ESSDEarth System Science DataESSDEarth Syst. Sci. Data1866-3516Copernicus PublicationsGöttingen, Germany10.5194/essd-10-2069-2018The Berkeley High Resolution Tropospheric NO2 productThe BEHR NO2 productLaughnerJoshua L.https://orcid.org/0000-0002-8599-4555ZhuQindanhttps://orcid.org/0000-0003-2173-4014CohenRonald C.rccohen@berkeley.eduhttps://orcid.org/0000-0001-6617-7691Department of Chemistry, University of California, Berkeley, Berkeley, CA 94720, USADepartment of Earth and Planetary Sciences, University of California, Berkeley, Berkeley, CA 94720, USAnow at: Department of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125, USARonald C. Cohen (rccohen@berkeley.edu)27November20181042069209516May201825June20189October201830October2018This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this licence, visit https://creativecommons.org/licenses/by/4.0/This article is available from https://essd.copernicus.org/articles/10/2069/2018/essd-10-2069-2018.htmlThe full text article is available as a PDF file from https://essd.copernicus.org/articles/10/2069/2018/essd-10-2069-2018.pdf
We describe upgrades to the Berkeley High Resolution (BEHR) NO2
satellite retrieval product. BEHR v3.0B builds on the NASA version 3 standard
Ozone Monitoring Instrument (OMI) tropospheric NO2 product to provide
a high spatial resolution product for a domain covering the continental
United States and lower Canada that is consistent with daily variations in
the 12 km a priori NO2 profiles. Other improvements to the BEHR v3.0
product include surface reflectance and elevation, and factors affecting the
NO2 a priori profiles such as lightning and anthropogenic emissions.
We describe the retrieval algorithm in detail and evaluate the impact of
changes to the algorithm between v2.1C and v3.0B on the retrieved NO2
vertical column densities (VCDs). Not surprisingly, we find that, on average,
the changes to the a priori NO2 profiles and the update to the new
NASA slant column densities have the greatest impact on the retrieved VCDs.
More significantly, we find that using daily a priori profiles results in
greater average VCDs than using monthly profiles in regions and times with
significant lightning activity.
The BEHR product is available as four subproducts on the University of
California DASH repository, using monthly a priori profiles at native OMI
pixel resolution (https://doi.org/10.6078/D1N086) and regridded to
0.05∘×0.05∘ (https://doi.org/10.6078/D1RQ3G)
and using daily a priori profiles at native OMI
(https://doi.org/10.6078/D1WH41) and regridded
(https://doi.org/10.6078/D12D5X) resolutions. The subproducts using
monthly profiles are currently available from January 2005 to July 2017, and
will be expanded to more recent years. The subproducts using daily profiles
are currently available for years 2005–2010 and 2012–2014; 2011 and 2015 on
will be added as the necessary input data are simulated for those years.
Introduction
Nitrogen oxides
(NO+NO2≡NOx) are trace gases
in the atmosphere and are key species controlling air quality and affecting
radiative balance. NOx regulates the chemical production of
tropospheric ozone , which affects the radiative balance in
the upper troposphere and is harmful to plants
, animals, and humans at the
surface. It also plays a role in the formation of aerosol particles
, which also affect the
radiative balance of the atmosphere . Exposure to fine
particles is also a strong factor controlling life expectancy
. Additionally, NOx
itself is harmful, as, for example, exposure causes bronchoconstriction and
associated difficulty breathing, especially for those affected by asthma
.
NOx is emitted from a variety of sources, both anthropogenic
and natural. Anthropogenic sources typically involve combustion, including
motor vehicles and fossil fuel electrical generation. Natural sources include
biomass burning, lightning, and soil bacteria. Understanding all of these
sources is crucial to understanding the reactive nitrogen budget and
predicting how future changes in emissions will affect air quality and
climate change.
Satellite observations provide uniquely comprehensive spatial maps of
NO2, allowing inference of NOx emissions. The
spatial resolution available with early instruments (i.e., the Global Ozone
Monitoring Experiment, GOME, 40×320 km2, ;
the SCanning Imaging Absorption SpectroMeter for Atmospheric CHartographY,
SCIAMACHY, 30×60 km2, ) allowed inferences at
the scale of entire continents or entire metropolitan regions, including
cities and their surroundings. More recent instruments have much higher
resolution (e.g., the Ozone Monitoring Instrument, OMI, 13×24 km2, ; the Tropospheric Monitoring Instrument,
TROPOMI, 7×7 km2, ), allowing inferences
about individual point sources and urban cores. Ground based measurements
sample emissions at specific points in great detail; however, extrapolating
such measurements to an entire region requires assumptions that are difficult
to test, such as fleet composition and operating mode
e.g.,, that can bias estimates of the total
vehicle emissions from a region. Satellite observations cannot currently
provide the same level of detail as a roadside measurement, but by observing
the entire city, provide a top–down constraint on its total
NOx emissions that include observations on every point in
the domain. Satellite observations have been used in a wide variety of
applications in this vein, including direct observation of emissions and
trends e.g.,, plume analysis to derive emissions and
chemical lifetime e.g.,,
model constraint e.g.,, and data assimilation
e.g.,.
Satellite measurements have been used to constrain natural
NOx sources as well, predominantly biomass burning
e.g.,, lightning e.g.,, and soil
NOxe.g.,. The episodic and geographically disparate nature of these sources
(especially lightning and biomass burning) make satellite observations an
ideal method to constrain their emissions, given satellites' continuous data
record and broad geographic coverage.
The current fleet of space-based sensors measures NO2, not total
NOx, but due to the rapid daytime equilibrium between
NO and NO2, this allows inferences about tropospheric
NOx to be made from NO2 measurements. For a
measurement of tropospheric NO2, several steps are required. First, a
UV-visible spectrometer records geolocated solar reflectances from the
Earth's surface and a reference spectrum of the sun. Then, absorbances in
backscattered sunlight are fit using differential optical absorption
spectroscopy (DOAS) or a similar technique to yield a total slant column
density (SCD). This quantity represents the amount of NO2 per unit
area, integrated along all light paths that reach the detector
. Next, the tropospheric and
stratospheric NO2 columns are separated. There are several
approaches; some examples include using a data assimilation system to
constrain modeled stratospheric columns and an iterative
process assuming that areas known a priori to have little tropospheric
NO2 are all stratospheric NO2 and interpolating to fill in
polluted areas . Finally, the tropospheric SCD is converted
into a vertical column density (VCD) in order to account for pixel-to-pixel
differences in path length and sensitivity to NO2. The conversion
factor from the SCD to the more geophysically relevant and easily understood
VCD is the air mass factor AMF,.
An AMF is computed by simulating an SCD and VCD for each retrieved pixel.
Typically, an a priori NO2 profile is simulated with a chemical
transport model (CTM) such as GEOS-Chem, WRF-Chem, the GMI-CTM, TM4, or TM5.
The modeled VCD can be calculated by integrating this profile over the
troposphere. The modeled SCD requires a radiative transfer model, such as
TOMRAD, SCIATRAN, or VLIDORT, in combination with the a priori NO2
profile in order to compute the light absorbed by NO2 and thus the
SCD that yields that absorbance. The radiative transfer calculations also
require a priori inputs: the sun-satellite geometry, surface reflectance, and
surface elevation are all necessary. Knowledge of the cloud and aerosol
properties in the pixel is also necessary to account for their effects on
light scattering in the radiative transfer calculations. Aerosol effects are
often assumed to be implicitly accounted for in cloud properties
e.g.,, but have been treated explicitly by some products
e.g.,.
The accuracy of these input data has a significant impact on the accuracy of
the AMFs and therefore the vertical columns. compared seven
retrievals and found that input assumptions were responsible for a 42 %
structural uncertainty in AMFs over polluted areas. A key concern is the
resolution of the input data. CTMs are computationally expensive, requiring a
trade-off between spatial and temporal resolution and domain size. For global
products, model resolutions of 3∘×2∘ to 1∘×1∘ are
typical. found that increasing the resolution of the
NO2 profiles from 2.5∘×2∘ to 4 km altered
the retrieved VCDs by up to 75 %, primarily by capturing the urban–rural
gradient in surface NO2 concentrations. found that
increasing the a priori profiles' resolution from 3∘×2∘ to 15 km resulted in a factor of 2 increase in NO2
column over the Canadian oil sands. examined the effect of
the profiles' temporal resolution, and identified up to 40 % changes in
individual VCDs using day-to-day NO2 profiles compared to monthly
averaged profiles. The current trade-off to obtain such high-resolution
profiles is that the resulting product is only available over a subset of the
world, rather than globally.
The Berkeley High Resolution (BEHR) Ozone Monitoring Instrument (OMI)
NO2 retrieval is one such regional product that provides tropospheric
NO2 VCDs over part of North America (approximately 125 to
65∘ W, 25 to 50∘ N) using high-resolution a priori inputs.
The BEHR product has been used in numerous studies covering areas of research
such as NOx trends , anthropogenic emissions , soil emissions
, land use regression modeling , and model
evaluation .
Here we describe the updates from v2.1C to v3.0B. (For information on v2.1C,
see , and the changelog at
http://behr.cchem.berkeley.edu/Portals/2/Changelog.txt, last access:
14 November 2018.) There are seven primary changes.
Updated to use the v3.0 NASA tropospheric SCDs
Surface reflectance updated from version 5 MODIS black sky albedo to version 6 MODIS BRF product
New a priori NO2 profiles, with specific changes:
Lightning NO2 included
Monthly profiles use 2012 emissions, instead of 2005 emissions used in v2.1C and prior
Daily profiles, with year-specific emissions, used for as many years as possible
Temperature profiles taken from WRF-Chem instead of the previous coarse climatology
A new gridding method was implemented that corrected issues with grid cells on the border between two pixels not being allocated a value
A variable tropopause height derived from WRF simulations replaced the previous fixed 200 hPa tropopause in the AMF calculations.
Surface pressure calculation was changed to follow using GLOBE terrain elevation and WRF surface pressure
These changes all affect the tropospheric VCDs. BEHR also provides a
“visible-only” VCD, that is, the VCD excluding NO2 below clouds for
users interested in, e.g., cloud slicing methods . These
visible-only VCDs are computed by dividing the tropospheric slant columns by
the corresponding visible-only AMF. BEHR v3.0A implemented a more physically
intuitive form of the visible-only AMF than that in v2.1C. This change is
described in the Supplement for interested users.
In this paper, we describe each change in detail and examine the effect of
each individual change on the calculated VCDs. v3.0A was available on the
BEHR website (behr.cchem.berkeley.edu, last access: 14 November 2018)
between November 2017 and July 2018; v3.0B replaced v3.0A on the website and
the static repositories in July 2018. Therefore, in this
paper, we will separate changes implemented in v3.0A from those in v3.0B, so
that the differences between v3.0A and v3.0B can be accounted for if any
results are published using v3.0A. Changes implemented in v3.0A are described
first, followed by those implemented in v3.0B. Validation of v3.0B is
described separately in .
Because of the computational resources required to simulate daily a priori
NO2 profiles, BEHR v3.0B is produced for all years from 2005 on using
monthly average NO2 profiles, and for as many years as possible with
daily NO2 profiles. The latter is available for 2005–2010 and
2012–2014, with the remaining years following as the simulations of the
necessary NO2 profiles are completed. In this paper, we focus on the
2012 data as an example to understand the effect each change to the algorithm
has on the final VCDs.
Methods: BEHR
Unless otherwise noted, the following methods description applies to both
BEHR versions 3.0A and 3.0B. A summary of the differences in methods between
v3.0A and v3.0B is listed in Table .
NO2 VCD calculation
The BEHR product calculates tropospheric vertical column densities (VCDs)
starting from the tropospheric slant column densities (SCDs) from the NASA
Standard Product, v3.0 , by
VBEHR=SNASAABEHR,
where VBEHR and SNASA are the BEHR VCD and NASA
SCD, respectively, and ABEHR is a custom tropospheric air mass
factor (AMF), computed with
ABEHR=(1-f)∫psurfptropwclear(p)g(p)dp+f∫pcloudptropwcloudy(p)g(p)dp∫psurfptropg(p)dp
where f is the cloud radiance fraction, and wclear and
wcloudy are the scattering weights for clear and cloudy
subscenes (i.e., parts of the pixel), respectively, ptrop is
the tropopause pressure, psurf is the ground surface pressure,
pcloud is the cloud optical centroid pressure, and g(p) is
the NO2 a priori profile in mixing ratio
(Sect. ). The calculation of both
psurf and ptrop differ between v3.0A and v3.0B;
see Sects. and ,
respectively.
This method produces VCDs that include an estimated below-cloud component,
and thus can be considered a total tropospheric column. This is desirable for
applications focusing on near-surface NO2, and are stored in the BEHR
data as “BEHRColumnAmountNO2Trop”. Other applications (e.g., cloud slicing)
benefit from having a “visible-only” tropospheric AMF that only retrieves
NO2 above the cloud in a cloudy subscene. For these “visible-only”
AMFs, Eq. () is replaced with
ABEHR,vis=(1-f)∫psurfptropwclear(p)g(p)dp+f∫pcloudptropwcloudy(p)g(p)dp(1-fg)∫psurfptropg(p)dp+fg∫pcloudptropg(p)dp,
where fg is the geometric cloud fraction. The numerator is the
same as in Eq. (), in both cases representing a modeled
slant column density. The denominator in Eq. () is the
total modeled tropospheric column, while in Eq. () it
is only the visible modeled column. Replacing ABEHR in
Eq. () with ABEHR,vis yields a
visible-only NO2 column as the output, stored in the variable
“BEHRColumnAmountNO2TropVisOnly” in the BEHR files. The form of this
visible AMF changed from v2.1C to v3.0A; please see Sect. S1 in the
Supplement for details of the old calculation.
The scattering weights (wclear and wcloudy) are
computed from the same look-up table (LUT) as the NASA SP v2.1 and v3.0
. The scattering weights depend on the solar
zenith angle (SZA, θS), viewing zenith angle (VZA, θV),
relative azimuth angle (RAA, ϕR), surface reflectance
(Sect. ), and surface pressure
(Sect. ). A vector of scattering weights is looked
up using 5-D multilinear interpolation to obtain the scattering weights for
the above input parameters. Note that the RAA is calculated as
ϕR,tmp=|180+ϕS-ϕV|,ϕR=ϕR,tmpifϕR,tmp∈[0,180],360-ϕR,tmpifϕR,tmp>180,
where ϕS and ϕV are the solar and viewing
azimuth angles, respectively, defined in degrees, and ϕR,tmp
is a temporary variable. The extra factor of 180 in Eq. ()
accounts for the RAA definition used in the scattering weight look-up table
(where ϕR=0 indicates that the satellite is opposite the
sun, i.e., in the forward scattering position), while
Eq. () ensures that ϕR is between 0 and
180∘.
