Summer evapotranspiration in western Siberia: a comparison between eddy covariance and Penman method formulations

Fleischer Elisa, Bölter Jana, Klemm Otto

Abstract

Evapotranspiration is difficult to quantify because of the many factors and complex processes that influence it. Several empiricalmethods have been developed over the years to estimate potential evapotranspiration based on easily available parameters.Directly measured data of actual evapotranspiration have been rather sparse in the past and still need to be improved in particularregions like western Siberia. The transition zone between the warm temperate and cold temperate continental climates is verysensitive to climate change, and water stress is an increasingly important issue in these regions with a highly dynamic agriculturalactivity. So there is a growing need to estimate actual evapotranspiration. Widely usable approximations are needed. In thisstudy, the values of potential evapotranspiration computed with the original version, and eight modifications of the Penmanformulation were compared and related to the actual evapotranspiration measured by eddy covariance over a grassland area inwestern Siberia. The original 1948 and 1963 Penman formulations are best for estimating potential evapotranspiration in the transition zone between the forest steppes and the pre-taiga. A nearly linear relationship between the potential and actual evapotranspiration was found. A simple modification of the Penman equation (i.e. the multiplication of the result by a factor of0.47) is suggested for approximating the actual evapotranspiration based on standard meteorological data for the region.The original Penman formulation is most robust and will provide the widest applicability in the future under changing climateand environmental conditions. In this context, it is further recommended not to neglect the ventilation term of the Penman equation, which is often assumed to be negligibly small. A detailed correlation analysis showed that under dry soil conditions,the vegetation largely contributed to the actual evapotranspiration and, in contrast to widely held expectations, that the Penmanequation is best adapted to vegetated surfaces.