Abstract. To better characterize anthropogenic emission-relevant aerosol
species, the Gridpoint Statistical Interpolation (GSI) and Weather Research and
Forecasting with Chemistry
(WRF/Chem) data assimilation system was updated from the
GOCART aerosol scheme to the Model for
Simulating Aerosol Interactions and Chemistry (MOSAIC) 4-bin (MOSAIC-4BIN) aerosol scheme. Three years (2015–2017) of
wintertime (January) surface PM2.5 (fine particulate matter with an aerodynamic
diameter smaller than 2.5 µm) observations from more than 1600 sites
were assimilated hourly using the updated three-dimensional
variational (3DVAR) system. In the control
experiment (without assimilation) using Multi-resolution Emission Inventory for China 2010 (MEIC_2010) emissions, the modeled
January averaged PM2.5 concentrations were severely overestimated
in the Sichuan Basin, central China, the Yangtze River Delta and the Pearl River
Delta by 98–134, 46–101, 32–59 and 19–60 µg m−3,
respectively, indicating that the emissions for 2010 are not appropriate for
2015–2017, as strict emission control strategies were implemented in recent
years. Meanwhile, underestimations of 11–12, 53–96 and
22–40 µg m−3 were observed in northeastern China, Xinjiang
and the Energy Golden Triangle, respectively. The assimilation experiment
significantly reduced both high and low biases to within
±5 µg m−3. The observations and the reanalysis data from the assimilation experiment
were used to investigate the year-to-year changes and the driving factors.
The role of emissions was obtained by subtracting the meteorological impacts
(by control experiments) from the total combined differences (by assimilation
experiments). The results show a reduction in PM2.5 of
approximately 15 µg m−3 for the month of January from 2015 to
2016 in the North China Plain (NCP), but meteorology played the dominant role
(contributing a reduction of approximately 12 µg m−3). The
change (for January) from 2016 to 2017 in NCP was different; meteorology
caused an increase in PM2.5 of approximately
23 µg m−3, while emission control measures caused a decrease
of 8 µg m−3, and the combined effects still showed a
PM2.5 increase for that region. The analysis confirmed that
emission control strategies were indeed implemented and emissions were
reduced in both years. Using a data assimilation approach, this study helps
identify the reasons why emission control strategies may or may not have an
immediately visible impact. There are still large uncertainties in this
approach, especially the inaccurate emission inputs, and neglecting
aerosol–meteorology feedbacks in the model can generate large uncertainties
in the analysis as well.