A temperature correction, α(p), is
applied to the scattering weights interpolated from the look-up table, such
that w(p) in Eqs. () and () is equal to
α(p)w0(p), where w0(p) is the pressure-dependent scattering
weights from the look-up table and α(p) is
α(p)=1-0.003⋅(T(p)-220),α(p)∈[0.1,10],
where Eq. () indicates that α(p) is constrained
to the range 0.1 to 10. T(p) is a temperature profile taken from the same
WRF-Chem simulation as the NO2 a priori profiles
(Sect. ).
Surface reflectivityOver land
BEHR v3.0 uses a bidirectional reflectance factor (BRF) to represent surface
reflectivity over land. The BRF is given by as
R(θS,θV,ϕR,Λ)=fiso(Λ)+fvol(Λ)Kvol(θS,θV,ϕR)+fgeo(Λ)Kgeo(θS,θV,ϕR),
where R is the surface reflectivity, fiso, fvol,
and fgeo are coefficients representing the relative contributions
of different types of scattering, and Kvol and Kgeo
are kernels representing the directional dependence of the reflectivity.
Λ represents a wavelength band, which here is band 3 of the MODIS
instrument (459–479 nm).
Kvol is the RossThick kernel and
Kgeo is the LiSparse kernel , corrected to be
reciprocal in θS and θV. BEHR calculates
both kernels using the formulations given in . The
coefficients, fiso, fvol, and fgeo, are
taken at 30 arcsec resolution from the MODIS MCD43D07 ,
MCD43D08 , and MCD43D09 BRF products,
respectively. Quality information for these coefficients is obtained from the
MCD43D31 product . (The combination of these four products
will henceforth be referred to as MCD43Dxx.) These products represent a
16-day average; in version 006 (used here), the file date is in the middle of
that 16-day averaging window. BEHR uses the file dated for the day being
retrieved for the BRF coefficients; i.e., for 1 June 2012, the MODIS files
with 1 June 2012 in the file name are used. This means that the surface
reflectivity used in BEHR incorporates land data from 8 days before and after
the OMI observation.
An average surface reflectance for a given OMI pixel is calculated by
computing R for each set of MCD43Dxx coefficients within the bounds of the
pixel given by the FoV75 corners from the OMPIXCOR product
and using the SZA, VZA, and RAA of the pixel as inputs to the kernels. All
values of R from MCD43Dxx coefficients with non-fill quality flags are
averaged to produce the overall surface reflectance for the pixel; however,
since coefficients with quality 3 are significantly lower quality than
quality 0 to 2, if the average quality of all MCD43Dxx coefficients within
the OMI pixel is ≥2.5, the pixel is flagged as low quality. The pixel
is also flagged if ≥50% of the MCD43Dxx coefficients have a fill
value for the quality (see Sect. ).
Over water
The MCD43Dxx products do not contain coefficients over deep water; therefore,
an alternate measure of surface reflectance is needed. We use the University
of Maryland land map
(ftp://rsftp.eeos.umb.edu/data02/Gapfilled/Land_Water_Mask_7Classes_UMD.hdf, last
access: 28 November 2017) to classify OMI pixels as land or water. Land classes 0
(shallow ocean), 6 (moderate or continental ocean), and 7 (deep ocean) are
considered ocean; all others are considered land. The mask is given at 30 arcsec resolution; if >50% of the mask data points within the FoV75
bounds of the OMI pixel are ocean, the pixel is treated with an ocean surface
reflectance.
Ocean surface reflectance is parameterized by SZA using output from the
Coupled Ocean Atmosphere Radiative Transfer (COART) model hosted at
https://satcorps.larc.nasa.gov/jin/coart.html, last access: 2 March
2018. The ratio of upwelling to downwelling radiation was simulated
for 18 solar zenith angles (0 to 85∘ at 5∘ increments).
Additional settings are given in Table . The ratio of
upwelling to downwelling radiation is linearly interpolated to the SZA of the
OMI pixel, and that interpolated ratio is taken as the surface reflectance of
the ocean pixel. In v3.0A, COART-simulated reflectance at 430 nm was used;
in v3.0B, reflectance at 460 nm was used.
Surface pressure
The surface elevation of each OMI pixel is computed by averaging all surface
elevation values from the Global Land One-kilometer Base Elevation (GLOBE)
database within the FoV75 bounds of the pixel. From v3.0B on,
pixel surface pressure is calculated using the method recommended by
:
p=pWRFTWRFTWRF+Γ⋅(hWRF-hGLOBE)-g/RΓ,
where p is the pixel surface pressure, pWRF,
TWRF, and hWRF are the surface pressure,
temperature, and elevation from the WRF model, hGLOBE is the
averaged GLOBE surface elevation, g is gravitational acceleration (9.8 m s-2), R is the gas constant for dry air (287 J kg-1 K-1) and
Γ the lapse rate (0.0065 K m-1).
Prior to v3.0B, the surface pressure was computed by converting the average
GLOBE surface elevation to a pressure using a fixed scale height calculation:
p=(1013.25hPa)e-z/7400m,
where z is the average surface elevation in meters.
Additional settings for the COART model used to simulate ocean
reflectivity. “Atmospheric profile” refers to the distribution of total
precipitable water, O3, CO2, and
CH4.
For the upper integration limit in Eq. (), BEHR v3.0A and
prior versions used a fixed tropopause pressure (200 hPa). BEHR v3.0B
utilizes a thermal tropopause pressure derived from temperature profiles from
the same WRF-Chem simulation as the NO2 a priori profiles. The
thermal tropopause is defined as the lowest level at which the average lapse
rate between this level and all higher levels within 2 km does not exceed 2
K km-1 by World Meteorological Organization (1957). The calculation
operationally works in most regions; however, occasionally a discontinuity
occurs between adjacent pixels where both pixels approach the 2 K km-1
threshold at the same model level but only one exceeds the threshold at that
level. As this discontinuity is only due to the choice of the standard
threshold for lapse rate in the criteria, an additional filtering is
implemented to identify pixels with abrupt transition in calculated
tropopause pressure. New tropopause pressures for these pixels are derived by
linear interpolation of tropopause pressures from the nearest valid pixels
after filtering.
Summary of differences in methods between v3.0A and
v3.0B.
Componentv3.0Av3.0BSectionOcean reflectanceCalc. for 430 nmCalc. for 460 nmSurface pressureScale heightWRF pressure adjusted with GLOBE elevationTropopause pressureFixed at 200 hPaCalculated from WRF temperature profilesDaily prof. hourLast hour before overpassClosest hour to overpassCloud products
BEHR contains several cloud fraction products: a geometric cloud fraction
derived from the O2–O2 algorithm , a cloud
radiance fraction calculated by NASA from the O2–O2 product,
and a geometric cloud fraction derived from the Aqua MODIS instrument (which
currently makes observations ∼8 min before OMI). Additionally, cloud
pressure from the OMI O2–O2 algorithm is
included. The OMI-derived quantities are the same as those in the NASA SP
v3.0. The MODIS cloud product used is MYD06_L2 .
found that the MODIS cloud product was less likely to give
erroneously large cloud fractions due to high surface reflectivity over the
California and Nevada desert, and concluded that this more than offset any
error caused by the small separation between the overpass times (currently
∼8 min) of OMI onboard the Aura satellite and MODIS onboard the Aqua
satellite. We continue to provide the MODIS cloud product for cloud
filtering; however, because it does not cover the full OMI swath, we use the
OMI cloud fractions in the AMF calculations.
As with the MODIS BRF product, all values of cloud fraction given in
MYD06_L2 within each OMI pixel's
bounds defined by the FoV75 pixel corners are averaged to yield the
MODIS-derived cloud fraction for that OMI pixel. Unlike the BRF product, only
Level 2 MODIS granules with times between the start and end times of the
current OMI orbit are used.
A priori profiles
From v3.0A onward, BEHR is divided into two subproducts which differ in the
temporal resolution of the a priori NO2 profiles. Based on the
results in , using a priori profiles specifically simulated
for each day of BEHR observations is preferable; however, the computational
cost of doing so limits the time periods that such profiles can be simulated
for. Therefore a second subproduct using monthly average profiles derived
from the 2012 a priori profiles is available that covers all years of the OMI
data record. This assumes that monthly average profiles are applicable to
years other than that for which they were simulated; while not a perfect
assumption, it has successfully been used in previous NO2 products
e.g.,.
In this section, we describe the model configuration used to generate the a
priori profiles. General model settings will be described first, followed by
information specific to the implementation of daily and monthly average
profiles in the BEHR algorithm.
WRF-Chem configuration
NO2 and temperature a priori profiles are generated using version
3.5.1 of WRF-Chem run at 12 km resolution across the
continental United States (Fig. S6 in the Supplement). The North American
Regional Reanalysis (NARR) dataset is used to drive the meteorological
initial and boundary conditions, as well as four-dimensional data analysis (FDDA) nudging . U and V winds,
temperature, and water vapor are nudged at all levels with nudging
coefficients of 0.0003 s-1.
Anthropogenic emissions are driven by the National Emissions Inventory 2011
(NEI 11) gridded to 12 km resolution. Each year's emissions are scaled by
the ratio of that year's total annual emissions to 2011 emission. These total
emissions are provided by the Environmental Protection Agency
. Biogenic emissions are driven by the Model of
Emissions of Gases and Aerosols from Nature MEGAN,.
Lightning emissions are driven by the recommended settings in
for a simulation using FDDA nudging.
Chemistry in WRF-Chem is simulated using the RACM2_Berkeley2 mechanism
, which is based on the Regional Atmospheric Chemistry Mechanism,
version 2 RACM2, with updates to alkyl nitrate and
nighttime chemistry and the inclusion of
methylperoxy nitrate (MPN) chemistry .
Chemical boundary conditions for WRF-Chem are taken from two different global
models. For model years 2007 and later, chemical concentrations from the
Model for Ozone and Related chemical Tracers MOZART,
provided by the National Center for Atmospheric Research (NCAR) at
https://www.acom.ucar.edu/wrf-chem/mozart.shtml (last access:
14 November 2018) are used, converted to boundary conditions for WRF-Chem
using the MOZBC utility. MOZART data are not available from NCAR for years
prior to 2007; instead, the chemical data are taken from GEOS-Chem model
v9-02 (at 2.5∘×2.0∘ resolution), with updates from
. These updates are detailed in Sect. S3. GEOS-Chem
instantaneous output is sampled every 3 h. This output is transformed into
netCDF files for input into the MOZBC utility by use of the gc2moz utility of
the AutoWRFChem package .
Each year is simulated with a 1-month spinup at the anthropogenic emissions
levels for that year. The year is simulated continuously, without
reinitialization. Instantaneous WRF-Chem output is sampled hourly. For 2007,
since MOZBC data were not available for December 2006, boundary conditions
for 1 January 2007 were repeated for the first 32 days of the simulation
(1 December 2006 to 1 January 2007) to allow the model time to spin up from
the initial conditions.
In the BEHR AMF calculation, the profiles are interpolated to the same
pressures that the scattering weights are defined on. The NO2 mixing
ratio profiles are interpolated in log–log space e.g.,
ln(NO2) given at ln(pWRF) is interpolated to
ln(pBEHR),. Temperature is interpolated in
semilog space (T given at ln(pWRF) is interpolated to
ln(pBEHR), since lapse rates assume a linear relationship
between temperature and altitude, and altitude is proportional to ln(p).
The profiles are also extrapolated to one scattering weight pressure level
above and below the top and bottom of the WRF profile, respectively. This
accounts for the possibility that, e.g., a pixel's surface pressure may be
slightly below the WRF surface pressure, but limiting the extrapolation to
only one level should minimize errors due to extrapolation. Once interpolated
and extrapolated, all profiles within the FoV75 bounds of the OMI pixel are
averaged to give the profiles used in calculating the AMF.
Daily a priori profiles
We make use of daily profiles for as much of the OMI data record as it is
computationally feasible to simulate these profiles. Both NO2 a
priori profiles and the temperature profiles necessary for the scattering
weight temperature correction are drawn from the same simulation. WRF-Chem is
configured to provide instantaneous output at the top of every hour. In
v3.0A, the last WRF-Chem profile before the average time of the OMI pixels
over the domain is chosen to provide the a priori NO2 and temperature
profiles. In v3.0B, the profile closest in time to the average OMI time is
used. These profiles are binned to OMI pixels as described in
Sect. .
As of this writing, daily profiles have been simulated for 2005 to 2010 and
2012 to 2014. Profiles for 2011 are in progress, and profiles for 2015 and
later years will be simulated as time and computational resources permit.
Monthly a priori profiles
Given the computational cost in producing daily a priori profiles, we
continue to use monthly average profiles as well to cover years for which
daily a priori profiles have not yet been simulated. Monthly profiles are
generated from 2012 WRF-Chem output. As in , an average of
all available hourly profiles for a given month weighted by weights wl is
given by
wl=1-13.5-(l/15)-h,wl∈[0,1],
where l is the profile longitude and h is the UTC hour of the profile.
This formulation gives highest weight to profiles near OMI overpass time
(approximately 13:30 local standard time) while smoothly interpolating
between adjoining time zones. The appropriate month's profiles are spatially
matched to OMI pixels in the same manner as the daily profiles
(Sect. ).
Paper structure
In Sects. and , we evaluate the
effect each change to the BEHR algorithm between v2.1C and v3.0B had on the
tropospheric VCDs. In order to provide a clear history, changes introduced in
v3.0A will be discussed first (Sect. ), followed by
changes introduced in v3.0B (Sect. ). V3.0A incorporated
all changes up through the introduction of the new gridding algorithm; the
remainder are added in v3.0B. Changes to the visible-only VCDs (i.e., those
excluding the below-cloud column) are discussed in the Supplement (Sect. S1).
Following this the overall difference between v2.1C and v3.0B will be
presented in Sect. . Recommendations for the use of
the product are given in Sect. . A description
of the data format is given in Appendix .
For the discussion of how changes to the algorithm affect the NO2
VCDs, Figs. and and Tables
and are the central focus.
Each panel shows the change in the BEHR NO2 VCDs resulting from a
specific change to the algorithm. To generate these figures, BEHR VCDs were
computed after adding each change to the algorithm incrementally. Each panel
in the figures and line in the tables shows the percent change in VCDs due to
the corresponding change to the algorithm. These are computed relative to
VCDs with one fewer change to the algorithm; for example,
Fig. b is the percent difference between VCDs using the
new NASA SCDs and the new MODIS BRF surface reflectance versus VCDs using
just the new NASA SCDs. Figures a and
a and the first lines in
Tables and are relative to
BEHR v2.1C.
