Global perspective of the secondary organic aerosol lifecycle in the present-day and future

Abstract

Secondary organic aerosol (SOA) is a major component of aerosol. Aerosol affect the radiation budget of the Earth and are detrimental to human health. Robust assessments of the impact of SOA on air quality and climatic are hindered by uncertainties in the SOA lifecycle. There are approximately 37 million unique organic compounds in the atmosphere. With such a large number of potential sources of SOA, representing the SOA lifecycle in a global model is challenging. SOA schemes within models vary in several ways, including: the emissions source types considered, and how volatile organic compound (VOC) physicochemical processing is treated. As a result, estimates of the global SOA production rate from models and observations range several orders of magnitude. Furthermore, simulated SOA concentrations from global models are typically lower than observed. The objectives of this study are to (i) quantify the impact of biogenic, anthropogenic and biomass burning VOC emissions on the global SOA budget and model agreement with observations, (ii) explore the sensitivity of the global SOA budget and model agreement with observations to variations in the physicochemical processing of VOCs, and (iii) examine how future changes in climate and emissions influence the SOA lifecycle. Throughout this study, a global chemistry-climate model (UKCA) is used, developed, and tested against a suite of surface and aircraft observations Firstly, the impact of biogenic, anthropogenic and biomass burning VOC emissions on the global SOA budget and model agreement with observation is quantified. This is achieved by introducing new VOC emission source types, whilst maintaining a fixed VOC oxidation mechanism. As source of SOA, monoterpene (C10H16) has been studied under laboratory conditions extensively. This VOC is commonly included in SOA schemes and, in many cases, is the only source of SOA. In this study, when monoterpene is the only source of SOA, the simulated global SOA production rate is 20 Tg (SOA) a-1 and the normalised mean bias (NMB) with respect to observed SOA is -91 %. In response to the addition of new emission source types, isoprene (C5H8), a lumped anthropogenic VOC (VOCANT) and a lumped biomass burning VOC (VOCBB), to the SOA scheme the global SOA production rate increases by 275 % (to 75 Tg (SOA) a-1), and model agreement with observations improves (NMB = -51 %). The improvement in agreement between simulated and observed SOA is primarily due to the inclusion of VOCANT, as opposed to isoprene or VOCBB. These results demonstrate that, under a single-step oxidation scheme, with a fixed yield of SOA, biogenic, anthropogenic, and biomass burning VOC emissions account for around half of the observed SOA abundance. With the new SOA scheme which considers all major VOC sources of SOA, the next objective is to explore the sensitivity of the SOA budget and model agreement with observations to variations in the physicochemical processing of VOCs. This is achieved by performing simulations with varying VOC deposition and oxidation mechanisms, whilst maintaining fixed VOC emissions. In light of recent field and explicit modelling studies, the sensitivity of SOA to VOC deposition is quantified. By including both dry and wet deposition of all VOC precursors of SOA, the global SOA production rate from all VOC sources reduces by 37 % (to 47 Tg (SOA) a-1) and model agreement with observations worsens (NMB = -66 %). Hence, neglecting VOC deposition can have significant impacts on SOA formation. According to chamber and field studies, VOCs form SOA after several generations of oxidation, and with yields which are sensitive to nitrogen oxide (NOX) concentrations. Therefore, for the anthropogenic and biomass burning VOC precursors of SOA (VOCANT/BB), model simulations are performed varying (a) the parent VOC reactivity, (b) the number of reaction intermediates, and (c) accounting for the influence of NOX on SOA yields. Both variations in parent VOC reactivity and accounting for the NOX-sensitive SOA yields have a substantial impact on simulated SOA, whereas SOA is mostly unaffected by the introduction of the reaction intermediate. In response to these variations in oxidation, the global SOA production rate from VOCANT/BB ranges from 18 Tg (SOA) a-1 to 45 Tg (SOA) a-1 (+150 %) and the NMB with respect to observed SOA ranges from -46 to -71 %. SOA is extremely sensitive to variations in parent VOC reactivity and accounting for the NOX-sensitive SOA yields, but is unaffected by the introduction of the reaction intermediate. These simulations highlight how the use of simplified VOC oxidation mechanisms within SOA schemes can have profound impacts on the global SOA budget and model agreement with observations. Finally, the impact of future changes in climate and emissions on the SOA lifecycle is quantified. This is achieved by driving the UKCA model with the Intergovernmental Panel on Climate Change (IPCC) Representative Concentration Pathway (RCP) 8.5 for the 2090s and the present-day (2000s). Compared to the present-day, climate change alone results in a 23 % increase in the global SOA burden due to increases in both SOA production (10 %) and the SOA lifetime (12 %). This climate-induced increase in SOA production is driven by an 82 % increase in monoterpene emissions due to the warming associated with RCP8.5 (4.6 °C). Global isoprene emissions reduce by 1 % under future climate change due to the opposing effects of warming and rising carbon dioxide concentrations which suppress isoprene synthesis (‘CO2 inhibition’). Projected changes in anthropogenic and biomass burning emissions alone result in a 3 % decrease in the global SOA load compared to the present-day due to a reduction in the SOA production rate (-6 %) and an increase in the SOA lifetime (4 %). This emissions-driven reduction in global SOA production is driven by a projected 11 % decrease in anthropogenic and biomass burning VOCs. When future changes in climate and emissions are combined, the global SOA burden increases by 20 % in the future compared to the present-day, which is due to increases in SOA production (4 %) and a lengthening of the SOA lifetime (15 %). Therefore, these results imply a growth in the global SOA burden due to rising biogenic VOC emissions and a lengthening of the SOA lifetime, despite reductions in anthropogenic and biomass burning emissions. This thesis contributes to the understanding of the SOA lifecycle in the present-day and the future. In relation to uncertainty in SOA and the impacts of SOA on air quality and climate, greater observational constraints on VOC emissions, deposition and oxidation mechanisms are required. Furthermore, these results imply a growing importance of natural sources of SOA in the future.