MODELED AND FIELD-MEASUREMENTS OF BIOGENIC HYDROCARBON EMISSIONS FROMA CANADIAN DECIDUOUS FOREST

The Biogenic Emission Inventory System (BEIS) used by the United State s Environmental Protection Agency (Lamb et al., 1993, Atmospheric Envi ronment 21, 1685-1705; Pierce and Waldruff, 1991, J. Air Waste Man. As s. 41, 937-941) was tested for its ability to provide realistic microc limate descriptions within a deciduous forest in Canada. The microclim ate description within plant canopies is required because isoprene emi ssion rate's from plants are strongly influenced by foliage temperatur e and photosynthetically active radiation impinging on leaves while mo noterpene emissions depend primarily on leaf temperature. Model microc limate results combined with plant emission rates and local biomass di stribution were used to derive isoprene and alpha-pinene emissions fro m the deciduous forest canopy. In addition, modelled isoprene emission estimates were compared to measured emission rates at the leaf level. The current model formulation provides realistic microclimatic condit ions for the forest crown where modelled and measured air and foliage temperature are within 3 degrees C. However, the model provides inadeq uate microclimate characterizations in the lower canopy where estimate d and measured foliage temperatures differ by as much as 10 degrees C. This poor agreement may be partly due to improper model characterizat ion of relative humidity and ambient temperature within the canopy. Th ese uncertainties in estimated foliage temperature can lead to underes timates of hydrocarbon emission estimates of two-fold. Moreover, the m odel overestimates hydrocarbon emissions during the early part of the growing season and underestimates emissions during the middle and latt er part of the growing season. These emission uncertainties arise beca use of the assumed constant biomass distribution of the forest and con stant hydrocarbon emission rates throughout the season. The BEIS model , which is presently used in Canada to estimate inventories of hydroca rbon emissions from vegetation, underestimates emission rates by at le ast two-fold compared to emissions derived from field measurements. Th e isoprene emission algorithm proposed by Guenther et al. (1993), appl ied at the leaf level, provides relatively good agreement compared to measurements. Field measurements indicate that isoprene emissions chan ge with leaf ontogeny and differ amongst tree species. Emission rates defined as function of foliage development stage and plant species nee d to be introduced in the hydrocarbon emission algorithms. Extensive m odel evaluation and more hydrocarbon emission measurements from differ ent plant species are required to fully assess the appropriateness of this emission calculation approach for Canadian forests.