A modeling approach for reconstruction of annual land surface evapotranspiration using palaeoecological data

A modeling approach to reconstruct the annual land surface evapotranspiration from palaeoecological data was suggested. It is based on assumption that the actual evapotranspiration is proportional to potential surface evapotranspiration and to some decoupling factor characterizing the surface moisture conditions. It was described in our study as a function of the climate moisture index (CMI). The potential evapotranspiration rate was derived using palaeoecological data about past land use and land cover, forest coverage, plant species composition and mean annual air temperature. The value of unknown decoupling factor was approximated using the results of evapotranspiration measurements at selected FLUXNET experimental sites located in the areas with different (ranged from extremely wet to moderately dry) moisture conditions. Comparisons of modeled evapotranspiration rates with results of the field flux measurements showed their good agreement for various forest, shrubland and grassland ecosystems in different geographical regions.