Is forest albedo measured correctly?

The forest albedo is an important meteorological and ecological variable ch aracterising a forest ecosystem. In order to measure this variable two 2 pi -sensors are mounted above the forest canopy. The first sensor records an u pward radiant energy flux from the forest canopy into the atmosphere, and t he second one a downward radiant energy flux from the atmosphere into the f orest canopy. The ratio of their responses is usually interpreted as the fo rest albedo. This measurement method is included in standard observation pr ograms at many research stations. In the present paper we quantify an inade quacy of this method to evaluate an albedo of a forest stand for photosynth etically active region (PAR, 400-700 nm) of the solar spectrum. We reproduc ed the three-dimensional PAR radiative field above a coniferous forest stan d at the scientific research station "Solling" near Gottingen with high lev el of realism. A good agreement, between simulated sensor response to the s imulated radiative field and a real sensor response to the real radiative f ield was achieved. However, the calculated value of forest albedo was found to be systematically higher than its value derived from sensors on a clear sunny day by about a factor of 3 during morning and evening hours and by a bout a factor of 1.9 during day time. This deviation is mainly caused by fo rest heterogeneity. A daily mean of the measured solar radiant energy fluxe s the ratio of which is usually assumed to represent the albedo does not av erage the effect of the forest heterogeneity.