Figure shows the percent change of average BEHR
tropospheric VCDs due to each algorithm improvement for the subproduct using
monthly average NO2 a priori profiles, while
Fig. shows the changes to the subproduct using
daily NO2 a priori profiles. (Figure has
fewer panels than Fig. as daily profiles were only
possible in increments after the change to the algorithm to introduce the new
a priori profiles was implemented.) Both figures are for summer
(June–August) 2012. Winter changes are presented in the Supplement.
Table gives the mean and median changes for each
incremental improvement shown in Figs.
and ; that is, it gives the domain-wide mean and
median values of the time-averaged changes shown in the figures.
Table is similar, but is the statistics for
individual pixels, rather than the time-averaged changes.
Percent change in the tropospheric NO2 column due to each of
the algorithm improvements. Changes due to (a) new NASA SCDs, (b) new surface
reflectance, (c) new monthly NO2 profiles, (d) new temperature
profiles, (e) new gridding method, (f) change in ocean reflectance LUT from
430 to 460 nm, (g) switch to WRF-derived tropopause pressure, (h) switch to
surface pressure methodology. Note that the color scale varies
among the plots. Averages are for June–August 2012 and exclude pixels affected
by the row anomaly and with cloud fraction >0.2. Monthly average a priori
profiles are used for all differences. Wintertime changes and histograms are
given in Sect. S5.
Changes in BEHR v3.0ANASA v3.0 slant columns
Version 3.0 of the NASA Standard Product introduced a new method of fitting
the observed Earthshine radiances to yield total SCDs
. This new fitting approach eliminates a
positive bias identified by , and reduces the total
SCDs retrieved. For much of the globe, this reduction is attributed to the
stratospheric SCD, but over the continental US, it is attributed to the
tropospheric SCD. Thus, the broad reduction in tropospheric VCDs seen here
(Fig. a, Tables
and ) due to the new SCD fitting is consistent with
.
Percent change in the total tropospheric NO2 column due to
each of the algorithm improvements for the subproduct using daily profiles.
Changes due to (a) new NO2 profiles, (b) new
temperature profiles, (c) new gridding, (d) change in
profile time selection and ocean reflectance LUT from 430 to 460 nm,
(e) switch to WRF-derived tropopause pressure, (f) switch
to the surface pressure methodology. Note that
in (a), the difference is against an increment using monthly average
profiles; also note that the color scale varies among the plots. Averages are
for June–August 2012 and exclude pixels affected by the row anomaly and with
cloud fraction >0.2. Wintertime changes and histograms are given in
Sect. S5.
Percent differences in averaged NO2 VCDs for each increment.
Means are given with 1σ uncertainties; medians are given with
uncertainties as the distance to the upper and lower quartiles. “Monthly”
and “Daily” in the first column indicate which subproduct is considered
(Sect. ). “Ocean LUT” refers to the ocean surface
reflectance LUT. Outliers were removed before calculating these
statistics.
Percent differences in individual pixels' NO2 VCDs for each
increment. Means are given with 1σ uncertainties; medians are given
with uncertainties as the distance to the upper and lower quartiles.
“Monthly” and “Daily” in the first column indicate which subproduct is
considered (Sect. ). “Ocean LUT” refers to the
ocean surface reflectance LUT. Outliers were removed before calculating these
statistics.
* Statistics only for ocean pixels. n/a: not applicable.
Surface reflectanceLand reflectance
BEHR v3.0A calculated the land surface reflectance by using the BRF
coefficients computed from the MODIS instruments to compute the directional
surface reflectance for the solar and viewing angles specific to each pixel.
Previous versions of BEHR used a black-sky albedo with no directional
dependence. The version of the MODIS surface reflectance products used was
also upgraded from version 5 in BEHR v2.1C and prior to version 6 in BEHR
v3.0A.
Figure b shows the difference in summertime NO2
VCDs resulting from the change in surface reflectance products.
Figure , panels a and d, show the overall summer
and winter changes in surface reflectance. Panels b, c, e, and f decompose
this change into the change in the MODIS product version (version 5 to 6,
panels b and e) and from black sky to BRF (panels c and f).
(a, d) Difference in surface reflectance between BEHR v2.1C
(MODIS MCD43C3 black sky albedo, old ocean look-up table) and BEHR v3.0B
(MODIS MCD43Dxx BRF, new look-up table). (b, e) Difference in
surface reflectance between versions 5 and 6 of the MODIS black sky albedo
(no change in ocean look-up table). (c, f) Difference in surface
reflectance between the MODIS black sky and BRF product and the change in the
ocean look-up table. (a–c) are for summer (JJA) and (d–f)
are for winter (DJF). (g) The ocean albedo look-up table values for
v2.1C, v3.0A, and v3.0B. (The change between v3.0A and v3.0B is discussed in
Sect. .)
Generally, UV-vis AMFs increase (thus NO2 VCDs decrease) with
increasing surface reflectance, due to greater sensitivity to near surface
NO2. This pattern is apparent when comparing
Figs. b and a, as changes in
the NO2 VCDs show the expected inverse relationship to the changes in
surface reflectance. These changes in average surface reflectance are due
primarily to the upgrade from version 5 to version 6 of the MODIS product, as
we see larger average changes in land surface reflectance between the version
5 and 6 black-sky product (Fig. b, e) than between
the version 6 black-sky and BRF products
(Fig. c, f). Further, we see that the spatial
pattern due to the surface reflectance seen in the BEHR VCDs
(Fig. b) is well correlated with the spatial pattern of
changes between versions 5 and 6 of the MODIS black-sky product
(Fig. b). Differences between versions 5 and 6
were listed at
https://lpdaac.usgs.gov/dataset_discovery/modis/modis_products_table/mcd43c3_v006
(last access: 14 November 2018) as of 5 February 2018. Two improvements of
note are the following.
Change from a land cover-based backup database to one based on full inversions.
Notably, the summertime decreases in surface reflectance along the east coast
(Fig. a, b) are
somewhat spatially correlated with deciduous broadleaf forest, mixed forest,
and woody savanna land cover types that are rare elsewhere in the country
(Fig. S8).
Change from using the majority snow or no-snow status from the 16-day
observation window to the current-day status. In
Fig. , panels d and e, the largest changes are
seen sporadically in the northern half of the country, which suggests snow
cover is impacting the surface reflectance.
We have not rigorously tested these specific changes as the cause for the
spatial pattern of changes in surface reflectance; rather, our point is that
the change from version 5 to 6 of the MODIS products is a larger driver of
the change in average surface reflectance than the change from black-sky to
BRF. However, when we consider the changes of individual pixels, we find that
the difference between the the black sky and BRF surface reflectance is much
more variable (Fig. S7). The switch to a BRF surface
reflectance is expected to improve retrieval accuracy of individual pixels
and therefore is valuable to users interested in day-to-day variations in
NO2 VCDs .
Ocean reflectance
BEHR v2.1C used an ocean reflectance look-up table embedded in the core code
that defined the dependence of the ocean reflectance on solar zenith angle
(SZA). As documentation of the source of this table is not available, BEHR
v3.0A switched to a new look-up table calculated explicitly using the Coupled
Ocean-Atmosphere Radiative Transfer (COART) model . The
difference in the SZA dependence of the look-up tables is shown in
Fig. g. The overall shape is similar, but the
difference between small and large SZAs is less pronounced in the new ocean
look-up table. Both are similar to the ocean surface reflectance calculated
by for an atmospheric aerosol optical depth of 1, but for
different wind speeds: the BEHR v2.1C look-up table is more characteristic of
slow (<1 m s-1) winds, while the v3.0A table assumes a wind speed
of 5 m s-1.
At small SZAs characteristic of summer OMI observations (<35∘),
the new look-up table yields a ∼50% greater ocean reflectance than
the old table, which leads to the off-shore reflectance changes seen in
Fig. a. At larger SZAs more characteristic of
winter (∼40 to 60∘), the difference between the old and new
look-up tables shrinks, resulting in less change in the wintertime ocean
surface reflectance (Fig. d).
Especially in summer, since the relative change in the ocean surface
reflectance is large, using the new ocean look-up table does result in large
relative changes to the NO2 VCDs. Along the coasts, these changes can
reach 2×1015 to 3×1015molec.cm-2
(or more near New York, NY), but away from the coasts, the absolute
differences are quite small.
New WRF-Chem profilesUpdate to new monthly average profiles
There are three significant changes from the old monthly average profiles
used in v2.1C and before to those used in v3.0A.
Lightning NOx emissions are included in the profiles; these were not available in WRF-Chem when the previous profiles were simulated.
The anthropogenic emissions used now are from the National Emissions Inventory (NEI, 2011), scaled based on total annual emission to 2012 levels. 2012 boundary conditions and meteorology are also used. In v2.1C and earlier, NEI 2005 emissions were used.
The chemical mechanism was updated from the Regional Acid Deposition Model, version 2 to the custom mechanism described in Sect. .
The changes in the summer average VCDs due to the update to the monthly
profiles are shown in Fig. c. The effect of including
lightning NOx emissions is most apparent, causing the ∼30% decrease (5th/95th percentiles: 8 % and 55 %) in VCDs in
the southeastern US (averaged east of 95∘ and south of 45∘).
This is due to the increased contribution of upper tropospheric (UT)
NO2 to the a priori profiles compared to the v2.1C profiles. As this
NO2 is located at higher sensitivity altitudes, the AMF is increased
(and the retrieved VCD decreased) to reflect that higher sensitivity.
The increased VCDs along the western coast are caused by changes to the UT
NO2 profiles. The UT NO2 over the west decreased compared to
the old a priori profiles. This may be due either to the change in chemical
mechanism or to a change in the O3 boundary condition, which would
affect the simulated UT NO:NO2 ratio.
(a, b) Percent difference in v3.0A NO2 VCDs using
daily instead of monthly profiles averaged over (a) June–August and
(b) January, February, and December 2012. Averages exclude pixels
affected by the row anomaly and with cloud fraction >0.2.
Daily vs. monthly profiles
Figure a shows the difference in summer
NO2 VCDs using the new daily profiles compared to the old v2.1C
monthly profiles. Figure shows the difference
in v3.0A of the average total tropospheric NO2 columns when using
daily NO2 profiles rather than monthly average profiles. Figure
a is the summer (JJA) average, and shows a
significant increase in VCDs along the eastern US, which is not present in
the winter (DJF) average (Fig. b). The timing
and location suggest that this difference is due to lightning, as the
southeastern US especially has very active lightning .
Ultimately, the fact that lightning is an intermittent but significant
NOx source in the upper troposphere (UT) is the cause of
this difference. Figure a shows the
statistical distribution of NO2 in the UT for two regions in the US:
the southeast, which has significant lightning activity, and the northwest,
which has very little lightning. The distribution is highly skewed with a
long tail in the southeastern US due to the lightning activity, but not in
the northwestern US. Because of the nonlinear nature of the AMF calculation,
this skewed distribution translates into different average VCD values.
Figure , panels b and c, show average
shape factors derived from monthly averaged and daily a priori profiles for
the southeastern and northwestern US. A shape factor is a profile divided by
its integral:
S(p)=g(p)∫psurfptropg(p)dp.
A shape factor can be interpreted as the relative vertical distribution of
NO2. It appears implicitly in the AMF calculation
(Eq. ).
Here we see how the skewed UT NO2 distribution affects the
southeastern US AMFs through the shape factor.
Figure b shows that the statistically
skewed UT NO2 distribution causes shape factors calculated from the
monthly average a priori profiles in the southeastern US to have a larger
fraction of the column NO2 in the UT than that calculated from the
daily profiles. Through Eq. (), this leads to
systematically greater AMFs (and therefore smaller VCDs) in the southeast
when using the monthly profiles if the scattering weights (w(p) in Eq.
) are greater in the UT than near the surface, which is
usually the case. In contrast, Fig. c
shows no difference in the monthly or daily shape factors for the
northwestern US. For interested readers, a more mathematical argument is
given in Sect. S2.
The implication is that, for regions with long-tailed statistical
distributions of NO2 concentrations, there will be systematic
differences between a product using monthly average and daily a priori
profiles. It is likely that the VCDs calculated using the daily a priori
profiles are more accurate, because in theory daily a priori profiles should
properly account for that long tail on days when it is relevant, whereas
monthly profiles will average in the extreme values.
Finally we note that this difference between daily and monthly profiles may
change in the future. found that the simulation providing
the NO2 profiles had too much lightning in the southeastern US.
Correcting that may reduce the skewness of the UT NO2 distribution.
Work is underway to improve the representation of lightning for the
southeastern US NO2 profiles.
(a) Frequency distribution (normalized to maximum) of
average NO2 above 400 hPa in the a priori profiles for the
southeastern and northwestern US, from June to August 2012.
(b, c) Mean a priori NO2 shape factors over the southeastern
US (b) and northwestern US (c) for June–August 2012. Shape
factors are defined as the NO2 profile in a mixing ratio divided by
its integral in molec. cm-2. The error bars are ±1σ. The
regions (southeastern and northwestern US) are shown in
Fig. S4.
WRF-Chem temperature profiles
Simulated or recorded temperature profiles are necessary to correct for the
temperature dependence of the NO2 cross section
Sect. of this paper, also.
BEHR v2.1C used temperature profiles provided to us by NASA at 5∘×2∘ resolution . Recently, an error was
identified in the temperature profile lookup used in BEHR v2.1C. Correcting
this error changes the v2.1C VCDs by -1.7%±3.8% in the
summer, and -0.9%±11.2% winter
(Fig. S9). Therefore the impact was small in both
seasons, but more variable in the winter.
BEHR v3.0A uses temperature profiles from WRF-Chem at 12 km resolution
instead. The effect on total tropospheric VCDs is shown in
Fig. d (monthly a priori profiles) and
Fig. b (daily a priori profiles). It is small,
0.5%±0.4% on average in summer using monthly average profiles.
Using daily temperature profiles, the change is slightly more variable
(0.6%±0.5%). Therefore, high-resolution temperature profiles
are significantly less important than NO2 profiles, which is
expected, as temperature should not vary as rapidly in space as NO2.
Gridding method
BEHR v2.1C used a constant value method (CVM) gridding algorithm to
oversample the native pixel data to a fixed 0.05∘×0.05∘ grid. A constant value method assigns the VCD of a given pixel
to any grid points within the pixel bounds; this works well when the grid
resolution is significantly finer than the native pixel resolution. It was
found that the existing algorithm was at times overly conservative, and did
not assign values to grid cells near the border of two pixels.
