An overview of satellite-based evapotranspiration products in the framework of the CEOS Land Product Validation Subgroup

Actual evapotranspiration (ET) is commonly the largest extractive term in the land surface water balance, thus representing a key component of any water management activity and water resource quantification. Unfortunately, in-situ ET observations are often expensive, sporadically collected, and representative only of local conditions. In this context, modelling approaches represent a widespread alternative for the characterization of ET over large areas and for log time periods. While most of the spatially-distribute ET estimation approaches relies on satellite data to some extent, not all these estimates can be considered as satellite ET products. Like other satellite-based datasets, ET estimates are indirect in nature, and often depend on modelling approaches characterized by a variety of approaches and input requirements integrating a mixture of satellite and non-satellite datasets. With continuous advancements and developments in satellite data, the number of continental to global satellite ET products are increasing and they are characterized by a vast variety of sensors and modelling methods. This increasing number of available ET products underscores the need for a concerted effort in defining the standards and protocols for validation and evaluation exercises, which is the main goal of the Committee on Earth Observation Satellite (CEOS) Land Product Validation (LPV) subgroup. In this research, an overview of the methodologies adopted for the assessment of satellite-based ET will be provided, with a focus on the key hypotheses and forcings representing the “satellite” component of the approaches. This overview will provide a common reference of what constitute a satellite-based ET product, to be investigated by the CEOS LPV subgroup in the definition of recommended protocols to assess the accuracy and reliability of current and future continental to global satellite ET datasets.