Quantifying evaporation of intercepted rainfall: a hybrid correction approach for eddy-covariance measurements

Abstract. Precipitation and interception have a significant influence on the reliability of eddy-covariance (EC) measurements, primarily of vapor fluxes. As evaporation data need to fit both to the energy and the water budget, a balanced approach is necessary to arrive at reasonable values of evaporation associated to interception. EC data of the investigated ICOS site DE-Tha (dense conifers) suggest a large and systematic underestimation of evaporation during and shortly after a rainfall event. Total evaporation of selected frost-free and liquid interception events accounted for only 24 % of precipitation, which is an untypically low proportion for a dense coniferous forest under a temperate climate. We show that our Rutter based 2D model approach, including spatially variable vegetation information, reproduces reliable estimates of interception evaporation to compare and integrate the results for different source areas. For the EC footprint area, modelled interception evaporation accounts for 45 % of precipitation for the evaluated events. The standard Bowen ratio based energy balance adjustment and the energy balance residual approach are not justified to account for underestimated fluxes during interception events. As a consequence, we propose a hybrid correction approach complementing EC measurements with our 2D model estimates of evaporation under interception conditions to adjust for underestimated fluxes of LE. Our approach uses LE as a link between the energy and water balance and provides appropriate evaporation rates from intercepted liquid precipitation for the analyzed forest ecosystem. The correct redistribution of the heat fluxes will lead to a better parametrization of surface fluxes in weather and climate models and supports to properly include land use in water management needs under climate change.