Deriving surface PM2.5 from satellite observations
We derive satellite-based PM2.5 (hereafter PM2.5_MAIAC)
over the northeastern US for 2011 by taking the product of daily average CMAQ
modeled PM2.5/AOD relationships
(PM2.5_CMAQ/AODCMAQ) with MAIAC AOD (AODMAIAC,
Eq. 1). These unconstrained PM2.5 estimates (Fig. 1) are
independent of surface observations. As PM2.5_MAIAC is
determined as the product of observed AODMAIAC and modeled
PM2.5_CMAQ/AODCMAQ, the spatial patterns of
PM2.5_MAIAC will be affected by the spatial variations in both
AODMAIAC and PM2.5_CMAQ/AODCMAQ. Figure 1a shows
the summertime average (June, July, and August, JJA) AODMAIAC at 1 km resolution overlaid
with AERONET observed AOD. While we find high AOD over some populated urban
areas such as New York City (NYC), high AODMAIAC is also found
over central New York State (NYS), away from major anthropogenic sources. In
CMAQ, PM2.5 (PM2.5_CMAQ) occurs over regions with major
anthropogenic sources such as NYC. AODCMAQ also shows a
latitudinal dependence, with higher AOD at lower latitudes, which reflects (1) relatively high emissions of aerosol and its precursors from anthropogenic
and biogenic sources over Maryland, Pennsylvania, and NYC and (2) latitudinal variations in RH
that affect aerosol hygroscopic growth. The modeled
PM2.5_CMAQ/AODCMAQ varies spatially (1 standard deviation
(SD) is 45 µg m-3 per unit of AOD), mainly driven by the spatial
variations in PM2.5_CMAQ (R=0.86). We find the overall spatial
pattern of satellite-derived PM2.5 correlates more strongly with modeled
PM2.5_CMAQ/AODCMAQ (R=0.97) than observed
AODMAIAC (R=0.8), suggesting that the large-scale spatial
variability reflects modeled rather than satellite-based distributions, at
least under our framework for the northeastern US in summer. The temporal
variability in PM2.5_MAIAC is also mainly
driven by variability in PM2.5_CMAQ/AODCMAQ (R=0.61),
with little temporal correlation between regional average
AODMAIAC and PM2.5_MAIAC (R=0.05, Fig. 2). At short timescales, the daily variability in regional
average PM2.5_MAIAC shows stronger
correlation with PM2.5_CMAQ/AODCMAQ in all seasons except
for JJA, when PM2.5_MAIAC values are driven by
variability in both AODMAIAC (R=0.5) and
PM2.5_CMAQ/AODCMAQ (R=0.4, Fig. 2). Summertime
AODMAIAC is higher than wintertime AOD by 50 %, while
summertime PM2.5_MAIAC is lower than in winter by 46 %.
Previous studies also found inconsistent seasonal cycles in AOD and
PM2.5 (Ford et al., 2013; Kim et al., 2015). We attribute the opposite
seasonal cycle in PM2.5_MAIAC and AODMAIAC to
three factors: (1) weak boundary layer ventilation in winter that leads to
sharp vertical gradients of aerosol distribution (Kim et al., 2015), (2) higher RH in summer that leads to larger hygroscopic growth, and (3) model
overestimates of PM2.5 (especially OC) in wintertime and underestimates
of PM2.5 in summertime, leading to an overestimate of the
winter-to-summer decrease in PM2.5_CMAQ/AODCMAQ (see
Sect. 3.3).
Summertime (JJA) average (a) MAIAC AOD
(AODMAIAC), (b) satellite-derived PM2.5
(PM2.5_MAIAC), (c) CMAQ model AOD
(AODCMAQ), (d) CMAQ model PM2.5
(PM2.5_CMAQ), and (e) CMAQ modeled
PM2.5/AOD
(PM2.5_CMAQ/AODCMAQ) ratio overlaid with
ground-based observations (AERONET, AQS, co-located AERONET and AQS sites)
over the northeastern US with zoom-in maps over the New York City region in the
upper left corner. (f) Density plot of AOD showing the distribution
of MAIAC, CMAQ, and AERONET observed AOD sampled at AERONET sites.
(g) Density plot of PM2.5 showing the distribution of
satellite-derived, CMAQ, and AQS observed PM2.5 sampled at AQS sites.
Regional 10-day running average of (a) MAIAC AOD
(AODMAIAC, blue), (b) CMAQ modeled
PM2.5/AOD relationship
(PM2.5_CMAQ/AODCMAQ, red), and
(c) satellite-derived PM2.5 (PM2.5_MAIAC,
green). The numbers on the upper left corner show the Pearson correlation
coefficients (R) of PM2.5_MAIAC with
PM2.5_CMAQ/AODCMAQ (red) and
AODMAIAC (blue).