BEHR v3.0A also uses a CVM gridding algorithm, however the implementation was
changed. The new CVM algorithm is a slightly modified version of that
provided by , with a custom interface to allow
communication between the Python code from https://github.com/gkuhl/omi
(last access: 14 November 2018) and the BEHR Matlab code.
We also tested the parabolic spline method (PSM) of gridding described in
, with updates from . The PSM attempts to
recover maxima in NO2 between adjacent pixels by fitting the
NO2 VCDs with 2-D splines and sampling the grid points along those
splines. While this algorithm should be an ideal match with our retrieval (as
our high-resolution profiles are able to better resolve urban–rural
NO2 gradients), two technical challenges persisted. First,
non-physical oscillations in the NO2 VCDs would appear, especially on
the edge of the row anomaly. Second, in one test, the PSM algorithm resulted
in much greater VCDs than the CVM algorithm over a large area. As this is not
the expected behavior, v3.0A uses the new CVM method from .
Figures e and c shows the
percent change in the VCDs resulting from the change in gridding method for
the subproducts using monthly and daily a priori profiles, respectively. The
average effect is small and no spatial pattern is evident, as would be
expected, although individual effects are quite variable (Table
). The new CVM algorithm correctly assigns grid cells
near the border of two pixels to one or the other. If two pixels overlap, an
average of their values weighted by the inverse of their area (FoV75Area from
the OMPIXCOR product) is assigned.
Changes in BEHR v3.0B
v3.0B implemented six main changes from v3.0A.
Retrievals using daily WRF-Chem profiles use the profile nearest in time
to OMI overpass, rather than the last profile before the OMI overpass
Ocean surface reflectance calculated at 460 nm instead of 430 nm
Variable tropopause pressure (derived from WRF simulations) implemented in the AMF calculation
The method for calculating surface pressure from was implemented
Clear and cloudy scattering weights are included separately in the native pixel files
The summary bits in the BEHRQualityFlags field were corrected.
Changes nos. 1–4 directly affect the retrieved VCDs. No. 5 is intended for
advanced users who wish to implement custom profiles. No. 6 makes rejecting
low-quality data easier for standard users.
Profile time effects
In v3.0A, when using daily profiles, the last set of profiles before the OMI
overpass time was used. In v3.0B, this was changed to be the nearest profile
in time. The overall average difference is near 0
(Fig. d, Tables and
), and the absolute magnitude of the average
changes is <4×1014 molec. cm-2. As expected, a difference
of 1 h in some of the selected profiles makes very little difference to the
average retrieved column density.
Ocean surface reflectance LUT effects
Figures f and d show the
changes to the NO2 VCD (for the subproducts using monthly and daily a
priori profiles, respectively) caused by the change to the wavelength of the
ocean reflectance LUT and the selection of the closest profile in time.
Figure f only shows the effect due to the ocean surface
reflectance LUT, as the monthly a priori profiles are not affected by the
change in how the closest daily profile in time is selected.
In v3.0A, the ocean surface reflectance was calculated at 430 nm as the
approximate midpoint of the wavelength fitting window for an NO2
retrieval 402–465 nm,. In v3.0B, this was changed to
be 460 nm, which is within the MODIS band used (459–479 nm). While both
approaches have merit, we chose to move towards calculating the surface
reflectance at similar wavelengths for consistency between the ocean and land
data. The change in VCD retrieved over ocean is very small (<1%,
Tables and ), as expected.
Implementation of variable tropopause height
BEHR v3.0B uses variable tropopause pressure derived from WRF simulations
while in prior versions the tropopause pressure is set to be 200 hPa.
Figures g and e reflect the
effect of changes in tropopause pressure on NO2 VCD. The changes in
NO2 are consistent with the variation in tropopause pressure. In
summertime, the WRF-derived thermal tropopause pressure in lower latitudes
(<45∘ N) is less than 200 hPa. This increases the contribution
of the UT, where OMI is highly sensitive to NO2, to the AMF, which in
turn reduces the retrieved NO2 VCDs. In higher latitudes (>45∘ N), the thermal tropopause pressure is greater than 200
hPa and leads to a slight increase in NO2 VCD. The changes in
average NO2 VCD caused by changes in tropopause pressure are small,
-1.6%±5.3% using monthly average profiles and -1.1%±8.2% using daily profiles. In wintertime, the WRF tropopause is below
the previous 200 hPa value over most of the US and it causes a broad
enhancement of NO2 VCD in most US domain (>30∘ N) by
approximately 2 % (Figs. S12, S15, Tables ,
). Evaluation of the tropopause pressure
calculation is ongoing; the calculation of the tropopause pressure may be
revised in future versions of BEHR.
Surface pressure calculation
Figures h and f show the
impact of switching from a fixed scale height calculation to using the
hypsometric equation to adjust WRF modeled surface pressure to the GLOBE
terrain elevation. As expected, the changes are similar whether monthly or
daily WRF output is used and are greatest over the Rocky and Appalachian
mountains (up to a maximum of ∼10 %). This is similar to the
5 % effect found in the summer, indicating that the
meteorological surface pressure correction in mountainous regions does impact
the NO2 columns even with a high-resolution terrain elevation
database.
Publishing separate clear and cloudy scattering weights
Very advanced users may wish to recalculate custom AMFs using their own
NO2 profiles but with the scattering weights used in BEHR. To
facilitate this, an array of scattering weights used in the BEHR AMF
calculation is included in the published native pixel resolution files. In
BEHR v3.0A and prior, these scattering weights were the cloud radiance
fraction weighted average of the temperature-corrected clear and cloudy
scattering weights:
w′(p)=(1-f)wclear(p)α(p)+fwcloudy(p)α(p),
where α(p) is defined by Eq. () and
wclear(p) and wcloudy are set to 0 below the
surface and cloud pressures, respectively.
Using these scattering weights along with the published a priori profiles,
users could reproduce BEHR AMFs well, to within 0.5%±1.9%,
using
A′=∫psurfptropw′(p)g(p)dp∫psurfptropg(p)dp,
where g(p) is the a priori profile also provided in the BEHR product.
However, publishing the clear and cloudy weights separately increases the
precision of reproduced AMFs by three orders of magnitude. Using these with
the provided BEHR a priori profiles allows users to reproduce BEHR AMFs
effectively exactly using Eq. () (Fig. S10). The primary
purpose is to allow users to replace the BEHR NO2 profiles with their
own for a custom AMF calculation. In theory, this also permits advanced users
to use different cloud fractions in their custom AMF calculations, but doing
so would require careful attention to possible errors, as the scattering
weights are tied to the cloud pressure used in BEHR.
BEHR quality flags
Starting with v3.0A, the BEHRQualityFlags field summarized key quality issues
from both the NASA and BEHR processing steps. The first and second bits in
these values are summary bits, so that users who want high-quality data can
very easily identify such data. Due to a bug in v3.0A, these bits did not
filter out all low-quality data. This has been rectified in v3.0B. See
Sect. for the proper use of these flags. These
flags may be updated in the future if additional causes of low-quality
NO2 VCDs are identified. Users should be sure to check the changelog
at http://behr.cchem.berkeley.edu/Portals/2/Changelog.txt (last access:
14 November 2018) for any changes to the flags. If the quality flags are
updated, the BEHR version number will be incremented, either by a major
version (e.g., v3.0B to v3.0C) or a minor revision (e.g., v3.0B to
v3.0Brev1).
Overall average differences in total tropospheric NO2 VCDs
between v2.1C and v3.0B for June–August (a, c) and January,
February, December (b, d) of 2012. (a, b) using monthly
NO2 profiles in v3.0B, (c, d) using daily profiles in
v3.0B.
Overall difference
Overall, the two changes that had the largest impact on the retrieved VCDs
were the new NASA slant column fitting and the new a priori NO2
profiles (-14%±14% and 0.86%±20.14%,
respectively, Table ). Although the overall average
effect of the new profiles is small, this is only because it causes both
positive and negative changes to the VCDs. The large standard deviation
reflects how different areas do have very significant changes. The effects of
the a priori profiles were especially strong in the SE US where lightning has
a strong influence on the profile shape in the summer
(Fig. ). Given the high sensitivity of NO2
retrievals to upper tropospheric NO2, this is not surprising. The
omission of lightning NO2 from the original BEHR product was a
limitation of WRF-Chem at the time the product was created ;
lightning NOx emission was not added to WRF-Chem until
v3.5.0, released in April, 2013
(http://www2.mmm.ucar.edu/wrf/users/download/get_sources.html#WRF-Chem,
last access: 14 November 2018). The change due to the SCD fitting resulted in
a fairly uniform decrease in NO2 VCDs across the domain.
The difference in the averages using daily (Fig. c, d)
vs. monthly profiles (Fig. a, b) is variable.
, did not see a significant difference in average VCDs
using daily versus monthly a priori profiles. Those results were obtained
using a model simulation without lightning. In regions without significant
lightning, this is expected because averaging over time periods greater than
a month eliminates the temporal variability captured by the daily profiles.
However, when there is significant lightning, the skewed UT NO2
distribution results in significant differences to the averages. The effect
of the daily profiles is on the average strongest in the SE US, as discussed
in Sect. , and is still an overall decrease
compared to the v2.1C profiles, due to the inclusion of lightning and the
reduction in surface emissions.
It should be noted that the difference between retrievals with daily and
monthly profiles will be greater in years other than 2012, since the daily
profiles incorporate year-specific emissions, while monthly profiles always
assume 2012 anthropogenic emissions.
Recommendations for use
In our experience, the most common use of the BEHR data falls into three categories:
Direct use of the NO2 VCDs for various purposes, including
calculation of NO2 trends, direct inference of lightning or other
emissions, etc.
Inverting the VCDs to obtain surface NO2 concentrations
Comparing the BEHR VCDs to modeled VCDs to evaluate the model, infer
emissions by constraining the model, etc.
Here we will give a brief summary of recommendations to use BEHR for each of
these applications, as well as general recommendations that apply to all uses
of the data.
General recommendationsQuality filtering
It is vital in any use of BEHR data to filter out low-quality data. The BEHR
algorithm attempts to calculate an NO2 VCD for as many pixels as
possible, even if some of those pixels are known to be of poor quality. The
philosophy is that it is better to have data for a pixel if at all possible
and to remove it only if the quality is too low for a particular application.
Some causes of low quality (e.g., the row anomaly,
https://projects.knmi.nl/omi/research/product/rowanomaly-background.php,
last access: 14 November 2018) make the NO2 column unusable in any
case, while others (e.g., high cloud fraction, low-quality surface
reflectance) only affect certain uses.
The quality of the pixel is summarized in the first two (least-significant)
bits of the BEHRQualityFlags field. The second bit is a critical error bit,
if set (i.e., if a bitwise AND of BEHRQualityFlags with 2 is >0) then the
NO2 columns for that pixel should not be used under any conditions.
The first bit is a quality flag bit; if it is set (if a bitwise AND of
BEHRQualityFlags with 1 is >0) then the use of the column for typical
applications wanting information down to the surface is not recommended;
however, other applications may still find use for this pixel. For example,
the first bit is set if the OMI geometric cloud fraction is >0.2, since
the uncertainty of the total tropospheric column increases greatly as more
NO2 is obscured by clouds, but cloud slicing approaches
e.g., will actually prefer large cloud fractions,
and so will need to do their own cloud filtering. For most applications
however, it is recommended to ignore pixels that have the first (i.e.,
quality summary) bit set to 1.
Users must also be sure to remove fill values. The fill value for each field
is defined in the “fillvalue” attribute. Generally, checking whether a
value is exactly equal to a fill value is not recommended unless the value is
an integer type, as floating point error on some systems may cause fill
values to be missed. It is better practice to check for values within some
relative tolerance of the fill value:
|x-f|<|f|⋅t,
where x is the data, f the fill value, and t the tolerance. t=10-4 works in our experience.
Choice of daily or monthly profile subproduct
Users will also need to choose whether to use the subproduct with daily
profiles. Use of the subproduct with daily profiles is strongly encouraged if
possible, for two reasons. First, the daily profiles also use year-specific
emissions (Sect. ), and so will better capture
trends in VCDs as the surface contribution to the a priori profiles is
reduced. Second, showed that using daily profiles
significantly changes day-to-day VCDs, and that some applications of
satellite data can be biased when monthly profiles are used. Applications
similar to those studied in , where upwind or downwind
columns are systematically averaged together, are particularly vulnerable to
bias when monthly average profiles are used.
Caution is advised if comparing 2005 or 2006 data using daily profiles to
other years; the different WRF-Chem boundary conditions
(Sect. ) may also bias observed trends. This effect
is likely small, as in a test of 1 week of data using two sets of profiles,
one using GEOS-Chem boundary conditions and one using MOZART boundary
conditions, the mean change was <1014molec.cm-2, and
only 0.7 % of pixels with any cloud fraction had a change exceeding 1×1015molec.cm-2 (0.05 % of pixels with cloud
fraction <0.2).
Mixing daily and monthly profile subproducts is strongly
discouraged, as systematic differences between them i.e.,
Sect. of this paper; will bias
any trends observed.
Application no. 1: direct observation of VCDs
Direct observation of VCDs has a number of applications, including
elucidating trends in NO2 burdens e.g.,
or inferring lightning emissions e.g.,. Users wanting
to average BEHR data over a given time period, e.g., to compare summer
average NO2 columns for different years, will find this easiest using
the gridded data, as this places the NO2 columns on a consistent
equirectangular latitude–longitude grid (i.e., the data in grid cell (1,1)
will be at the same lat–lon in each orbit, whereas in the native data, pixel
(1,1) will not), so it is easy to average across different days. When
averaging, each grid cell should be weighted by the area weight value given
in the gridded product; this is the inverse of the pixel area, so weighting
by this inherently gives more weight to smaller, more representative pixels.
Users interested in VCDs from individual days (e.g., to find NO2
downwind of an episodic event such as lightning) can use either the native
pixel or gridded products, whichever is easier. In this case, it is important
to keep in mind that pixel sizes vary from day to day. Therefore, if the
source signal of interest is smaller than a single pixel, it will be more
diluted if it falls in a larger pixel on the edge of the OMI swath than a
small one near the center.
Since a VCD is a measurement integrated over the troposphere, it does not
directly provide information about the surface concentration of NO2.
The simplest approach to infer ground-level NO2 concentrations from
VCDs is to multiply the BEHR VCD by the ratio of surface concentration to VCD
obtained from a modeled NO2 profile :
NO2surf=gpsurf∫psurfptropg(p)dpVBEHR,
where g(p) is the modeled profile, psurf the surface press,
ptrop the tropopause pressure, and VBEHR the BEHR
VCD. g(p) may be obtained in many ways; for users without model output or
measurements of NO2 profiles, the a priori profiles used in BEHR are
included in the native pixel subproduct and may be used for this purpose. In
this case, using the subproduct with daily profiles is highly recommended so
that the profiles respond to changes in meteorology day to day, especially wind fields.
