1. Using in situ data to better understand Chinese air pollution events
Air pollution in China is a major health issue, resulting in over 1.3 million deaths per year.
New a emissions inventory and in situ datasets allow us to better understand air quality in China.
5/3/17
IGC8
Pictures taken in Beijing during summer 2016
Jonathan M. Moch, Loretta Mickley, Hong Liao, Yuan Cheng, Meng Li

2. DJF monthly means PM2.5 for 6 sites in North China Plain (2012-2013
In situ data and GEOS-Chem can help us understand Chinese pollution events
GEOS-Chem has previously been shown to do a bad job capturing winter PM2.5 variability and magnitude over the North China Plain (NCP).
A new standard simulation with MERRA2 meteorology and updated emissions inventories from Tsinghua University shows a better match with observations but now an overestimate of PM2.5.
DJF monthly means PM2.5 for 6 sites in North China Plain (
Nested China Simulation using MERRA2
DJF ( )
Each point represents monthly mean from an individual site
1:1
Model PM2.5 (μg/m3)
PM2.5 (μg/m3)
R = 0.86
MB = 38.6
Observed PM2.5 (μg/m3)
For 29 sites across China: R = 0.66; MB = 4.81

3. Setup of new China simulation
MERRA2 meteorology
Nested Asia grid (0.5x0.625°)
Multi-resolution Emissions Inventory for China (MEIC) for 2012 and 2013
China component of default MIX inventory for Asian emissions in GEOS-Chem v11.01
Individual power plant database with unit specific parameters
County level transportation information and digital road map
MEIC 2012 emissions relative to 2010 MIX emissions:
China is implementing policies such as closing coal plants in the North China Plain and attempting to reduce SO2 emissions, so total emissions and spatial distributions can change rapidly.
MEIC
Standard scaling
NOx
+1%
+5%
SO2
+3%
-1%
NH3
+8%
+0% BC OC
-7%
+4%
# ~ 41% reduction in SO2 emissions from 2006 levels according to MEIC inventory
# ~17% reduction in SO2 just between 2006 and 2010 according to published MEIC
# additional 29% decline in SO2 from 2010 to 2013 according to MEIC
# ~31% increase in NOx emissions from 2006 to 2013
# ~23% icrease in NOx from 2006 to 2010 and additional 7% from 2010 to 2013
Coal-fired power plant in Shanxi, 2015 (Getty Images)

4. Relatively good PM2.5 match hides component issues
DJF monthly mean components for North China Plain
Sulfate
Nitrate 2012
Each point represents monthly mean from an individual site
During winter
sulfate is underestimated and nitrate is overestimated across the North China Plain.
1:1
Model Sulfate (μg/m3)
Model Nitrate (μg/m3)
R = 0.72
MB =
R = 0.75
MB = 21.25
Organonitrate removal
Observed Sulfate (μg/m3)
Observed Nitrate (μg/m3)
Hypotheses to be tested:
Missing sulfate oxidation mechanism is leading to too little sulfate, which allows for excess nitrate in the presence of ammonia.
Excessively high ammonia emissions allow too much nitrate to enter the aerosol phase.
There is an issue with NOx partitioning or missing NOx sink

5. Comparison with Beijing daily data in 2011 shows similar bias
PM2.5 BC OC
Component daily means from Tsinghua
R = 0.64
MB = 7.29
Model daily mean
R = 0.74
MB = 0.5
R = 0.89
MB = 3.23
US Embassy
Observations
Concentration μg/m3
Sulfate
Nitrate
Ammonium
R = 0.93
MB = -2.26
R = 0.90
MB = 12.15
R = 0.93
MB = 2.57
Concentration μg/m3
December
December
December
Comparison of GEOS-Chem with daily observations in Beijing 2011 also show an overestimate of PM2.5 driven primarily by excess model nitrate during pollution episodes.
Component observations provided by Yuan Cheng

6. Switching gears to brown carbon: possibly an important absorber in China
Imaginary refractive index of BrC in Beijing, December 2011
Brown carbon is the light absorbing portion of OC.
Brown carbon and BC may play an important role in strengthening Chinese pollution episodes by absorbing radiation and stabilizing the planetary boundary layer.
Surface measurements indicate brown carbon in China may be much more absorbing than current parameters for brown carbon in GEOS-Chem anticipate.
Measured kOA,350
Diesel
Imaginary Refractive Index at 350 nm (kOA,350)
Lignite
December
MAKE SURE EXPLAIN IMAG INDEX
Satellite tuned (GEOS-Chem default)
Biomass burning
Reasons why kOA,350 might be variable:
Changing emissions
Secondary brown carbon formation
Photobleaching
Cheng et al., 2016; Hammer et al., 2016; Lu et al., 2015

7. Brown carbon may be a significant absorber at short wavelengths
GEOS-Chem simulated AAOD at 350nm in Beijing 2011
Using in situ mass absorption coefficient for BrC leads to BrC making up on average 30% of total simulated AAOD over December versus 16% if using static satellite derived values.
Depending on mixing state assumptions about BC, brown carbon could make up 20-50% of total absorption over Beijing at short wavelengths.
BrC AAOD using measured refractive index
Aerosol Absorbing Optical Depth at 350nm
BC AAOD
9% versus 7% at 440nm…
December
BrC AAOD with satellite retrieved refractive index

8. Next steps
Examine importance of increased nitrate removal for nitrate and ammonium overestimates.
Examine importance of ammonia emissions for nitrate and total PM2.5 overestimates.
Use measurements of surface BC and BrC and aerosol optical depths to constrain GEOS-Chem simulations of BrC and BC.
Estimate the effects of BrC and BC on regional climate.
Thanks to: Loretta Mickley, Hong Liao, Yuan Cheng, Meng Li, Shuxiao Wang, Chris Nielsen, Meng Gao, Melissa Sulprizio

9. Backup: 2012-2013 Monthly PM has R=0.64, with seasonal bias
DJF
MAM
PM2.5 (μg/m3)
PM2.5 (μg/m3)
R = 0.66; MB = 4.81
R = 0.59; MB = -6.63
JJA
SON
PM2.5 (μg/m3)
PM2.5 (μg/m3)
R = 0.67; MB = -5.77
R = 0.57; MB = -4.41

10. Backup slide: sulfate is underestimated across Eastern China in winter
NCP model range
NCP model mean
Monthly observations
Rough bounds for NCP
Sulfate (μg/m3)
DJF Sulfate (μg/m3) 35 30
2012
2013
NCP DJF R = 0.72, MB = 25 20
Sulfate underestimate is not just a problem in Beijing, but January 2013 sulfate was anomalously high. 15 10 5

11. Backup slide: nitrate overestimate across NCP is persistent
NCP model range
NCP model mean
Monthly observations
Total R = 0.69, MB = 15.92
DJF R = 0.75, MB = 21.25
MAM R = 0.10, MB = 9.50
JJA R = 0.71 MB =
SON R = 0.46, MB = 16.35
Nitrate (μg/m3)
2012
2013
Nitrate overestimate is largest in winter, and consistent across all of China.
Observed Nitrate (μg/m3)
Model Nitrate (μg/m3)