While at larger spatial scales PM2.5_CMAQ/AODCMAQ
contributes more to the spatial and temporal variability in
PM2.5_MAIAC than AODMAIAC, at smaller scales,
over which we assume the spatial variability in PM2.5/AOD is
homogenous, incorporating fine-resolution satellite data reveals stronger
spatial gradients (e.g., enhancements along industrial corridors) than
PM2.5_CMAQ (Fig. 1b). In addition to refining spatial
resolution, satellite-derived PM2.5 can correct model summertime biases
in PM2.5. Observed AOD from AERONET and PM2.5 from AQS indicate an
overall underestimate in both AODCMAQ (Fig. 1c; normalized mean bias
(NMB) = -44 %) and PM2.5_CMAQ (Fig. 1d; NMB = -17 %) in summer. We find PM2.5_CMAQ/AODCMAQ is
overall consistent with the observed PM2.5/AOD sampled at co-located
AQS–AERONET sites (NMB = 0.9 %) as the ratio largely cancels out the
model underestimates in both PM2.5 and AOD. AOD distributions
retrieved from MODIS (AODMAIAC) agree better with AERONET AOD than
AODCMAQ (NMB = 5 %, Fig. 1f), though we find small low biases at
two sites in NYC and at most DRAGON sites over Maryland. Our
derived distribution of PM2.5_MAIAC is thus closer to
PM2.5 observed at AQS sites than PM2.5_CMAQ (NMB = 4.7 % vs. 44 % for PM2.5_CMAQ, Fig. 1g).
However, the PM2.5_MAIAC distribution is wider than
observed at AQS: the lowest 5 % is 5 vs. 7 µg m-3 for
PM2.5_MAIAC vs. AQS PM2.5, and the highest 5 % is
16 vs. 13 µg m-3. We find that PM2.5_MAIAC
is biased high over NYC, coastal regions of Massachusetts, the
borders of upstate New York, and northern Vermont. Evaluation of
PM2.5_MAIAC in other seasons shows larger biases and
uncertainties (Fig. S1 in the Supplement). In the following sections, we examine sources of
uncertainties and biases in satellite-derived PM2.5. We quantify the
uncertainties in terms of bias (systematic) and random uncertainty. The bias
uncertainty is linked to the overall accuracy, while the random uncertainty
reflects random fluctuations in the measurements or the imprecision of the
model resulting from imperfect modeling assumptions and simplifications.
Evaluation of satellite-observed AOD products
AODMAIAC in general agrees well with AERONET observations
(spatial R=0.83, temporal R=0.85, MB = -0.01, and
RMSE = 0.07). The performance of AODMAIAC evaluated at
northeastern US AERONET sites is consistent with the evaluation of
Superczynski et al. (2017) over North America (R=0.82,
MB = -0.008). We find, however, that AODMAIAC in winter
(December, January, and February, DJF) is biased high by 49 %
(MB=+0.02) on average. The wintertime overestimate is likely due
to residual snow contamination, which is below the detection limit, even
though we applied a stringent data quality filter to remove pixels flagged as
snow. We find the wintertime overestimate is most evident over northern
latitudes (e.g., AERONET sites in Massachusetts; NMB ranges from 80 % to
180 %), where snow occurs more often. The NMBs of AODMAIAC
are 15 % in March, April, and May (MAM), -5 % in JJA, and 17 %
in September, October, and November (SON), though the quantile range of the
error is large, suggesting that single observations have large random
uncertainties (Fig. 3). Taking the 1σ standard deviation of the
normalized biases as a metric of random uncertainty, we estimate the
uncertainties of daily satellite observations to be around 80 % in DJF,
60 % in MAM and SON, and 50 % in JJA. Spatial and/or temporal
averaging can reduce these random errors of satellite observations, which is
evidenced as the smaller spread of errors than for monthly averages at the
same spatial resolution, or daily data at coarser (10 km) resolution, but it
does not reduce the overall MB between AODMAIAC and
AODAERONET (Fig. 3). We find that spatially averaging
AODMAIAC to 10 km leads to an overall increase in
AODMAIAC. Temporal averaging, however, leads to an overall
decrease in AODMAIAC except for DJF, leading to a smaller
positive MB in SON (7 %) and MAM (7 %), but larger negative MB in JJA
(-8 %) and positive MB (67 %) in DJF.
Distribution of normalized biases of AODMAIAC evaluated
at 52 AERONET (including DRAGON, only available for JJA) sites in four
seasons of 2011 over the northeastern US using daily MAIAC AOD at 1 km
resolution and 10 km resolution and monthly average MAIAC AOD composite (only
including days when both satellite and AERONET measurements are available) at
1 km resolution. The box shows the interquartile range (IQR) while the whiskers
extend to show the rest of the distribution with outliers (points that are
either 1.5× IQR or more above the third quantile or below the first
quantile) removed. The red triangles show the seasonal mean normalized
biases. Note that the normalized bias is an asymmetric metric, for which model
overestimates are unbounded, whereas model underestimates are bounded by
-100 %; therefore the mean of normalized biases is typically higher
than the median of the normalized biases.
