Abstract:
Land evapotranspiration (ET) is an essential component of the climate system, but has been long linked with uncertainties in most regions of the world due to a lack of reference global observations. This has important consequences for climate research and hydrology: Land ET constitutes one of the missing terms of the global water cycle, and land-atmosphere exchanges have been shown to strongly impact climate-change projections, in particular with regard to changes in climate variability and extreme events. Several recently developed datasets based in part on observations, now provide global ET estimates: dedicated satellite-derived products, observation-driven land surface model simulations, reanalysis data products, and datasets based on atmospheric water balance estimates.

The LandFlux-EVAL initiative (http://www.iac.ethz.ch/url/research/LandFlux-EVAL), which is part of the LandFlux activity spearheaded by the GEWEX radiation panel, aims at evaluating and inter-comparing these newly available ET datasets. In this presentation, the rationale for the LandFlux-EVAL initiative will be highlighted, together with first results from on-going investigations and their implications. In particular we will provide estimations of biases and uncertainties in current land ET estimates, and corresponding evaluations of IPCC AR4 simulations.