Summary Results from:
Global estimates of evapotranspiration for climate studies using multi-sensor remote sensing data: Evaluation of three process-based approaches
As they relate to the validation of mod16
Authors: Vinukollu, R.K., Wood, E.F., Ferguson, C.R. and Fisher, J.B.
Source: Remote Sensing of Environment, Volume 115, Issue 3, Pages 801-823
Link to: Access Publication
Abstract:
Three process based models are used to estimate terrestrial heat fluxes and evapotranspiration (ET) at the global scale: a single source energy budget model, a PenmanÐMonteith based approach, and a PriestleyÐTaylor based approach. All models adjust the surface resistances or provide ecophysiological constraints to account for changing environmental factors. Evaporation (or sublimation) over snow-covered regions is calculated consistently for all models using a modified Penman equation. Instantaneous fluxes of latent heat computed at the time of satellite overpass are linearly scaled to the equivalent daily evapotranspiration using the computed evaporative fraction and the day-time net radiation. A constant fraction (10% of daytime evaporation) is used to account for the night time evaporation. Interception losses are computed using a simple water budget model. We produce daily evapotranspiration and sensible heat flux for the global land surface at 5 km spatial resolution for the period 2003–2006. With the exception of wind and surface pressure, all model inputs and forcings are obtained from satellite remote sensing.
Satellite-based inputs and model outputs were first carefully evaluated at the site scale on a monthly-mean basis, then as a four-year mean against a climatological estimate of ET over 26 major basins, and finally in terms of a latitudinal profile on an annual basis. Intercomparison of the monthly model estimates of latent and sensible heat fluxes with 12 eddy-covariance towers across the U.S. yielded mean correlation of 0.57 and 0.54, respectively. Satellite-based meteorological datasets of 2 m temperature (0.83), humidity (0.70), incident shortwave radiation (0.64), incident longwave radiation (0.67) were found to agree well at the tower scale, while estimates of wind speed correlated poorly (0.17). Comparisons of the four year mean annual ET for 26 global river basins and global latitudinal profiles with a climatologically estimated ET resulted in a Kendall's τ > 0.70. The seasonal cycle over the continents is well represented in the Hovmöeller plots and the suppression of ET during major droughts in Europe, Australia and the Amazon are well picked up. This study provides the first ever moderate resolution estimates of ET on a global scale using only remote sensing based inputs and forcings, and furthermore the first ever multi-model comparison of process-based remote sensing estimates using the same inputs.
Satellite-based inputs and model outputs were first carefully evaluated at the site scale on a monthly-mean basis, then as a four-year mean against a climatological estimate of ET over 26 major basins, and finally in terms of a latitudinal profile on an annual basis. Intercomparison of the monthly model estimates of latent and sensible heat fluxes with 12 eddy-covariance towers across the U.S. yielded mean correlation of 0.57 and 0.54, respectively. Satellite-based meteorological datasets of 2 m temperature (0.83), humidity (0.70), incident shortwave radiation (0.64), incident longwave radiation (0.67) were found to agree well at the tower scale, while estimates of wind speed correlated poorly (0.17). Comparisons of the four year mean annual ET for 26 global river basins and global latitudinal profiles with a climatologically estimated ET resulted in a Kendall's τ > 0.70. The seasonal cycle over the continents is well represented in the Hovmöeller plots and the suppression of ET during major droughts in Europe, Australia and the Amazon are well picked up. This study provides the first ever moderate resolution estimates of ET on a global scale using only remote sensing based inputs and forcings, and furthermore the first ever multi-model comparison of process-based remote sensing estimates using the same inputs.