Relative importance of satellite AOD versus modeled PM2.5/AOD to
uncertainties in satellite-derived PM2.5
We have shown that both satellite AOD and modeled PM2.5/AOD are subject
to large uncertainties at the daily timescale. To directly compare the
relative importance of the biases of satellite AOD vs. model PM2.5/AOD
for the satellite-derived PM2.5, we scale the biases of modeled
PM2.5/AOD with daily AODMAIAC, so that the biases are expressed in
units of PM2.5 (µg m-3):
ΔPM2.5_AOD=AODMAIAC-AODAERONET×PM2.5_CMAQAODCMAQ.
We then scale the biases of AODMAIAC with the daily modeled
PM2.5/AOD relationship:
ΔPM2.5_Rel=PM2.5_CMAQAODCMAQ-PM2.5_AQSAODAERONET×AODMAIAC.
We can also interpret ΔPM2.5_AOD and ΔPM2.5_Rel as the changes in derived PM2.5 if we
use “true” observed AOD or PM2.5/AOD instead of AODMAIAC or
modeled PM2.5/AOD. As shown in Fig. 5a, mean biases caused by modeled
PM2.5/AOD are +9.2 µg m-3 in DJF, +2.8 µg m-3
in MAM, -3.3 µg m-3 in JJA, and +7.7 µg m-3 in SON, which introduces larger biases to the derived PM2.5 than
the MAIAC satellite product in all seasons (7.6 µg m-3 in DJF,
+1.3 µg m-3 in MAM, -0.7 µg m-3 in JJA, and 0.9 µg m-3 in SON). Using the root-mean-squared ΔPM2.5_AOD to quantify the random uncertainty, satellite
AOD contributes an overall random error of 8.3 µg m-3 to daily
satellite PM2.5_MAIAC with the smallest error in JJA
(5.1 µg m-3) and largest error in DJF (13.2 µg m-3),
while modeled PM2.5/AOD contributes an error of 10.8 µg m-3
(root-mean-squared ΔPM2.5_Rel), with the smallest
error in JJA (6.5 µg m-3) and largest error in SON (15.2 µg m-3). The spread of the biases is larger for modeled PM2.5/AOD
than that for MAIAC AOD except for DJF. Our findings are consistent with
Ford and Heald (2016), who use a higher-resolution (nested) version of
the GEOS-Chem model and MODIS Dark Target AOD (Collection 6) to estimate 2
times larger uncertainties in surface PM2.5 resulting from modeled
PM2.5/AOD relationships than in satellite AOD.
(a) Box plots comparing the distribution of biases in daily
PM2.5_MAIAC due to observational uncertainties in
AODMAIAC (green, ΔPM2.5_AOD) versus model
uncertainties in PM2.5_CMAQ/AODCMAQ (blue,
ΔPM2.5_Rel), evaluated consistently at 11 co-located
AQS–AERONET sites over the northeastern US. (b) Pearson correlation
coefficient between the biases in daily satellite-derived PM2.5
(ΔPM2.5_MAIAC, evaluated with AQS observations) and the
biases in PM2.5_AOD attributed to observational uncertainties
in AODMAIAC (ΔPM2.5_AOD) versus model
uncertainties in PM2.5_CMAQ/AODCMAQ
(ΔPM2.5_Rel). ΔPM2.5_AOD is
calculated by multiplying the biases of AODMAIAC with daily
modeled PM2.5/AOD relationships (Eq. 8).
ΔPM2.5_Rel is calculated by multiplying the modeled
PM2.5/AOD biases with daily AODMAIAC (Eq. 9). The
red triangles show the seasonal mean biases.
At the daily timescale, both ΔPM2.5_AOD and
ΔPM2.5_Rel show large day-to-day variability:
the 1σ standard deviation is 10.5 µg m-3 for daily ΔPM2.5_AOD and 8.3 µg m-3 for daily ΔPM2.5_Rel. Next, we evaluate the dependence of the
biases of satellite-derived PM2.5 (denoted as ΔPM2.5_MAIAC; evaluated with AQS observed PM2.5) on
ΔPM2.5_Rel versus ΔPM2.5_AOD by evaluating the Pearson correlation
coefficients (R). Overall, ΔPM2.5_MAIAC is more
strongly correlated with ΔPM2.5_Rel (R=0.85)
than with ΔPM2.5_AOD (R=0.53),
indicating the uncertainties of modeled PM2.5/AOD are a more important
driving factor to the uncertainties of daily satellite-derived PM2.5,
which could explain 72 % variance (R2) in ΔPM2.5_MAIAC. In JJA, however, ΔPM2.5_MAIAC is moderately correlated with both ΔPM2.5_AOD (R=0.48) and ΔPM2.5_Rel (R=0.49), suggesting uncertainties of
modeled PM2.5/AOD and satellite AOD contribute equally to the
uncertainties of satellite-derived PM2.5. We note that there is no
statistically significant correlation between ΔPM2.5_Rel and ΔPM2.5_AOD,
with R ranging from -0.4 in SON to 0.23 in JJA, which suggests that the
errors caused by modeled PM2.5/AOD and by satellite AOD are independent
of each other.