Application no. 3: comparing to models
Users wishing to compare BEHR VCDs to model output should follow the
suggestions in . This requires calculating the overlap
between the BEHR pixels and the user's model grid cells and applying the BEHR
averaging kernel to the user's model profile before calculating the model
VCD, so the native pixel product must be used, since it contains the
averaging kernels and the pixel corners.
The averaging kernels would be applied to the model profile as
Vmodel=∑kckak,
where Vmodel is the modeled VCD after applying the averaging
kernels, k is the level index, ck is the model profile converted to a
partial column for level k, and ak is the averaging kernel for level
k.
There are three important considerations in this application. First, since
BEHR provides only a tropospheric VCD, it must be compared against a modeled
tropospheric column, no stratospheric component may be included.
Second, the model NO2 profile should be interpolated to the pressure
levels on which the averaging kernels are defined (given in the BEHR files as
BEHRPressureLevels) rather than the other way around. This is because the
averaging kernels may have sharp changes between levels (usually at the cloud
pressure, since OMI's sensitivity increases dramatically over a bright
cloud), so interpolating the averaging kernels to the model pressures is more
likely to introduce errors.
Third, the model profile is best converted to partial columns before applying
the averaging kernels. This may be done in several ways, such as the
following.
Interpolate the profile to the averaging kernels' pressure levels, then
multiply the profile concentration as number density by the layer height.
Interpolate the profile to the edges of the averaging kernels'
levels, then integrate over each layer to obtain the partial column.
Both methods need the edge of the pressure levels, either to calculate the
box height or to define the limits of the integration. Since the pressures
given for the averaging kernels are the level centers, the edges are most
easily defined as the midpoints between those layers; with the surface
pressure serving as the lower limit of the bottom layer and the tropopause
pressure serving as the upper limit of the top layer.
Converting from pressure to altitude for either method can either be done
using a scale height relation (e.g., Eq. ), though this
will likely introduce some error as we saw in
Sect. that the meteorological correction can
be significant. A better option, if the user's model output includes altitude
and pressure vectors, is to interpolate the altitude from the model to the
averaging kernels' pressure levels alongside the NO2. Alternatively,
in the second method, NO2 profiles in mixing ratio can be directly
integrated over pressure Appendix B. This is done
internally in BEHR using the integPr2 code at
https://github.com/CohenBerkeleyLab/BEHR-core-utils/blob/develop/AMF_tools/integPr2.m
(last access: 14 November 2018).
BEHR data are stored in monthly compressed files as
four subproducts on the University of California DASH archive
. All BEHR data are also available for download at
http://behr.cchem.berkeley.edu (last access: 14 November 2018). The
BEHR code is hosted on GitHub at
https://github.com/CohenBerkeleyLab/BEHR-core/tree/master (last access:
14 November 2018) . WRF-Chem simulations for 2005–10, and
2012–2014 are available at the time of writing. Full model output is
available for 2005, 2007–2009, and 2012–2014; a reduced set of variables is
stored for 2006 and 2010 to save space. Due to the large file size, access
currently must be arranged by contacting the corresponding author; work is
underway to make it available through http://behr.cchem.berkeley.edu/.
The analysis code for this paper (and its dependencies) along with the
incremental averages are available at
https://doi.org/10.5281/zenodo.1247564.
The v3.0 NASA Aura OMI NO2 standard product and
OMI/Aura Ground Pixel Corners product was obtained from the
Goddard Earth Science Data and Information Services Center (GES DISC) in
Greenbelt, MD, USA. The MODIS Aqua Clouds 5-Min L2 Swath 1 and 5 km
MYD06_L2 and MODIS Terra+Aqua BRDF/Albedo Parameters 1–3
Band3 and QA BRDF Quality Daily L3 Global 30ArcSec CMG V006 MCD43D07,
MCD43D08, MCD43D09, MCD43D31 were
acquired from the Level-1 and Atmospheric Archive and Distribution System
(LAADS) Distributed Active Archive Center (DAAC) located in the Goddard Space
Flight Center in Greenbelt, MD (https://ladsweb.nascom.nasa.gov/, last
access: 14 November 2018).
Conclusions
Here we present v3.0 of the Berkeley High Resolution OMI
NO2 product (BEHR NO2). This version incorporates a number of
changes, including updated a priori NO2 profiles with lightning
NOx emissions, daily NO2 profiles for select years,
a directional surface reflectance product, variable tropopause height, a new
gridding algorithm, and improved surface pressure calculation, in addition to
using the current NASA OMI NO2 Standard Product. The new a priori
profiles and the upgrade to the new NASA product had the largest effect on
the retrieved total tropospheric VCDs. Retrieved visible-only tropospheric
VCDs were most strongly affected by the new visible-only AMF formulation, but
otherwise were similarly affected by each change.
Published formatFile structure
BEHR data are published as HDF version 5 files. Each file contains a single,
top-level group “Data”, which in turn contains each orbit as a child group
named “SwathX” where X is the orbit number. The datasets for each orbit are
contained in the “SwathX” groups.
Separate HDF files contain data at the native OMI pixel resolution and
regridded to 0.05∘×0.05∘ resolution. The regridded
files only contain a subset of the variables stored in the native pixel
files. The regridded files contain each orbit gridded separately; each
orbit's grid covers the entire domain retrieved. Grid cells outside each
orbit's observed swath contain fill values. Users can identify whether a file
contains gridded information by the dataset level attribute
“gridding_method”, if present, the file is a gridded file; if absent, the
file is a native pixel file. Additionally, the “Description” attribute
contained in each swath indicates whether the data are at native or regridded
resolution.
Retrievals using daily vs. monthly NO2 a priori profiles are
available separately. Retrievals using monthly profiles will be updated as
new OMI and MODIS data becomes available. Retrievals using daily profiles are
limited by the need to model said profiles; these will become available as
modeled NO2 profiles are simulated.
BEHR files are named with the format
“OMI_BEHR-profile_region_version_yyyymmdd.hdf”,
where:
profile will be DAILY or MONTHLY, indicating whether daily or monthly
NO2 a priori profiles were used
region region retrieved, currently, US = continental United States.
version is the version string (Sect. ).
yyyymmdd is the date of the observation
This information is also contained as swath level attributes
“BEHRProfileMode”, “BEHRRegion”, “Version”, and “Date”, respectively.
Key variables
The BEHR files contain a large number of variables, including a large amount
of ancillary data used in the algorithm. All variables in the HDF files have
a “description” attribute that provides some information about what they
are. They also have a “product” attribute that indicates whether they are
taken verbatim from the NASA Standard Product (product = “SP”) or added by
BEHR (product = “BEHR”). The primary variables that most users should focus
on are:
BEHRColumnAmountNO2Trop: this is the tropospheric VCD calculated
using Eqs. ()
and (). It is the concentration of NO2 integrated
from the surface to the tropopause, including NO2 below clouds. This
is the NO2 value that most users should use. Historical note: the BEHR v2.1A documentation indicated that this was a visible-only
VCD; that was incorrect. This value has been the total tropospheric column in
all BEHR versions.
BEHRColumnAmountNO2TropVisOnly: this is the visible-only
tropospheric VCD calculated with Eqs. ()
and (). It excludes below-cloud NO2. Generally
the use for this quantity is more specialized; most users should use the
previous value.
BEHRQualityFlags: a 32-bit unsigned integer value where each bit
represents a boolean flag indicating the presence of a specific error or
warning for that pixel. See Sect. for
details.
Areaweight (gridded products only): a weight calculated
of the inverse of the area of the pixel that each grid cell falls within.
This should be used to weight the gridded data during temporal averaging (see
Sect. ).
Longitude, Latitude: the coordinates of the pixel or
grid cell center.
CloudFraction: this is a geometric cloud fraction from the OMI
O2–O2 cloud product . It is the default
used to filter for cloudy pixels, and is the same as the corresponding
variable in the NASA Standard Product.
CloudRadianceFraction: this is a radiance cloud fraction
(i.e., one weighted by the amount of light coming from the cloud vs. the
ground). It is the same as the corresponding field in the NASA Standard
Product.
MODISCloud: this is a geometric cloud fraction from the Aqua
MODIS instrument averaged to the OMI pixels. It is an alternate
way of filtering for cloudy pixels that may be less susceptible to false
positives from highly reflective ground . Some pixels near
the edge of the swath may be missing this data since the MODIS swath width is
slightly smaller than OMI's.
More advanced users may find the 3-D variables included in the native pixel
subproducts useful. These variables give a unique vector of values for each
pixel. In Matlab, the vector for each pixel runs along the first dimension,
so if the NO2 VCDs are the 2-D array V and one of the 3-D
arrays is A, then the vector corresponding to V(i,j) would
be A(:,i,j). However, some languages reverse the order of the
dimensions. In BEHR v3.0B, the vector dimension can be identified as the one
with a length of 33.
In BEHR, these 3-D variables are defined on a vertical grid of 30 standard
pressure levels (ranging from 1020 to 60 hPa) with values interpolated to
the surface pressure, cloud pressure, and tropopause pressure included,
bringing the total length of the vertical dimension to 33. If one of the
interpolated pressure levels is the same as a standard pressure level, the
value is not duplicated, and the vector of values will be padded with fill
values at the end.
BEHRPressureLevels: this dataset defines the pressure levels that
the other 3-D variables are defined on.
BEHRNO2apriori: this dataset gives the NO2 a priori
profiles used in the BEHR retrieval in mixing ratio.
BEHRAvgKernels: these are the averaging kernels referenced in
Sect. . They are defined as
a(p)=(1-f)wclear(p)α(p)+fwcloudy(p)α(p)A,
where a(p) is the averaging kernel, f the cloud radiance fraction,
α(p) the temperature correction (Eq. ) A the BEHR
AMF, and wclear(p) and wcloudy(p) the clear and
cloudy scattering weights, which are set to 0 below the surface and cloud
pressure, respectively.
BEHRScatteringWeightsClear, BEHRScatteringWeightsCloudy:
the temperature corrected clear and cloudy scattering weights, set to 0 below
the surface and cloud pressure, respectively, i.e.,
wclear′(p)=wclear(p)α(p),wcloudy′(p)=wcloudy(p)α(p).
Quality flagging
BEHR data contains a 32-bit unsigned integer quality flag field that
summarizes quality errors and warnings from both the NASA processing and BEHR
processing. Each bit in the integer value represents a specific error or
warning flagged during processing. The bits are divided into three
categories; the bit number is the position of the bit (1-based) starting from
the least significant bit.
Bits 1 and 2: summary bits. These summarize the other 30 bits.
Users interested in simple filtering can focus only on these.
Bits 3–16: error bits. These are set to 1 for significant
errors in the retrieval that preclude the use of the corresponding
NO2 data in any capacity.
Bits 17–32: warning bits. These are set to 1 as non-fatal
warnings about the processing of the corresponding data. These do not
automatically preclude the use of the corresponding data, but rather provide
warnings of potentially lower-quality data or information about decisions
made during the retrieval. The flags for low-quality BRF data
(Sect. ) fall into this category.
The meaning of each used bit is given in the “FlagMeanings” attribute of
the BEHRQualityFlags dataset; here, we will only discuss the two summary
bits.
Bit 2 is the error summary bit; it is set to 1 if any error bit is set.
Therefore, NO2 columns from any pixel with this bit set should not be
used. In v3.0B, this is set if the NASA VcdQualityFlags or XTrackQualityFlags
fields indicate the pixel should not be used, or if the BEHR AMF is invalid
(usually because a WRF profile is not available for that pixel).
Bit 1 is the quality summary bit; in v3.0B, it is set to 1 if bit 2 is set,
the MODIS BRF coefficients are of low quality, or the OMI geometric cloud
fraction exceeds 20 %. Therefore, the NO2 data can be restricted
to high-quality, total tropospheric column data by using only pixels where
this bit is not set.
As an example, if a pixel is in the row anomaly, then the 5th bit will be
set, which also requires bits 1 and 2 to be set. The (little-endian) binary
representation for this pixel would be 11001 (followed by 0s), so the value
stored would be 20+21+24=19. The easiest way to figure out if the
nth bit is set is to do a bitwise AND operation between the quality flag
value and 2n-1, for n∈[1,32].
These quality flags focus on the quality of the NO2 retrieval;
therefore ancillary data (such as the MODIS surface reflectance or MODIS
clouds) is not necessarily unusable for pixels flagged with a retrieval
error.
In the gridded product, the quality flags field is a bitwise OR of all
contributing pixels' quality flags. Therefore, any error or warning in a
pixel that contributes to a grid cell is propagated to the grid cell.
Versioning
BEHR versions follow the format “vX-XYrevZ”, e.g., v3-0Arev0. The “X-X”
indicates the version of the NASA Standard Product that was ingested as the
basis for that BEHR retrieval. “Y” is a sequential letter (A, B, C, etc.)
indicating the major version of BEHR produced from the same NASA SP base;
i.e., v3-0A indicates the first major BEHR version based on the NASA SPv3.
“revZ” (short for “revision”) indicates a small update to the BEHR
product. Revisions are reserved for small changes that are not expected to
significantly affect scientific results obtained from the data, e.g., updates
to file format or attributes, or very uncommon error corrections. A revision
of 0 may be omitted from the version string; i.e., “v3-0A” and
“v3-0Arev0” are the same version.
Traceability
To ensure traceability, files ingested during processing from other satellite
products or models are recorded in the swath level attributes “OMNO2File”
(NASA NO2 SP data), “OMPIXCORFile” (pixel corner data),
“MODISCloudFiles” (MYD06 files that MODIS cloud data are taken from),
“MODISAlbedoFile” (MCD43Dxx files that BRF parameters are taken from), and
“BEHRWRFFile” (WRF-Chem output files the NO2 profiles are taken
from are post-processed for monthly average profiles).
The BEHR code is available on GitHub at
https://github.com/CohenBerkeleyLab/BEHR-core. Each
release will be tagged with the same version string as the data.
