurate estimation is essential for agricultural irrigation and drought monitoring, water resources management and climate change prediction. Retrieval of land surface ET based remotely sensed surface temperature versus vegetation index triangular/trapezoidal characteristic space is one of the international hot spots and frontier topics in the quantitative remote sensing of surface evapotranspiration.Due to the unclear understanding of the soil evaporation/vegetation transpiration processes and mechanisms, the interpretation schemes of retrieval models differ significantly for the scientific question of how soil evaporation/vegetation transpiration in characteristic space changes with surface soil moisture and vegetation coverage (including cover type and coverage), resulting in significant differences in ET retrieval results between different models. This paper comprehensively, systematically and deeply reviews the research on surface ET retrieval and soil evaporation/vegetation transpiration separation based on surface temperature versus vegetation index triangular/trapezoidal characteristic space. The basic theories, advantages and disadvantages of each method for the determination of dry/wet edges and ET retrieval modeling are described in detail, the applicable conditions of each method are clarified, and the problems to be solved are sorted out. Based on these, this paper finally points out the future development direction of the remote sensing retrieval of ET in the triangular/trapezoidal characteristic space.To correctly reveal the variation law of land surface ET in the triangular/trapezoidal characteristic space of surface temperature-vegetation index and improve the accuracy and operational capability of surface ET estimation based on characteristic space, several suggestions for future research are proposed as follows: First, using model simulation data, ground observation data, and combining model comparison, theoretical analysis, mathematical derivation to comprehensively and thoroughly study the agreement, differences and compatibility between the triangular and trapezoidal characteristic space in terms of the physical concepts, causes, and retrieval results. Then clarifying the applicability, the connection and difference in the separation of soil evaporation/vegetation transpiration between the "simultaneous separation method" and the "two-stage separation method", and investigating the variation of ET and soil evaporation/vegetation transpiration in characteristic space with the change of soil moisture and vegetation index. Finally, with the use of long-term remote sensing data and the successful application of the negative correlation between land surface temperature and vegetation index (i.e., spatial information) in the retrieval of soil moisture, developing a new model for remote sensing retrieval of ET of a characteristic space without the determination of dry and wet edges or end-member land surface temperature. Meanwhile, based on the recently proposed shortwave infrared reflectance-vegetation index trapezoidal characteristic space, which has the advantages of high soil moisture retrieval accuracy and less impact by atmospheric forcing changes, the establishment of surface ET retrieval models can be based on the shortwave infrared reflectance-vegetation index characteristic space.Through this paper, it is helpful to further understand the mechanism of land surface ET retrieval based on the triangle/trapezoid characteristic space, provide inspiration for the establishment of new methods for remote sensing retrieval of ET and soil evaporation/vegetation transpiration separation, and promote the research level of quantitative remote sensing of evapotranspiration in China. © 2021, Science Press. All right reserved.