Factors leading to uncertainties in modeled PM2.5/AOD
relationship
Uncertainties in the modeled PM2.5/AOD relationship mainly reflect
uncertain aerosol speciation, aerosol vertical profiles, ambient RH, and
parameterizations for aerosol optical properties including aerosol density,
size distribution, refractive index, and hygroscopic growth. Here we quantify
the uncertainties from each factor and evaluate their impacts on the derived
PM2.5.
Aerosol speciation
Aerosol optical properties vary with chemical composition. Model biases in
the aerosol composition also affect the overall representation of particle
hygroscopicity. For the same PM2.5 abundance, variations in the aerosol
composition may alter the particle optical properties, especially
hygroscopicity, and consequently the PM2.5/AOD relationship. Figure 6a
compares the modeled aerosol composition with ground-based observations
averaged for each season. High biases in PM2.5_CMAQ in
winter are largely due to model overestimates of OC by a factor of 3,
and low biases in summer are due to a combination of underestimated SNA and
OC. As a result, CMAQ overestimates the fraction of OC by about 20 % in
DJF, 15 % in MAM, and less than 10 % in other seasons, while it
underestimates the fraction of SNA by 5 % to 20 % in all seasons.
(a) Seasonal average PM2.5 speciation from CMAQ vs.
AQS observations in 2011 evaluated at 54 CSN and IMPROVE sites.
(b) Box plots showing the distribution of estimated biases of daily
satellite-derived PM2.5 due to model biases in PM2.5 speciation
(ΔPM2.5_spe) by season for 2011. Red triangles show the
seasonal mean biases. (c) Pearson correlation coefficient between
the biases in PM2.5_MAIAC (ΔPM2.5_MAIAC)
and ΔPM2.5_spe.
To estimate the impacts of model biases in aerosol speciation on
AODCMAQ and PM2.5_MAIAC, we keep the total aerosol
mass the same and redistribute AOD (AODCMAQ_ir) of each
species i based on the observed fraction of each species (i.e., SNA, OC, elemental carbon (EC), soil
dust; sea salt was excluded due to the limited ground-based measurements and
its negligible contribution):
AODCMAQ_ir=PMi_obsPMTOT_obs×PMTOT_CMAQPMi_CMAQ×AODCMAQ_i,
where PMTOT_obs and PMTOT_CMAQ are the
total aerosol mass from observations and CMAQ, respectively, which are
reconstructed by summing up SNA, OC, EC, and soil dust. Next, we estimate the
uncertainty due to speciation as the differences in derived
PM2.5_MAIAC (ΔPM2.5_spe)
using the redistributed AODCMAQ_ir instead of the original
AODCMAQ, shown in Fig. 6b. As SNA generally has the largest mass
extinction efficiency (MEE), a low bias in SNA leads to an overall underestimate
of AODCMAQ and therefore an overestimate of PM2.5_MAIAC, which is largest in winter (MB = 2.2 µg m-3,
SD = 2.6 µg m-3) and smallest in summer (MB = 0.7 µg m-3,
SD = 3.0 µg m-3). The estimated biases due to speciation show
seasonal cycles similar to those of the modeled PM2.5/AOD biases (Fig. 4),
suggesting that aerosol speciation errors contribute to the seasonality in
modeled PM2.5/AOD biases. Overall, model–observation discrepancy in
speciation causes an error (root-mean-squared ΔPM2.5_spe) of 4.0 µg m-3. On a daily basis,
the correlation (R) between ΔPM2.5_spe and
ΔPM2.5_MAIAC is over 0.5 for all seasons except
JJA, which means model biases in speciation alone can explain more than
25 % variance (R2) in ΔPM2.5_MAIAC.
Biases in speciation in JJA have relatively smaller impacts on the derived
PM2.5, which contribute less than 1 µg m-3 MB and show weak
correlation with ΔPM2.5_MAIAC (R=0.15).
Aerosol vertical profile
A caveat on the results in the Sect. 3.5.1 is that we assume the model
errors in speciation are constant across all vertical layers, as AQS sites
only provide observations near the surface. The DISCOVER-AQ aircraft
campaign measured vertical variations in aerosol composition, although
spatial and temporal coverage is limited. Figure 7a compares the modeled and
observed vertical distributions of SNA, OC, and BC averaged over the
DISCOVER-AQ campaign. We do not discuss sea salt and dust here since they
contribute a negligible portion of the total aerosol mass in this region.