Additionally, 11 swath level attributes contain the Git SHA-1 hash of the
most recent commit of the core BEHR code and additional dependencies
at the time each of the three major steps in processing BEHR data is executed. These attribute names have the form
“GitHead_repo_step”, where repo will be one of
Core: the core BEHR repository
(https://github.com/CohenBerkeleyLab/BEHR-core, last access:
14 November 2018),
BEHRUtils: the repository of BEHR satellite utility functions
(https://github.com/CohenBerkeleyLab/BEHR-core-utils, last access:
14 November 2018),
GenUtils: the repository of general Matlab utilities
(https://github.com/CohenBerkeleyLab/Matlab-Gen-Utils, last access:
14 November 2018),
PSM: the repository containing the modified “omi” Python
package used for gridding
(https://github.com/CohenBerkeleyLab/BEHR-PSM-Gridding, last access:
14 November 2018),
MatPyInt: the Matlab–Python type conversion interface
(https://github.com/CohenBerkeleyLab/MatlabPythonInterface, last
access: 14 November 2018), and
WRFUtils: the repository containing Matlab utilities for working with WRF
data (https://github.com/CohenBerkeleyLab/WRF_Utils, last access:
14 November 2018),
and step will be one of
Read: step in which OMI, MODIS, and GLOBE data are ingested into
Matlab and (where necessary) averaged to OMI pixels,
Main: step in which scattering weights and NO2 profiles
are matched to OMI pixels, the BEHR AMFs and VCDs are calculated, and the
data are gridded, and
Pub: step in which the BEHR Matlab files are converted to HDF files.
The supplement related to this article is available online at: https://doi.org/10.5194/essd-10-2069-2018-supplement.
JLL, QZ and RCC articulated a vision for a new BEHR algorithm.
JLL and QZ converted those ideas to the code of the BEHR algorithm; JLL led
the writing of the manuscript with input from QZ and RCC. RCC provided
guidance and mentoring throughout the project and secured funding. All
the authors reviewed the manuscript.
The authors declare that they have no conflict of
interest.
Acknowledgements
The authors gratefully acknowledge support from NASA ESS Fellowship
NNX14AK89H, NASA grant NNX15AE37G, and TEMPO project grant SV3-83019.
We would like to acknowledge high-performance computing support from Cheyenne
(10.5065/D6RX99HX) provided by NCAR's Computational and Information
Systems Laboratory, sponsored by the National Science Foundation. This
research also used the Savio computational cluster resource provided by the
Berkeley Research Computing program at the University of California, Berkeley
(supported by the UC Berkeley Chancellor, Vice Chancellor for Research, and
Chief Information Officer).
We acknowledge use of WRF-Chem preprocessor tools MOZBC, fire_emiss, etc.,
provided by the Atmospheric Chemistry Observations and Modeling (ACOM)
laboratory of NCAR. We also thank Eric Bucsela and Jim Gleason for very
helpful discussions about the new formulation of the visible-only
AMF.
Edited by: David Carlson
Reviewed by: two anonymous referees
ReferencesAcarreta, J. R., De Haan, J. F., and Stammes, P.: Cloud pressure retrieval
using the O2-O2 absorption band at 477 nm, J. Geophys. Res.,
109, D05204, 10.1029/2003JD003915, 2004.Anderson, D. C., Loughner, C. P., Diskin, G., Weinheimer, A., Canty, T. P.,
Salawitch, R. J., Worden, H. M., Fried, A., Mikoviny, T., Wisthaler, A., and
Dickerson, R. R.: Measured and modeled CO and NOy in
DISCOVER-AQ: An evaluation of emissions and chemistry over the eastern US,
Atmos. Environ., 96, 78–87,
https://doi.org/10.1016/j.atmosenv.2014.07.004, 2014.Bechle, M. J., Millet, D. B., and Marshall, J. D.: National Spatiotemporal
Exposure Surface for NO2: Monthly Scaling of a Satellite-Derived
Land-Use Regression, 2000–2010, Environ. Sci. Technol., 49, 12297–12305,
10.1021/acs.est.5b02882, 2015.Beirle, S., Platt, U., Wenig, M., and Wagner, T.: NOx
production by lightning estimated with GOME, Adv. Space Res., 34, 793–797,
10.1016/j.asr.2003.07.069, 2004.Beirle, S., Huntrieser, H., and Wagner, T.: Direct satellite observation of
lightning-produced NOx, Atmos. Chem. Phys., 10,
10965–10986, 10.5194/acp-10-10965-2010, 2010.
Beirle, S., Boersma, K., Platt, U., Lawrence, M., and Wagner, T.: Megacity
Emissions and Lifetimes of Nitrogen Oxides Probed from Space, Science, 333,
1737–1739, 2011.Belmonte Rivas, M., Veefkind, P., Boersma, F., Levelt, P., Eskes, H., and
Gille, J.: Intercomparison of daytime stratospheric NO2 satellite
retrievals and model simulations, Atmos. Meas. Tech., 7, 2203–2225,
10.5194/amt-7-2203-2014, 2014.Boersma, F., Bucsela, E., Brinksma, E., and Gleason, J. F.: NO2, in:
OMI Algorithm Theoretical Basis Document, edited by: Chance, K., Smithsonian
Astrophysical Observatory, vol. IV: OMI Trace Gas Algorithms, Report #:
ATBD-OMI-02 Version 2.0, 13–36, available at:
https://ozoneaq.gsfc.nasa.gov/media/docs/ATBD-OMI-04.pdf (last access:
14 November 2018), 2001.Boersma, K. F., Eskes, H. J., Dirksen, R. J., van der A, R. J., Veefkind, J.
P., Stammes, P., Huijnen, V., Kleipool, Q. L., Sneep, M., Claas, J.,
Leitão, J., Richter, A., Zhou, Y., and Brunner, D.: An improved
tropospheric NO2 column retrieval algorithm for the Ozone Monitoring
Instrument, Atmos. Meas. Tech., 4, 1905–1928, 10.5194/amt-4-1905-2011,
2011.Boersma, K. F., Vinken, G. C. M., and Eskes, H. J.: Representativeness errors
in comparing chemistry transport and chemistry climate models with satellite
UV-Vis tropospheric column retrievals, Geosci. Model Dev., 9, 875–898,
10.5194/gmd-9-875-2016, 2016.
Boucher, O., Randall, D., Artaxo, P., Bretherton, C., Feingold, G., Forster,
P., Kerminen, V.-M., Kondo, Y., Liao, H., Lohmann, U., Rasch, P., Satheesh,
S., Sherwood, S., Stevens, B., and Zhang, X.: Climate Change 2013: The
Physical Science Basis. Contribution of Working Group I to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change, edited
by: Stocker, T. F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S. K.,
Boschung, J., Nauels, A., Xia, Y., Bex, V., and Midgley, P. M., Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA, chap.:
Clouds and Aerosols, 571–657, 2013.Bousserez, N.: Space-based retrieval of NO2 over biomass burning
regions: quantifying and reducing uncertainties, Atmos. Meas. Tech., 7,
3431–3444, 10.5194/amt-7-3431-2014, 2014.Browne, E. C., Perring, A. E., Wooldridge, P. J., Apel, E., Hall, S. R.,
Huey, L. G., Mao, J., Spencer, K. M., Clair, J. M. St., Weinheimer, A. J.,
Wisthaler, A., and Cohen, R. C.: Global and regional effects of the
photochemistry of CH3O2NO2: evidence from ARCTAS, Atmos. Chem.
Phys., 11, 4209–4219, 10.5194/acp-11-4209-2011, 2011.Browne, E. C., Wooldridge, P. J., Min, K.-E., and Cohen, R. C.: On the role
of monoterpene chemistry in the remote continental boundary layer, Atmos.
Chem. Phys., 14, 1225–1238, 10.5194/acp-14-1225-2014, 2014.Bucsela, E., Celarier, E., Wenig, M., Gleason, J., Veefkind, J., Boersma, K.,
and Brinksma, E.: Algorithm for NO2 vertical column retrieval from
the ozone monitoring instrument, IEEE T. Geosci. Remote, 44, 1245–1258,
10.1109/tgrs.2005.863715, 2006.Bucsela, E. J., Krotkov, N. A., Celarier, E. A., Lamsal, L. N., Swartz, W.
H., Bhartia, P. K., Boersma, K. F., Veefkind, J. P., Gleason, J. F., and
Pickering, K. E.: A new stratospheric and tropospheric NO2 retrieval
algorithm for nadir-viewing satellite instruments: applications to OMI,
Atmos. Meas. Tech., 6, 2607–2626, 10.5194/amt-6-2607-2013, 2013.Bucsela, E. J., Perring, A. E., Cohen, R. C., Boersma, K. F., Celarier,
E. A., Gleason, J. F., Wenig, M. O., Bertram, T. H., Wooldridge, P. J.,
Dirksen, R., and Veefkind, J. P.: Comparison of tropospheric NO2 from
in situ aircraft measurements with near-real-time and standard product data
from OMI, J. Geophys. Res., 113, D16S31, 10.1029/2007JD008838, 2008.Bucsela, E. J., Pickering, K. E., Huntemann, T. L., Cohen, R. C., Perring,
A., Gleason, J. F., Blakeslee, R. J., Albrecht, R. I., Holzworth, R.,
Cipriani, J. P., Vargas-Navarro, D., Mora-Segura, I., Pacheco-Hernández,
A., and Laporte-Molina, S.: Lightning-generated NOx seen by the Ozone
Monitoring Instrument during NASA's Tropical Composition, Cloud and Climate
Coupling Experiment (TC4), J. Geophys. Res., 115, D00J10,
10.1029/2009JD013118, 2010.Burnett, R., Chen, H., Szyszkowicz, M., Fann, N., Hubbell, B., Pope, C. A.,
Apte, J. S., Brauer, M., Cohen, A., Weichenthal, S., Coggins, J., Di, Q.,
Brunekreef, B., Frostad, J., Lim, S. S., Kan, H., Walker, K. D., Thurston,
G. D., Hayes, R. B., Lim, C. C., Turner, M. C., Jerrett, M., Krewski, D.,
Gapstur, S. M., Diver, W. R., Ostro, B., Goldberg, D., Crouse, D. L., Martin,
R. V., Peters, P., Pinault, L., Tjepkema, M., van Donkelaar, A., Villeneuve,
P. J., Miller, A. B., Yin, P., Zhou, M., Wang, L., Janssen, N. A. H., Marra,
M., Atkinson, R. W., Tsang, H., Quoc Thach, T., Cannon, J. B., Allen, R. T.,
Hart, J. E., Laden, F., Cesaroni, G., Forastiere, F., Weinmayr, G., Jaensch,
A., Nagel, G., Concin, H., and Spadaro, J. V.: Global estimates of mortality
associated with long-term exposure to outdoor fine particulate matter,
P. Natl. Acad. Sci. USA, 115, 9592–9597, 10.1073/pnas.1803222115,
2018.Burrows, J. P., Weber, M., Buchwitz, M., Rozanov, V.,
Ladstätter-Weißenmayer, A., Richter, A., DeBeek, R., Hoogen, R.,
Bramstedt, K., Eichmann, K.-U., and Eisinger, M.: The Global Ozone Monitoring
Experiment (GOME): Mission Concept and First Scientific Results, J. Atmos.
Sci., 56, 151–175, 10.1175/1520-0469(1999)056<0151:TGOMEG>2.0.CO;2,
1999.Canty, T. P., Hembeck, L., Vinciguerra, T. P., Anderson, D. C., Goldberg, D.
L., Carpenter, S. F., Allen, D. J., Loughner, C. P., Salawitch, R. J., and
Dickerson, R. R.: Ozone and NOx chemistry in the eastern US:
evaluation of CMAQ/CB05 with satellite (OMI) data, Atmos. Chem. Phys., 15,
10965–10982, 10.5194/acp-15-10965-2015, 2015.Carlton, A. G., Wiedinmyer, C., and Kroll, J. H.: A review of Secondary
Organic Aerosol (SOA) formation from isoprene, Atmos. Chem. Phys., 9,
4987–5005, 10.5194/acp-9-4987-2009, 2009.Castellanos, P., Boersma, K. F., Torres, O., and de Haan, J. F.: OMI
tropospheric NO2 air mass factors over South America: effects of biomass
burning aerosols, Atmos. Meas. Tech., 8, 3831–3849,
10.5194/amt-8-3831-2015, 2015.Chauhan, A., Krishna, M., Frew, A., and Holgate, S.: Exposure to nitrogen
dioxide NO2) and respiratory disease risk, Rev. Environ. Health, 13,
73–90, 1998.Choi, S., Joiner, J., Choi, Y., Duncan, B. N., Vasilkov, A., Krotkov, N., and
Bucsela, E.: First estimates of global free-tropospheric NO2
abundances derived using a cloud-slicing technique applied to satellite
observations from the Aura Ozone Monitoring Instrument (OMI), Atmos. Chem.
Phys., 14, 10565–10588, 10.5194/acp-14-10565-2014, 2014.de Foy, B., Lu, Z., Streets, D. G., Lamsal, L. N., and Duncan, B. N.:
Estimates of power plant NOx emissions and lifetimes from
OMI NO2 satellite retrievals, Atmos. Environ., 116, 1–11,
10.1016/j.atmosenv.2015.05.056, 2015.Dockery, D. W., Pope, C. A., Xu, X., Spengler, J. D., Ware, J. H., Fay,
M. E., Ferris, B. G., and Speizer, F. E.: An Association between Air
Pollution and Mortality in Six U.S. Cities, New Engl. J. Med., 329,
1753–1759, 10.1056/nejm199312093292401, 1993.Emmons, L. K., Walters, S., Hess, P. G., Lamarque, J.-F., Pfister, G. G.,
Fillmore, D., Granier, C., Guenther, A., Kinnison, D., Laepple, T., Orlando,
J., Tie, X., Tyndall, G., Wiedinmyer, C., Baughcum, S. L., and Kloster, S.:
Description and evaluation of the Model for Ozone and Related chemical
Tracers, version 4 (MOZART-4), Geosci. Model Dev., 3, 43–67,
10.5194/gmd-3-43-2010, 2010.EPA: Air Pollutant Emissions Trends Data, avalable at:
https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data,
last access: 11 October 2016.Fujita, E. M., Campbell, D. E., Zielinska, B., Chow, J. C., Lindhjem, C. E.,
DenBleyker, A., Bishop, G. A., Schuchmann, B. G., Stedman, D. H., and Lawson,
D. R.: Comparison of the MOVES2010a, MOBILE6.2, and EMFAC2007 mobile source
emission models with on-road traffic tunnel and remote sensing measurements,
J. Air Waste Manage., 62, 1134–1149, 10.1080/10962247.2012.699016,
2012.Goliff, W. S., Stockwell, W. R., and Lawson, C. V.: The regional atmospheric
chemistry mechanism, version 2, Atmos. Environ., 68, 174–185,
10.1016/j.atmosenv.2012.11.038, 2013.Grell, G. A., Peckham, S. E., Schmitz, R., McKeen, S. A., Frost, G.,
Skamarock, W. C., and Eder, B.: Fully coupled “online” chemistry within the
WRF model, Atmos. Environ., 39, 6957–6975,
10.1016/j.atmosenv.2005.04.027, 2005.Guenther, A., Karl, T., Harley, P., Wiedinmyer, C., Palmer, P. I., and Geron,
C.: Estimates of global terrestrial isoprene emissions using MEGAN (Model of
Emissions of Gases and Aerosols from Nature), Atmos. Chem. Phys., 6,
3181–3210, 10.5194/acp-6-3181-2006, 2006.
