Incorporating Biogenic Hydrocarbon Emission Inventories into Mesoscale Meteorological Models Shelley Pressley, Brian Lamb, and Hal Westberg Washington State University, Dept. of Civil and Environmental Engineering, Pullman, WA 99164-2910 spressle@wsunix.wsu.edu ABSTRACT Isoprene flux measurements have been collected since 1996 at the AmeriFlux site located at the University of Michigan Biological Station (UMBS) as part of the Program for Research on Oxidants: Photochemistry, Emissions, and Transport (PROPHET). The isoprene flux data show that there is a very strong linear correlation between daily isoprene emissions and sensible heat fluxes for the predominantly aspen/oak stand located in northern Michigan. Our hypothesis is that the surface energy flux is a better model parameter for estimating isoprene emissions at the canopy scale than temperature and light levels, and the link to the surface energy budget will provide an improvement in isoprene emission models. Since surface energy budgets are an integral part of mesoscale meteorological models, this correlation could potentially be a useful tool for incorporating biogenic emissions explicitly into regional atmospheric modeling systems. The correlation can also be useful as a diagnostic tool for current biogenic emission inventory models, to determine if we are correctly predicting isoprene for the right reasons. Because sensible heat flux is a surrogate for the integration of temperature and light through the depth of the canopy, heat flux is potentially another predictor of isoprene fluxes based on physiological/biological control mechanisms. In this paper, correlations between observed sensible heat flux and isoprene flux are presented along with the results of a multiple linear regression for predicting isoprene flux as a function of sensible heat flux and maximum daily heat flux. We also examine the performance of the Biogenic Emission Inventory System (BEIS) canopy model for isoprene. INTRODUCTION The importance of isoprene at urban, regional, and global scales of atmospheric chemistry is well established (Fehsenfeld et al., 1992; Guenther et al., 1995; Cowling et al., 1998; Guenther et al., 2000; Fuentes et al., 2000). In rural areas, isoprene is almost always the dominant reactive hydrocarbon. In order to understand the chemistry of rural atmospheres, it is essential that the emissions of isoprene be well characterized. However, the details of how much isoprene, from what ecosystems, and under what conditions, remain troublesome aspects of accurately portraying isoprene in chemical cycles. In many cases, our ability to model isoprene and other biogenic hydrocarbons is limited to an accuracy of approximately a factor of two. We know that isoprene is emitted at high rates from oak, aspen, poplar and at lower rates from other deciduous vegetation and from spruce (Guenther et al., 1994; Geron et al., 1994; Guenther et al., 1996; Kempf et al., 1996). We know that isoprene is emitted in the presence of sunlight and exhibits an exponential increase with temperature to a maximum level near 40 o C (Guenther et al., 1993). At the same time, we know that isoprene is not directly tied to photosynthesis; isoprene emissions can increase while photosynthesis ceases—at least in the short term (minutes to hours) (Monson and Fall, 1989). Yet, it has also been theorized that there is long term (few days) control of isoprene emissions linked to the average temperature and PPFD over the previous 48 hours (Sharkey et al, 1999). We know that the basal isoprene emission rate can be different from sunlit leaves in the top of a canopy compared to shady leaves deep in a canopy (Harley et al., 1996, 1997). We know that the onset of isoprene emissions is delayed for several weeks after bud break in the spring and that isoprene emissions decrease in the fall at the approach of leaf senescence (Monson et al., 1994; Goldstein et al., 1998; Fuentes et al., 1999).