Both the model and observations show SNA contributes more than half of the total
aerosol across all vertical layers (Fig. 7). Aircraft observations show SNA
decreases gradually with altitude with a nearly constant vertical gradient,
while SNACMAQ is well mixed below 1.5 km and starts to decline at the
same rate as SNAaircraft above 1.5 km (Fig. 7). CMAQ underestimates
SNA below 1.5 km but overestimates SNA at higher altitudes. The positive
model bias of SNA at higher altitudes may be due to excessive vertical
transport or overestimation of RH (Sect. 3.4.3) and consequently
overestimation of SO2 oxidation rate and aerosol water uptake. OC, however, is biased low at all altitudes, which is likely due to
inaccurate treatment of the production of secondary organic aerosol (SOA)
(Zhang et al., 2009). The newer version of CMAQv5.1 shows higher SOA
concentration in summer with the introduction of new SOA species (Appel et
al., 2017). BC is generally low during the campaign (typically lower than
0.3 µg m-3). BCCMAQ generally agrees well with
BCaircraft, though BCCMAQ tends to overestimate BC between 1
and 3 km. Figure 7b compares CMAQ modeled and observed total aerosol mass
(SNA + OC + BC) averaged during the campaign. CMAQ modeled aerosol mass
is on average biased low below 2 km and biased high at higher altitudes
(Fig. 7b).
Campaign-mean vertical profiles of (a) aerosol
composition, (b) total mass (SNA + OC + BC), and
(c) extinction from CMAQ vs. observations from the DISCOVER-AQ 2011
Baltimore–Washington, D.C. campaign. (d) Campaign-mean vertical
profile of the model-to-observation ratio of extinction
(RatioEXT), total aerosol mass (RatioMass), and
RatioEXT/RatioMass. Aircraft observations are
first aggregated to match model layers, and corresponding model values are
sampled concurrently with the time of observations. CMAQ modeled extinction
is estimated with FlexAOD using the default parameters in Table 1. The
shading in (b) and (c) shows the standard deviation of the
day-to-day variability.
Next, we evaluate how the vertical distribution of aerosols relates to
extinction. Figure 7c compares the modeled and observed average vertical
extinction profiles. We find, consistent with the biases in mass, a low bias
in the modeled extinction profile below 2 km and a high bias above (Fig. 7c).
The biases in extinction and the biases in mass have the same signs for more
than 80 % of data pairs and are strongly correlated (R=0.85). This
suggests that the aerosol vertical profile of extinction is mainly
indicative of mass distribution. However, column AOD measures the vertical
integral of light extinction by aerosols, which means the modeled AOD biases
would be proportional to modeled surface PM2.5 biases only if the
biases in extinction are constant across all vertical layers. Since the
biases of extinction change sign at higher altitude, the AOD biases reflect
the competing effects of negative biases near the surface and positive
biases at high altitudes, which lead to an overall negative bias of the
PM2.5/AOD relationship, consistent with the negative NMB of
PM2.5/AOD in July shown in Fig. 4.
To explore the causes of the model–observation discrepancy in extinction and
the resulting impacts on the satellite-derived surface PM2.5, we
calculate the vertical extinction profile in CMAQ by replacing the modeled
aerosol mass distribution (SNA, OC, BC), total MEE (ratio of total mass to extinction), or RH with those of the
aircraft observations, as shown in Fig. 8a. Replacing the modeled aerosol
mass with observations, we find a decrease in extinction at high altitudes
(above 2.5 km) and increase at low altitudes (below 2.5 km), but replacing
the aerosol mass alone does not explain all of the model–observation
differences. At high altitudes, only replacing the modeled total MEE without changing the mass captures the observed
extinction. We attribute the model overestimate of extinction to model
overestimation of extinction efficiency at high altitudes. A major
contributor to the model overestimate of total MEE is its excessive RH at
high altitudes, which leads to an overestimate of the hygroscopic growth.
Replacing RH with observations largely corrects the high biases aloft but
does not correct the low biases below 2 km (Fig. 8a). At lower
altitudes, the model low biases are due to model underestimates of both
aerosol mass and total MEE. Model underestimates of MEE are likely due to
(1) uncertain optical properties, (2) other aerosols or gases (e.g., NO2,
O3), or liquid clouds that can scatter or absorb light.
(a) Campaign-mean vertical profiles of extinction
calculated from CMAQ speciated aerosol fields using FlexAOD and
those calculated by replacing modeled speciated aerosol mass (Mass), total column
mass (Column), vertical profile shape (Profile), total mass extinction
efficiency (MEE), and relative humidity (RH) with those observed during the
DISCOVER-AQ 2011 Baltimore–Washington, D.C. campaign. EXTamb is
the aircraft observed vertical extinction profile. (b) Box plots of
the distribution of biases of PM2.5_MAIAC attributed to each
factor shown in (a), and the biases of PM2.5_MAIAC
attributed to modeled PM2.5/AOD (Rel). Red triangles show the
mean biases. (c) Pearson correlation coefficient between the biases
in modeled PM2.5/AOD relationships and the biases in modeled
PM2.5/AOD attributed to individual factors shown in
(b).