Haagen-Smit, A., Darley, E., Zaitlin, M., Hull, H., and Noble, W.:
Investigation on Injury to Plants from Air Pollution in the Los Angeles Area,
Plant Physiol., 27, 18–34, 1952.
Hastings, D. and Dunbar, P.: Global Land One-kilometer Base Elevation (GLOBE)
Digital Elevation Model, Documentation, Volume 1.0, National Oceanic and
Atmospheric Administration, National Geophysical Data Center, 325 Broadway,
Boulder, Colorado 80303, USA, 1999.
Heath, R. L.: Responses of plants to air pollution, chap. Ozone, Academic
Press, New York, NY, USA; San Francisco, CA, USA; London, United Kingdom,
23–55, 1975.Hudman, R. C., Jacob, D. J., Turquety, S., Leibensperger, E. M., Murray,
L. T., Wu, S., Gilliland, A. B., Avery, M., Bertram, T. H., Brune, W., Cohen,
R. C., Dibb, J. E., Flocke, F. M., Fried, A., Holloway, J., Neuman, J. A.,
Orville, R., Perring, A., Ren, X., Sachse, G. W., Singh, H. B., Swanson, A.,
and Wooldridge, P. J.: Surface and lightning sources of nitrogen oxides over
the United States: Magnitudes, chemical evolution, and outflow, J. Geophys.
Res., 112, D12S05, 10.1029/2006JD007912, 2007.Hudman, R. C., Russell, A. R., Valin, L. C., and Cohen, R. C.: Interannual
variability in soil nitric oxide emissions over the United States as viewed
from space, Atmos. Chem. Phys., 10, 9943–9952,
10.5194/acp-10-9943-2010, 2010.Hudman, R. C., Moore, N. E., Mebust, A. K., Martin, R. V., Russell, A. R.,
Valin, L. C., and Cohen, R. C.: Steps towards a mechanistic model of global
soil nitric oxide emissions: implementation and space based-constraints,
Atmos. Chem. Phys., 12, 7779–7795, 10.5194/acp-12-7779-2012, 2012.Huijnen, V., Flemming, J., Kaiser, J. W., Inness, A., Leitão, J., Heil,
A., Eskes, H. J., Schultz, M. G., Benedetti, A., Hadji-Lazaro, J., Dufour,
G., and Eremenko, M.: Hindcast experiments of tropospheric composition during
the summer 2010 fires over western Russia, Atmos. Chem. Phys., 12,
4341–4364, 10.5194/acp-12-4341-2012, 2012.Izumi, K. and Fukuyama, T.: Photochemical aerosol formation from aromatic
hydrocarbons in the presence of NOx, Atmos. Environ. A-Gen.,
24, 1433–1441, 10.1016/0960-1686(90)90052-O, 1990.Jacob, D. J., Logan, J. A., Gardner, G. M., Yevich, R. M., Spivakovsky,
C. M., Wofsy, S. C., Sillman, S., and Prather, M. J.: Factors regulating
ozone over the United States and its export to the global atmosphere, J.
Geophys. Res., 98, 14817–14826, 10.1029/98JD01224, 1993.Jiang, Z., McDonald, B. C., Worden, H., Worden, J. R., Miyazaki, K., Qu, Z.,
Henze, D. K., Jones, D. B. A., Arellano, A. F., Fischer, E. V., Zhu, L., and
Boersma, K. F.: Unexpected slowdown of US pollutant emission reduction in the
past decade, P. Natl. Acad. Si USA, 115, 5099–5104,
10.1073/pnas.1801191115, 2018.Jin, Z., Charlock, T. P., Smith, W. L., and Rutledge, K.: A parameterization
of ocean surface albedo, Geophys, Res, Lett,, 31, L22301,
10.1029/2004GL021180, 2004.
Jin, Z., Charlock, T., Rutledge, K., Stamnes, K., and Wang, Y.: Analytical
solution of radiative transfer in the coupled atmosphere-ocean system with a
rough surface, Appl. Optics, 45, 7443–7455, 2006.
Kagawa, J.: Evaluation of biological significance of nitrogen oxides
exposure, Tokai J. Exp. Clin. Med., 10, 348, 1985.Kampa, M. and Castanas, E.: Human health effects of air pollution, Environ.
Pollut., 151, 362–367, 10.1016/j.envpol.2007.06.012, 2008.Kharol, S., Martin, R., Philip, S., Boys, B., Lamsal, L., Jerrett, M.,
Brauer, M., Crouse, D., McLinden, C., and Burnett, R.: Assessment of the
magnitude and recent trends in satellite-derived ground-level nitrogen
dioxide over North America, Atmos. Environ., 118, 236–245,
10.1016/j.atmosenv.2015.08.011, 2015.Krotkov, Nickolay, A. and Veefkind, P.: OMI/Aura Nitrogen Dioxide (NO2) Total
and Tropospheric Column 1-orbit L2 Swath 13x24 km V003, Greenbelt, MD, USA,
Goddard Earth Sciences Data and Information Services Center (GES DISC),
10.5067/Aura/OMI/DATA2017, 2016.Krotkov, N. A., Lamsal, L. N., Celarier, E. A., Swartz, W. H., Marchenko, S.
V., Bucsela, E. J., Chan, K. L., Wenig, M., and Zara, M.: The version 3 OMI
NO2 standard product, Atmos. Meas. Tech., 10, 3133–3149,
10.5194/amt-10-3133-2017, 2017.Kuhlmann, G., Hartl, A., Cheung, H. M., Lam, Y. F., and Wenig, M. O.: A novel
gridding algorithm to create regional trace gas maps from satellite
observations, Atmos. Meas. Tech., 7, 451–467, 10.5194/amt-7-451-2014,
2014.Kurosu, T. P. and Celarier, E. A.: OMI/Aura Global Ground Pixel Corners
1-Orbit L2 Swath 13x24km V003, Greenbelt, MD, USA, Goddard Earth Sciences
Data and Information Services Center (GES DISC),
10.5067/Aura/OMI/DATA2020, 2010.Lamsal, L. N., Martin, R. V., van Donkelaar, A., Steinbacher, M., Celarier,
E. A., Bucsela, E., Dunlea, E. J., and Pinto, J. P.: Ground-level nitrogen
dioxide concentrations inferred from the satellite-borne Ozone Monitoring
Instrument, J. Geophys. Res., 113, D16308, 10.1029/2007JD009235, 2008.Laughner, J., Zhu, Q., and Cohen, R.: Berkeley High Resolution (BEHR) OMI NO2
– Gridded pixels, daily profiles, v5, UC Berkeley Dash, Dataset,
10.6078/D12D5X, 2018a.Laughner, J., Zhu, Q., and Cohen, R.: Berkeley High Resolution (BEHR) OMI NO2
– Native pixels, daily profiles, v5, UC Berkeley Dash, Dataset,
10.6078/D1WH41, 2018b.Laughner, J., Zhu, Q., and Cohen, R.: Berkeley High Resolution (BEHR) OMI NO2
– Gridded pixels, monthly profiles, UC Berkeley Dash, Dataset,
10.6078/D1RQ3G, 2018c.Laughner, J., Zhu, Q., and Cohen, R.: Berkeley High Resolution (BEHR) OMI NO2
– Native pixels, monthly profiles, UC Berkeley Dash, Dataset,
10.6078/D1N086, 2018d.Laughner, J. L.: AutoWRFChem-Base v1.2.0: Automation for the WRF-Chem model,
Zenodo, 10.5281/zenodo.839040, 2017.Laughner, J. L. and Cohen, R. C.: Quantification of the effect of modeled
lightning NO2 on UV-visible air mass factors, Atmos. Meas. Tech., 10,
4403–4419, 10.5194/amt-10-4403-2017, 2017.Laughner, J. L. and Zhu, Q.: Code updated in response to ESSD first reviews,
Zenodo, 10.5281/zenodo.1450747, 2018a.Laughner, J. L. and Zhu, Q.: CohenBerkeleyLab/BEHR-Core: BEHR Core code,
Zenodo, 10.5281/zenodo.998275, 2018b.Laughner, J. L., Zare, A., and Cohen, R. C.: Effects of daily meteorology on
the interpretation of space-based remote sensing of NO2, Atmos. Chem.
Phys., 16, 15247–15264, 10.5194/acp-16-15247-2016, 2016.Laughner, J. L., Zhu, Q., and Cohen, R.: Evaluation of version 3.0B of the
BEHR OMI NO2 product, Atmos. Meas. Tech. Discuss.,
10.5194/amt-2018-248, in review, 2018e.Levelt, P., van der Oord, G., Dobber, M., Mälkki, A., Visser, H.,
de Vries, J., Stammes, P., Lundell, J., and Saari, H.: The Ozone Monitoring
Instrument, IEEE T. Geosci. Remote, 44, 1093–1101,
10.1109/TGRS.2006.872333, 2006.Lin, J.-T., Liu, M.-Y., Xin, J.-Y., Boersma, K. F., Spurr, R., Martin, R.,
and Zhang, Q.: Influence of aerosols and surface reflectance on satellite
NO2 retrieval: seasonal and spatial characteristics and implications
for NOx emission constraints, Atmos. Chem. Phys., 15,
11217–11241, 10.5194/acp-15-11217-2015, 2015.Liu, F., Beirle, S., Zhang, Q., Dörner, S., He, K., and Wagner, T.:
NOx lifetimes and emissions of cities and power plants in
polluted background estimated by satellite observations, Atmos. Chem. Phys.,
16, 5283–5298, 10.5194/acp-16-5283-2016, 2016.Liu, F., Beirle, S., Zhang, Q., van der A, R. J., Zheng, B., Tong, D., and
He, K.: NOx emission trends over Chinese cities estimated
from OMI observations during 2005 to 2015, Atmos. Chem. Phys., 17,
9261–9275, 10.5194/acp-17-9261-2017, 2017.
Liu, Y., Bourgeois, A., Warner, T., Swerdlin, S., and Hacker, J.:
Implementation of the observation-nudging based on FDDA into WRF for
supporting AFEC test operations, 6th WRF Conference, NCAR, Boulder, CO, USA,
2006.Lorente, A., Folkert Boersma, K., Yu, H., Dörner, S., Hilboll, A.,
Richter, A., Liu, M., Lamsal, L. N., Barkley, M., De Smedt, I., Van
Roozendael, M., Wang, Y., Wagner, T., Beirle, S., Lin, J.-T., Krotkov, N.,
Stammes, P., Wang, P., Eskes, H. J., and Krol, M.: Structural uncertainty in
air mass factor calculation for NO2 and HCHO satellite retrievals,
Atmos. Meas. Tech., 10, 759–782, 10.5194/amt-10-759-2017, 2017.Lu, Z., Streets, D. G., de Foy, B., Lamsal, L. N., Duncan, B. N., and Xing,
J.: Emissions of nitrogen oxides from US urban areas: estimation from Ozone
Monitoring Instrument retrievals for 2005–2014, Atmos. Chem. Phys., 15,
10367–10383, 10.5194/acp-15-10367-2015, 2015.Marais, E. A., Jacob, D. J., Choi, S., Joiner, J., Belmonte-Rivas, M., Cohen,
R. C., Beirle, S., Murray, L. T., Schiferl, L., Shah, V., and Jaeglé, L.:
Nitrogen oxides in the global upper troposphere: interpreting cloud-sliced
NO2 observations from the OMI satellite instrument, Atmos. Chem.
Phys. Discuss., 10.5194/acp-2018-556, in review, 2018.Marchenko, S., Krotkov, N. A., Lamsal, L. N., Celarier, E. A., Swartz, W. H.,
and Bucsela, E. J.: Revising the slant column density retrieval of nitrogen
dioxide observed by the Ozone Monitoring Instrument, J. Geophys. Res.-Atmos.,
120, 5670–5692, 10.1002/2014JD022913, 2015.Martin, R., Sauvage, B., Folkins, I., Sioris, C., Boone, C., Bernath, P., and
Ziemke, J.: Space-based constraints on the production of nitric oxide by
lightning, J. Geophys. Res., 112, D09309, 10.1029/2006JD007831, 2007.McKenzie, R., Johnstone, P., McElroy, C., Kerr, J., and Solomon, S.: Altitude
distributions of stratospheric constituents from ground-based measurements at
twilight, J. Geophys. Res., 96, 15499–15511, 10.1029/91JD01361, 1991.McLinden, C. A., Fioletov, V., Boersma, K. F., Kharol, S. K., Krotkov, N.,
Lamsal, L., Makar, P. A., Martin, R. V., Veefkind, J. P., and Yang, K.:
Improved satellite retrievals of NO2 and SO2 over the
Canadian oil sands and comparisons with surface measurements, Atmos. Chem.
Phys., 14, 3637–3656, 10.5194/acp-14-3637-2014, 2014.Mebust, A. and Cohen, R.: Observations of a seasonal cycle in
NOx emissions from fires in African woody savannas, Geophys.
Res. Lett., 40, 1451–1455, 10.1002/grl.50343, 2013.Mebust, A. K. and Cohen, R. C.: Space-based observations of fire
NOx emission coefficients: a global biome-scale comparison,
Atmos. Chem. Phys., 14, 2509–2524, 10.5194/acp-14-2509-2014, 2014.Mebust, A. K., Russell, A. R., Hudman, R. C., Valin, L. C., and Cohen, R. C.:
Characterization of wildfire NOx emissions using MODIS fire
radiative power and OMI tropospheric NO2 columns, Atmos. Chem. Phys.,
11, 5839–5851, 10.5194/acp-11-5839-2011, 2011.Menzel, D. B.: Ozone: An overview of its toxicity in man and animals,
J. Toxicol. Env. Health, 13, 181–204, 10.1080/15287398409530493, 1984.Miyazaki, K., Eskes, H. J., and Sudo, K.: Global NOx
emission estimates derived from an assimilation of OMI tropospheric
NO2 columns, Atmos. Chem. Phys., 12, 2263–2288,
10.5194/acp-12-2263-2012, 2012.Miyazaki, K., Eskes, H. J., Sudo, K., and Zhang, C.: Global lightning
NOx production estimated by an assimilation of multiple
satellite data sets, Atmos. Chem. Phys., 14, 3277–3305,
10.5194/acp-14-3277-2014, 2014.Miyazaki, K., Eskes, H., Sudo, K., Boersma, K. F., Bowman, K., and Kanaya,
Y.: Decadal changes in global surface NOx emissions from
multi-constituent satellite data assimilation, Atmos. Chem. Phys., 17,
807–837, 10.5194/acp-17-807-2017, 2017.
