In:
Atmospheric Chemistry and Physics, Copernicus GmbH, Vol. 18, No. 10 ( 2018-05-28), p. 7393-7422
Abstract:
Abstract. The global secondary organic aerosol (SOA) budget is highly uncertain, with
global annual SOA production rates, estimated from global models, ranging
over an order of magnitude and simulated SOA concentrations underestimated
compared to observations. In this study, we use a global composition-climate
model (UKCA) with interactive chemistry and aerosol microphysics to provide
an in-depth analysis of the impact of each VOC source on the global SOA
budget and its seasonality. We further quantify the role of each source on
SOA spatial distributions, and evaluate simulated seasonal SOA concentrations
against a comprehensive set of observations. The annual global SOA production
rates from monoterpene, isoprene, biomass burning, and anthropogenic
precursor sources is 19.9, 19.6, 9.5, and 24.6 Tg (SOA) a−1,
respectively. When all sources are included, the SOA production rate from all
sources is 73.6 Tg (SOA) a−1, which lies within the range of
estimates from previous modelling studies. SOA production rates and SOA
burdens from biogenic and biomass burning SOA sources peak during Northern
Hemisphere (NH) summer. In contrast, the anthropogenic SOA production rate is
fairly constant all year round. However, the global anthropogenic SOA burden
does have a seasonal cycle which is lowest during NH summer, which is
probably due to enhanced wet removal. Inclusion of the new SOA sources also
accelerates the ageing by condensation of primary organic aerosol (POA),
making it more hydrophilic, leading to a reduction in the POA lifetime. With
monoterpene as the only source of SOA, simulated SOA and total organic
aerosol (OA) concentrations are underestimated by the model when compared to
surface and aircraft measurements. Model agreement with observations improves
with all new sources added, primarily due to the inclusion of the
anthropogenic source of SOA, although a negative bias remains. A further
sensitivity simulation was performed with an increased anthropogenic SOA
reaction yield, corresponding to an annual global SOA production rate of
70.0 Tg (SOA) a−1. Whilst simulated SOA concentrations improved
relative to observations, they were still underestimated in urban
environments and overestimated further downwind and in remote environments.
In contrast, the inclusion of SOA from isoprene and biomass burning did not
improve model–observations biases substantially except at one out of two
tropical locations. However, these findings may reflect the very limited
availability of observations to evaluate the model, which are primarily
located in the NH mid-latitudes where anthropogenic emissions are high. Our
results highlight that, within the current uncertainty limits in SOA sources
and reaction yields, over the NH mid-latitudes, a large anthropogenic SOA
source results in good agreement with observations. However, more
observations are needed to establish the importance of biomass burning and
biogenic sources of SOA in model agreement with observations.
Type of Medium:
Online Resource
ISSN:
1680-7324
DOI:
10.5194/acp-18-7393-2018
Language:
English
Publisher:
Copernicus GmbH
Publication Date:
2018
detail.hit.zdb_id:
2092549-9
detail.hit.zdb_id:
2069847-1
Permalink