Figure 8b shows the biases of PM2.5_MAIAC due to model
uncertainties in vertical profiles of aerosol mass or MEE or RH, estimated
by calculating the changes in PM2.5 when we replace the model vertical
profiles with observations. Since the aircraft altitude ranges from 0.3 to
3.4 km, we use modeled values for the layers below 0.3 and above 3.4 km
while attempting to minimize the discontinuity at both boundaries through
vertical interpolation. As SNA and OC contribute most to extinction, we
also evaluate the biases of vertical profiles of SNA and OC separately. We
find that replacing modeled aerosol mass with observed mass leads to small
positive biases in PM2.5_MAIAC (MB = 0.05 µg m-3, SD = 4.3 µg m-3), due to the combined effects of
negative biases from SNA (MB = -2.5 µg m-3, SD = 4.7 µg m-3) and positive biases from OC (MB = +1.9 µg m-3,
SD = 4.3 µg m-3).
We further separate the model–observation discrepancy in the vertical
profiles as differences in total column mass versus in vertical profile
shape by (1) keeping the modeled vertical distribution but adjusting the mass
of each species uniformly so that the total column mass is equal to
observation and (2) keeping the total column mass the same as in the model but
redistributing the aerosol based on the observed vertical profiles. We find
that redistributing the aerosol vertical profile leads to a positive mean
bias in PM2.5_MAIAC (MB = 1.1 µg m-3, SD = 4.9 µg m-3), while the model–observation discrepancy in
column mass leads to a negative mean bias (MB = -0.6 µg m-3, SD = 3.6 µg m-3) (Fig. 8b). The positive biases in the profile
shape are mainly attributed to model biases of the vertical profile of SNA
(MB = 1.2 µg m-3, SD = 5.0 µg m-3), which shows a
larger fraction of SNA at higher altitude where aerosol is less effective at
scattering light due to lower RH. The negative MB of column mass reflects a
combination of negative biases of SNA (MB = -4.1 µg m-3, SD = 5.6 µg m-3) due to model overestimates of SNA column mass and
positive bias of OC (MB = 6.7 µg m-3, SD = 4.4 µg m-3) due to model underestimates of column mass of OC. Model biases in
MEE lead to a small positive MB of 0.6 µg m-3.
Using the observed PM2.5/AOD acquired from paired AQS–AERONET sites, we
estimate that model biases in modeled PM2.5/AOD lead to a negative MB
of -0.9 µg m-3 with large day-to-day variability (SD = 9.8 µg m-3) during the DISCOVER-AQ campaign, reflecting the model
biases from different sources as discussed above. Next, we evaluate which
factor drives the daily variability in the modeled PM2.5/AOD biases the
most by evaluating the R value between the estimated biases in modeled
PM2.5/AOD versus that attributed to individual factors. We find model
bias in aerosol mass is the most deterministic factor for the biases in
modeled PM2.5/AOD (R=0.82, Fig. 8c). Model biases in aerosol mass
can be due to either biases in column mass or vertical profile shape. We
find model biases in modeled PM2.5/AOD are more dependent on the biases
in aerosol column mass (R=0.79), instead of vertical profile shape.
Model biases in MEE show moderate correlation with
model biases of PM2.5/AOD (R=0.56). While model uncertainties in RH
lead to an overall negative bias (MB = -1.7 µg m-3, SD = 7.4 µg m-3) to PM2.5_MAIAC, they are negatively
correlated with model biases of PM2.5/AOD (R=-0.25).
RH
Figure 8 suggests model biases in RH contribute a negative bias to the
derived PM2.5_MAIAC during the DISCOVER-AQ aircraft
campaign. Here we evaluate the impacts of modeled RH (RHCMAQ) biases on
derived PM2.5 throughout the year using six atmospheric soundings over
the northeastern US. We only assess the impacts of RH on the optical
properties (i.e., hygroscopic growth) of aerosols. Comparing RHCMAQ with
observed RH (RHobs), RHCMAQ is overall biased high with the
largest biases in winter. To evaluate the resulting impacts on AODCMAQ,
we recalculate the extinction using observed ambient RH from the soundings
instead of RHCMAQ in Eq. (4). Replacing RHCMAQ with RHobs
decreases extinction by ∼50 % on average from the surface
to 5 km in both JJA and DJF (black lines in Fig. 9a and b). As AOD is the
vertical integral of extinction, the total area between EXTsonde and
EXTCMAQ (gray shading in Fig. 9a and b) indicates the differences in
AOD due to differences in RH. The differences in RH below 3 km in DJF, MAM,
and SON contribute more than 80 % to the total differences in AOD. In JJA,
the contribution from higher versus lower altitudes is similar, despite
small model RH biases below 2 km.
(a) DJF and (b) JJA average vertical profiles of
the CMAQ modeled versus observed RH at six atmospheric soundings over the
northeastern US, and the modeled extinction versus that calculated by replacing
modeled RH with observed values. The gray area shows the difference in the
two
extinction profiles, with the total area being the difference in AOD.
