Uncertainty assessment of potential biogenic volatile organic compound emissions from forests with the monte Carlo Method : Case study for an episode from 1 to 10 July 2000 in Poland

[1] An uncertainty assessment of a volatile organic compounds (VOCs) emission inventory using a Monte Carlo study according to the “Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories” has been performed. For the episode of 1–10 July 2000 hourly biogenic VOC (BVOC) emissions from forests in Poland were calculated in 10 km × 10 km resolution with a semiempirical BVOC model (seBVOC). Driving parameters of the model were land cover, temperature, light intensity, foliar biomass, leaf area index (LAI), and plant-specific emission factors. The hourly meteorology input has been modeled with the nonhydrostatic Multiscale Climate Chemistry Model (MCCM). For each of the driving parameters, probability distribution functions (PDFs) based on the normal and log-normal distributions have been identified. Repeated runs of the seBVOC model in the Monte Carlo study with random figures drawn from the probability distribution functions yield the error distribution and the uncertainties. The results show an uncertainty in isoprene emission of the entire modeled period and modeling domain in the range from −71% to 73%, in monoterpene emissions in the range of −57% to 140%, and in other VOC (OVOC) emissions in the range of −55% to 57%. Uncertainties in daily estimates for the domain were higher ranging between −84% and 98% for isoprene, −63% and 147% for monoterpenes, and 63% and 72% for other VOCs. Largest uncertainty results from errors of the emission factors followed by errors in temperature and foliar biomass. These uncertainties cover only a subset of possible variables and are less than the total uncertainty.