Temporal, spatial characteristics and uncertainty of biogenic VOC emissions in the Pearl River Delta region, China

Abstract

Using the Global Biosphere Emissions and Interactions System model (GloBEIS), 3 × 3 km gridded and hourly biogenic volatile organic compound (BVOC) emissions in the Pearl River Delta (PRD) were estimated for the year 2006. The study used newly available land cover database, observed meteorological data, and recent measurements of emission rates for tree species in China. The results show that the total BVOC emission in the PRD region in 2006 was 296 kt (2.2 × 1011 gC), of which isoprene contributes about 25% (73 kt, 6.4 × 1010 gC), monoterpenes about 34% (102 kt, 8.9 × 1010 gC), and other VOCs (OVOC) about 41% (121 kt, 6.8 × 1010 gC). BVOC emissions in the PRD region exhibit a marked seasonal pattern with the peak emission in July and the lowest emission in January, and are mainly distributed over the outlying areas of the PRD region, where the economy and land use are less developed. The uncertainties in BVOC emission estimates were quantified using Monte Carlo simulation; the results indicate high uncertainties in isoprene emission estimates, with a relative error of -82 to +177%, ranging from 12.4 to 186.4 kt; -41 to +58% uncertainty for monoterpenes emissions, ranging from 67.7 to 181.9 kt; and -26 to +30% uncertainty in OVOC emissions, ranging from 88.8 to 156.2 kt on the 95% confidence intervals. The key uncertainty sources include emission factors and the model empirical coefficients α, CT1, CL, and Eopt for estimating isoprene emission, and emission factors and foliar density for estimating monoterpenes and OVOC emissions. This implies that determining these empirical coefficient values properly and conducting more field measurements of emission rates of tree species are key approaches for reducing uncertainties in BVOC emission estimates. Improving future BVOC emission inventory work in the PRD region requires giving priority to research on shrub land, coniferous forests, and irrigated cropland and pasture. © 2010 Elsevier Ltd.