Myhre, G., Shindle, D., Bréon, F.-M., Collins, W., Fuglestvedt, J.,
Huang, J., Koch, D., Lamarque, J.-F., Lee, D., Mendoza, B., Nakajima, T.,
Robock, A., Stephens, G., Takemura, T., and Zhang, H.: Climate Change 2013:
The Physical Science Basis, Contribution of Working Group I to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change, edited
by: Stocker, T. F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S. K.,
Boschung, J., Nauels, A., Xia, Y., Bex, V., and Midgley, P. M., Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA, chap.:
Anthropogenic and Natural Radiative Forcing, 659–684, 2013.Nault, B. A., Garland, C., Pusede, S. E., Wooldridge, P. J., Ullmann, K.,
Hall, S. R., and Cohen, R. C.: Measurements of CH3O2NO2 in the
upper troposphere, Atmos. Meas. Tech., 8, 987–997,
10.5194/amt-8-987-2015, 2015.Nault, B. A., Garland, C., Wooldridge, P. J., Brune, W. H., Campuzano-Jost,
P., Crounse, J. D., Day, D. A., Dibb, J., Hall, S. R., Huey, L. G., Jimenez,
J. L., Liu, X., Mao, J., Mikoviny, T., Peischl, J., Pollack, I. B., Ren, X.,
Ryerson, T. B., Scheuer, E., Ullmann, K., Wennberg, P. O., Wisthaler, A.,
Zhang, L., and Cohen, R. C.: Observational Constraints on the Oxidation of
NOx in the Upper Troposphere, J. Phys. Chem. A, 120,
1468–1478, 10.1021/acs.jpca.5b07824, 2016.Nault, B. A., Laughner, J. L., Wooldridge, P. J., Crounse, J. D., Dibb, J.,
Diskin, G., Peischl, J., Podolske, J. R., Pollack, I. B., Ryerson, T. B.,
Scheuer, E., Wennberg, P. O., and Cohen, R. C.: Lightning
NOx Emissions: Reconciling Measured and Modeled Estimates
With Updated NOx Chemistry, Geophys. Res. Lett., 44,
9479–9488, 10.1002/2017GL074436, 2017.Noel, S., Bovensmann, H., Burrows, J. P., Frerick, J., Chance, K. V., Goede,
A. P. H., and Muller, C.: SCIAMACHY instrument on ENVISAT-1, in: Sensors,
Systems, and Next-Generation Satellites II, edited by: Fujisada, H., Proc.
SPIE, 3498, 10.1117/12.333621, 1998.
Palmer, P., Jacob, D., Chance, K., Martin, R., Spurr, R., Kurosu, T., Bey,
I., Yantosca, R., Fiore, A., and Li, Q.: Air mass factor formulation for
spectroscopic measurements from satellites: Applications to formaldehyde
retrievals from the Global Ozone Monitoring Experiment, J. Geophys. Res.,
106, 14539–14550, 2001.Pandis, S. N., Harley, R. A., Cass, G. R., and Seinfeld, J. H.: Secondary
organic aerosol formation and transport, Atmos. Environ. A-Gen., 26,
2269–2282, 10.1016/0960-1686(92)90358-R, 1992.Parker, L., Kemball-Cook, S., and Yarwood, G.: Final Report Hood County
NOx Trends, Tech. rep., contract #582-16-60185, available
at:
http://www.hoodcountycleanair.com/userfiles/file/HoodCountyNOxTrends_06Oct2017.pdf
(last access: 14 November 2018), 2017.Pickering, K. E., Bucsela, E., Allen, D., Ring, A., Holzworth, R., and
Krotkov, N.: Estimates of lightning NOx production based on
OMI NO2 observations over the Gulf of Mexico, J. Geophys.
Res.-Atmos., 121, 8668–8691, 10.1002/2015JD024179, 2016.Platnick, S., King, M., Wind, G., Ackerman, S., Menzel, P., and Frey, R.:
MODIS/Aqua Clouds 5-Min L2 Swath 1 km and 5 km, NASA MODIS Adaptive
Processing System, Goddard Space Flight Center, USA,
10.5067/MODIS/MYD06_L2.006, 2015.Pope, C. A., Ezzati, M., and Dockery, D. W.: Fine-Particulate Air Pollution
and Life Expectancy in the United States, N. Engl. J. Med., 360, 376–386,
10.1056/nejmsa0805646, https://doi.org/10.1056/nejmsa0805646,
2009.Pusede, S. E., Duffey, K. C., Shusterman, A. A., Saleh, A., Laughner, J. L.,
Wooldridge, P. J., Zhang, Q., Parworth, C. L., Kim, H., Capps, S. L., Valin,
L. C., Cappa, C. D., Fried, A., Walega, J., Nowak, J. B., Weinheimer, A. J.,
Hoff, R. M., Berkoff, T. A., Beyersdorf, A. J., Olson, J., Crawford, J. H.,
and Cohen, R. C.: On the effectiveness of nitrogen oxide reductions as a
control over ammonium nitrate aerosol, Atmos. Chem. Phys., 16, 2575–2596,
10.5194/acp-16-2575-2016, 2016.
Richter, A. and Wagner, T.: The Use of UV, Visible and Near IR Solar Back
Scattered Radiation to Determine Trace Gases, in: The Remote Sensing of
Tropospheric Composition from Space, edited by Burrows, J., Platt, U., and
Borrell, P., Springer, New York, 2011.Rollins, A. W., Browne, E. C., Min, K.-E., Pusede, S. E., Wooldridge, P. J.,
Gentner, D. R., Goldstein, A. H., Liu, S., Day, D. A., Russell, L. M., and
Cohen, R. C.: Evidence for NOx Control over Nighttime SOA
Formation, Science, 337, 1210–1212, 10.1126/science.1221520, 2012.Roujean, J.-L., Leroy, M., and Deschamps, P.-Y.: A bidirectional reflectance
model of the Earth's surface for the correction of remote sensing data,
J. Geophys. Res., 97, 20455–20468, 10.1029/92JD01411, 1992.Russell, A. R., Perring, A. E., Valin, L. C., Bucsela, E. J., Browne, E. C.,
Wooldridge, P. J., and Cohen, R. C.: A high spatial resolution retrieval of
NO2 column densities from OMI: method and evaluation, Atmos. Chem.
Phys., 11, 8543–8554, 10.5194/acp-11-8543-2011, 2011.Russell, A. R., Valin, L. C., and Cohen, R. C.: Trends in OMI NO2
observations over the United States: effects of emission control technology
and the economic recession, Atmos. Chem. Phys., 12, 12197–12209,
10.5194/acp-12-12197-2012, 2012.Schaaf, C. Z. W.: MCD43D07 MODIS/Terra+Aqua BRDF/Albedo Parameter1 Band3
Daily L3 Global 30ArcSec CMG V006, NASA EOSDIS Land Processes DAAC,
10.5067/modis/mcd43d07.006, 2015a.Schaaf, C. Z. W.: MCD43D08 MODIS/Terra+Aqua BRDF/Albedo Parameter2 Band3
Daily L3 Global 30ArcSec CMG V006, NASA EOSDIS Land Processes DAAC,
10.5067/modis/mcd43d08.006, 2015b.Schaaf, C. Z. W.: MCD43D09 MODIS/Terra+Aqua BRDF/Albedo Parameter3 Band3 Daily
L3 Global 30ArcSec CMG V006, NASA EOSDIS Land Processes DAAC,
10.5067/modis/mcd43d09.006, 2015c.Schaaf, C. Z. W.: MCD43D31 MODIS/Terra+Aqua BRDF/Albedo QA BRDFQuality Daily
L3 Global 30ArcSec CMG V006, NASA EOSDIS Land Processes DAAC,
10.5067/modis/mcd43d31.006, 2015d.Schreier, S. F., Richter, A., Kaiser, J. W., and Burrows, J. P.: The
empirical relationship between satellite-derived tropospheric NO2 and
fire radiative power and possible implications for fire emission rates of
NOx, Atmos. Chem. Phys., 14, 2447–2466,
10.5194/acp-14-2447-2014, 2014Schütt, A.: Improved Gridding Routine of OMI NO2 data and its
applications, Master's thesis, Ludwig-Maximilians-University Munich, 2017.Schwantes, R. H., Teng, A. P., Nguyen, T. B., Coggon, M. M., Crounse, J. D.,
St. Clair, J. M., Zhang, X., Schilling, K. A., Seinfeld, J. H., and Wennberg,
P. O.: Isoprene NO3 Oxidation Products from the RO2+HO2 Pathway, J. Phys. Chem. A, 119,
10158–10171,10.1021/acs.jpca.5b06355, 2015.Slusser, J., Hammond, K., Kylling, A., Stamnes, K., Perliski, L., Dahlback,
A., Anderson, D., and DeMajistre, R.: Comparison of air mass computations,
J. Geophys. Res., 101, 9315–9321, 10.1029/96JD00054, 1996.Stahler, A., Lucht, W., Schaaf, C., Tsang, T., Gao, F., Li, X., Muller,
J.-P., Lewis, P., Barnsley, M., Strugnell, N., Hu, B., Hyman, A.,
d'Entremont, R., Chen, L., Liu, Y., McIver, D., Liang, S., Disney, M.,
Hobson, P., Dunderdale, M., and Roberts, G.: MODIS BRDF/Albedo Product:
Algorithm Theoretical Basis Document Version 5.0, 7–15, avalable at:
https://lpdaac.usgs.gov/sites/default/files/public/product_documentation/atbd_mod09_v5.pdf
(last access: 14 November 2018), 1999.Travis, K. R., Jacob, D. J., Fisher, J. A., Kim, P. S., Marais, E. A., Zhu,
L., Yu, K., Miller, C. C., Yantosca, R. M., Sulprizio, M. P., Thompson, A.
M., Wennberg, P. O., Crounse, J. D., St. Clair, J. M., Cohen, R. C.,
Laughner, J. L., Dibb, J. E., Hall, S. R., Ullmann, K., Wolfe, G. M.,
Pollack, I. B., Peischl, J., Neuman, J. A., and Zhou, X.: Why do models
overestimate surface ozone in the Southeast United States?, Atmos. Chem.
Phys., 16, 13561–13577, 10.5194/acp-16-13561-2016, 2016.Valin, L., Russell, A., and Cohen, R.: “Variations of OH radical in an urban
plume inferred from NO2 column measurements”, Geophys. Res. Lett.,
40, 1856–1860, 10.1002/grl.50267, 2013.NOx
van der A, R. J., Eskes, H. J., Boersma, K. F., van Noije, T. P. C.,
Van Roozendael, M., De Smedt, I., Peters, D. H. M. U., and Meijer, E. W.:
Trends, seasonal variability and dominant NOx source derived
from a ten year record of NO2 measured from space, J. Geophys. Res.,
113, D04302, 10.1029/2007JD009021, 2008.van Marle, M. J. E., Kloster, S., Magi, B. I., Marlon, J. R., Daniau, A.-L.,
Field, R. D., Arneth, A., Forrest, M., Hantson, S., Kehrwald, N. M., Knorr,
W., Lasslop, G., Li, F., Mangeon, S., Yue, C., Kaiser, J. W., and van der
Werf, G. R.: Historic global biomass burning emissions for CMIP6 (BB4CMIP)
based on merging satellite observations with proxies and fire models
(1750–2015), Geosci. Model Dev., 10, 3329–3357,
10.5194/gmd-10-3329-2017, 2017.Vasilkov, A., Qin, W., Krotkov, N., Lamsal, L., Spurr, R., Haffner, D.,
Joiner, J., Yang, E.-S., and Marchenko, S.: Accounting for the effects of
surface BRDF on satellite cloud and trace-gas retrievals: a new approach
based on geometry-dependent Lambertian equivalent reflectivity applied to OMI
algorithms, Atmos. Meas. Tech., 10, 333–349, 10.5194/amt-10-333-2017,
2017.Veefkind, J., Aben, I., McMullan, K., Förster, H., de Vries, J., Otter, G.,
Claas, J., Eskes, H., de Haan, J., Kleipool, Q., van Weele, M., Hasekamp, O.,
Hoogeveen, R., Landgraf, J., Snel, R., Tol, P., Ingmann, P., Voors, R.,
Kruizinga, B., Vink, R., Visser, H., and Levelt, P.: TROPOMI on the ESA
Sentinel-5 Precursor: A GMES mission for global observations of the
atmospheric composition for climate, air quality and ozone layer
applications, Remote Sens. Environ., 120, 70–83,
10.1016/j.rse.2011.09.027, 2012.Wanner, W., Li, X., and Strahler, A. H.: On the derivation of kernels for
kernel-driven models of bidirectional reflectance, J. Geophys. Res., 100,
21077–21089, 10.1029/95JD02371, 1995.
Wegmann, M., Fehrenbach, A., Heimann, S., Fehrenbach, H., Renz, H., Garn, H.,
and Herz, U.: NO2-induced airway inflammation is associated with
progressive airflow limitation and development of emphysema-like lesions in
C57BL/6 mice, Exp. Toxicol. Pathol, 56, 341–350,
10.1016/j.etp.2004.12.004, 2005.Zare, A., Romer, P. S., Nguyen, T., Keutsch, F. N., Skog, K., and Cohen, R.
C.: A comprehensive organic nitrate chemistry: insights into the lifetime of
atmospheric organic nitrates, Atmos. Chem. Phys., 18, 15419–15436,
10.5194/acp-18-15419-2018, 2018.Zhou, Y., Brunner, D., Boersma, K. F., Dirksen, R., and Wang, P.: An improved
tropospheric NO2 retrieval for OMI observations in the vicinity of
mountainous terrain, Atmos. Meas. Tech., 2, 401–416,
10.5194/amt-2-401-2009, 2009.
Ziemke, J., Chandra, S., and Bhartia, P.: Cloud slicing: A new technique to
derive upper tropospheric ozone from satellite measurements, J. Geophys.
Res., 106, 9853–9867, 2001.Zörner, J., Penning de Vries, M., Beirle, S., Sihler, H., Veres, P. R.,
Williams, J., and Wagner, T.: Multi-satellite sensor study on
precipitation-induced emission pulses of NOx from soils in
semi-arid ecosystems, Atmos. Chem. Phys., 16, 9457–9487,
10.5194/acp-16-9457-2016, 2016.