(c) Box plots showing the impacts of model bias of RH on the derived
PM2.5_MAIAC (ΔPM2.5_RH) in four seasons
of 2011, which are calculated by comparing the PM2.5_MAIAC
minus the one calculated using observed RH. (d) Box plots show the
influence of model RH biases on the derived PM2.5_MAIAC
(ΔPM2.5_RH) as a function of observed near-surface RH.
We evaluate how the model–observation discrepancy in RH affects the derived
PM2.5 by calculating the changes in PM2.5_MAIAC
(ΔPM2.5_RH) if EXTsonde is used instead of
EXTCMAQ. As expected, model errors in RH lead to a negative bias in
derived PM2.5_MAIAC of 3 µg m-3 on average
(Fig. 9c). The negative biases in PM2.5_MAIAC due to RH
are largest in spring (-3.5 µg m-3) and smallest in summer
(-1.6 µg m-3). The hygroscopic growth factor is nonlinearly correlated
with RH, which increases more rapidly at high RH (> 80 %) than
at low to median RH (< 80 %, Fig. S2). Compared with median RH
conditions, model RH errors lead to more than double ΔPM2.5_RH (-6.4 µg m-3 vs. 3 µg m-3) when observed near-surface RH > 80 % (Fig. 9d). At
RH > 95 %, we find that the ΔPM2.5_RH can be as large as -20 µg m-3 (Fig. 9d). Despite the large
impacts of model errors of RH at humid conditions, there is no significant
correlation between ΔPM2.5_RH and ΔPM2.5_MAIAC (R=0.18, evaluated at nearby sites
within 10 km), suggesting that uncertainty in RH is not a main contributor
to the random uncertainties in satellite-derived PM2.5.
Uncertainties in the parameterization of aerosol optical
properties
In previous sections, we demonstrated that the satellite-derived PM2.5
depends on the accuracy of the model simulation. Even with a perfect
simulation, satellite-derived PM2.5 will be sensitive to the
parameterization of aerosol optical properties, which would affect the
MEE. We evaluate the uncertainties associated with
the parameterization of aerosol optical properties by varying each parameter
(Table 1), and we calculate the corresponding changes in the derived
PM2.5_MAIAC. Figure 10 shows the range of uncertainty in
annual average PM2.5_MAIAC due to uncertain aerosol size
distributions, hygroscopicity, refractive index, and aerosol species density.
Uncertainties in annual average satellite-derived
PM2.5_MAIAC due to uncertainties of size distribution,
hygroscopicity, refractive index, and aerosol species density of
sulfate–nitrate–ammonium (SNA; blue) and organic carbon (OC; green) sampled
over AQS sites. The circle shows the annual average satellite-derived
PM2.5_MAIAC using the default parameters to calculate
AODCMAQ in FlexAOD (Table 1). The error bars represent the range
of PM2.5_MAIAC using different values for each parameter. The
labels indicate the corresponding minimum or maximum parameter values that
produce the range shown in PM2.5_MAIAC. The horizontal line at
15 µg m-3 indicates the annual average
PM2.5_MAIAC calculated using default values for each aerosol
optical property in the base FlexAOD.
The size of a particle is a defining characteristic of aerosol light
extinction (Mishchenko et al., 1999). To evaluate model sensitivities to the
uncertainties in size distribution, we vary the r0 of SNA from 0.05 to
0.15 with a 0.02 increase each time to cover the range of values reported
in the literature. For OC, we calculate AODCMAQ with r0=0.02,
0.06, 0.09, and 0.12 µm, all values used in previous studies
(Hess et al., 1998; Chin et al., 2002;
Highwood, 2009; Drury et al., 2010). Annual average PM2.5_MAIAC could vary by up to 5 µg m-3 (32 %) with the choice
of a
modal radius of either rSNA or rOC, which is the largest source of
uncertainty among the four parameters (Fig. 10). We find that AODCMAQ
reaches a maximum with rSNA=0.07 µm (reff=0.12 µm) and minimum with rSNA=0.05 (reff=0.15 µm),
while PM2.5_MAIAC reaches a maximum with
rSNA=0.05 (reff=0.09 µm) and minimum with
rSNA=0.11 (reff=0.19 µm), suggesting the impacts
of size distribution are nonlinear and nonuniform (Fig. S3). Mie scattering
of a particle tends to be most effective when the particle's diameter is
near the wavelength of interest (0.55 µm). As hygroscopic particle
growth also affects the size distribution, depending on ambient RH and the
hygroscopic growth factor, reducing (or increasing) the dry effective radius
could move the bulk aerosol size either closer to or further from 0.55 µm and thus either increase or decrease the extinction. For OC, as
the effective radius and the hygroscopic growth factor are smaller than for
SNA, increasing the modal radius leads to more effective scattering, and thus
larger AODCMAQ and smaller PM2.5_MAIAC. Relative to
the default rOC=0.09 µm assumed by Drury et al. (2010), using the rOC (0.02 µm) recommended by Chin et
al. (2002a) increases PM2.5_MAIAC by 5 µg m-3
(32 %) on average, worsening the positive biases of
PM2.5_MAIAC. Increasing rOC to 0.12 µm as
recommended by Highwood et al. (2009) has little effect, decreasing
PM2.5_MAIAC by 2 % on average.
