Controls on evaporation in a boreal spruce forest

The surface energy balance over a boreal spruce forest is analyzed using 3 yr of 30-min-averaged data collected during the 1994-96 Boreal Ecosystem-At mosphere Study experiment 40 km west of Thompson, Manitoba, to show the cli matic controls on surface evapotranspiration. The seasonal variation of eva poration is shown: lowest in spring when the ground is frozen, highest in s ummer (although daytime evaporative fractions are only 0.4), and lower agai n in fall after frost. The surface sensible heat flux in contrast is high i n spring, when evaporation is low. Evaporation is much higher when the surf ace, including the moss layer, is wet. At all temperatures (in summer), eva porative fraction falls with increasing light level, because of the high ve getative resistance of the forest system. Using a Monin-Obukhov formulation and a bulk vegetation model, the vegetative resistance for the boreal spru ce forest system is calculated. This bulk vegetative resistance decreases w ith increasing photosynthetic radiation, decreases sharply with relative hu midity, decreases with increasing surface water storage, and is lower on cl oudy days than on sunny days with the same incoming photosynthetic radiatio n. Vegetative resistance at its midmorning minimum is lower by a factor of 4 when the moss surface is very wet. As over grassland sites, the lower sur face resistance to evaporation directly influences the diurnal cycle of lif ting condensation level and cloud-base height, which are much lower on days with a wet surface. The reduction of vegetative resistance under cloudy sk ies at the same incoming radiation level presumably reflects the more effic ient use of diffuse radiation by the canopy for photosynthesis. Vegetative resistance is roughly doubled in spring, when the ground is frozen, and is higher in fall after frost. About 63% of the observed variance in vegetativ e resistance can be explained in terms of meteorological variables using mu ltiple linear regression. Some measurement issues are addressed in an appen dix. The residual in the energy balance falls with increasing wind speed, w hich may be due to a small (10%-15%) underestimation of the sensible and la tent heat fluxes at low wind speeds. During spring melt, however, this resi dual has a high daytime value of 30% of net radiation. The residual is also much higher on sunny days than on cloudy days.