The uncertainty of hygroscopicity lies in two aspects: (1) the function
shape and (2) the parameters. Figure S2 compares the κ function
shape with the hygroscopic growth factors used by the IMPROVE network (Hand
and Malm, 2006), the default algorithm used to calculate AOD online in CMAQ,
with that proposed by Chin et al. (2002) (Table 1). Using the
DISCOVER-AQ aircraft data to evaluate the parameterization of hygroscopic
growth, we find that the κ parameter best characterizes the observed
hygroscopic growth factor (Fig. S2c). Latimer and Martin (2018) similarly
found that implementing a κ formulation instead of hygroscopic
growth based on OPAC improved the GEOS-Chem representation of mass
scattering efficiency. Thus, we choose the κ parameter to represent
the hygroscopic growth factor, and the uncertainty estimate here only
reflects uncertainties in the κ parameter. In practice, changes in
aerosol composition could have even larger effects on hygroscopicity than
uncertainties in κ as discussed in Sect. 3.5.1.
To test the sensitivity of satellite-derived PM2.5 to uncertainties in
the κ parameter, we compute AODCMAQ using the low (0.33) and
high end of κ (0.72) for SNA as suggested by Koehler et al. (2006). As the hygroscopic properties of inorganic salts are relatively
well known, the range of uncertainty for f(RH) of SNA is 30 % at most (Fig. S2b). OC, however, is composed of thousands of species with
distinct hygroscopicities. Assuming κOC ranges from 0
(non-hygroscopic) to 0.2 (Jimenez et al., 2009; Duplissy et al., 2011), the
range of f(RH) of OC can be as large as a factor of 2 at high RH > 96 % (Fig. S2a). Despite the larger uncertainty of κOC, we
find the overall impacts of the uncertainties of κOC on the
derived PM2.5 (0.3 µg m-3, 2 % of annual average
PM2.5_MAIAC) are smaller than those of κSNA
(1.6 µg m-3, 11 % of annual average PM2.5_MAIAC). The small impacts of κOC reflect the relatively small
portion and the less hygroscopic nature of OC. For single observations,
varying κSNA leads to a maximum increase in
PM2.5_MAIAC by 20 % and a maximum decrease by
28 %. Varying κOC increases PM2.5_MAIAC by 10 % or decreases PM2.5_MAIAC by 18 %
at most. The overall impact of the uncertainties of κSNA ranks second among the four parameters for SNA, while κOC has
the smallest impacts on the derived PM2.5 (Fig. 10).
The refractive index (m) determines the Mie extinction efficiency, which is
subject to uncertainties mostly due to the lack of measurements
(Kanakidou et al., 2005). mSNA in OPAC (default value) is slightly
different from that recommended in Chin et al. (2002) and Highwood
(2009). Moise et al. (2015) suggest mOC varies by species, with its
real part ranging from 1.37 to 1.65. We calculated another version of
AODCMAQ by varying the real part of mSNA and mOC using the
lowest and highest values reported in the literature. We find the annual
average PM2.5_MAIAC decreased by 0.8 µg m-3
(6 %) using the high end of mR_SNA, while it increased by
1.3 µg m-3 (9 %) on average using the low end. Though
mR_OC has a wider range of uncertainty, its
impacts on PM2.5_MAIAC (-4 % to +6 %) are smaller
than that of mR_SNA. While the overall impacts on
PM2.5_MAIAC due to uncertainties of
mR_SNA are generally within 10 % for single
observations, PM2.5_MAIAC can change by more than
20 % under SNA-dominated and high-RH environments. The overall uncertainty
due to mR_OC is generally within 5 % for single
observations, with a few cases (< 10 % of the total data) in
which
the relative change in PM2.5_MAIAC can exceed 10 %.
As aerosol density (ρ) is assumed to be constant for each species, varying
ρ has the same effect on the extinction of given species. We vary the
aerosol density of SNA from 1.65 to 1.83 g cm-3 based on the uncertainty
estimate from a laboratory study of Sarangi et al. (2016), which
translates to an uncertainty of -3 % to 7 % for AODSNA, and the
aerosol density of OC from 1.2 to 1.78 g cm-3 following Park et al. (2006), which translates to an uncertainty in AODOC ranging from -8 %
to 37 %. We find aerosol species density, in general, contributes least to
the overall uncertainty in satellite-derived PM2.5. Varying ρOC
across the range in Table 1 increases annual average PM2.5_MAIAC by 0.9 µg m-3 (6 %) or decreases it by 0.6 µg m-3 (3 %) at most. As the aerosol density of inorganic salt is less
uncertain, varying ρsulf leads to negligible changes in annual average
PM2.5_MAIAC at both the high (0.7 µg m-3, 5 %)
and low (-0.5 µg m-3, -2 %) ends.