Investigation Overview
Deriving Information on Surface Conditions from Column
and VERtically Resolved Observations Relevant to Air Quality
A NASA Earth Venture campaign intended to improve
the interpretation of satellite observations to diagnose
near-surface conditions relating to air quality
Objectives:
1. Relate column observations to surface conditions for
aerosols and key trace gases O3, NO2, and CH2O
2. Characterize differences in diurnal variation of surface and
column observations for key trace gases and aerosols
3. Examine horizontal scales of variability affecting satellites
and model calculations
Deployment Strategy
Systematic and concurrent observation of column-integrated, surface, and
vertically-resolved distributions of aerosols and trace gases relevant to air quality
as they evolve throughout the day.
Three major observational
components:
NASA UC-12 (Remote sensing)
Continuous mapping of aerosols
with HSRL and trace gas columns
with ACAM
NASA P-3B (in situ meas.)
In situ profiling of aerosols and
trace gases over surface
measurement sites
Ground sites
In situ trace gases and aerosols
Remote sensing of trace gas and
aerosols (Pandora, AERONET)
Ozonesondes
Aerosol lidar observations
Tethersondes, NO2 sondes
Deployment Locations
Maryland, July 2011
California, Jan-Feb 2013
Houston, September 2013
Colorado, Jul-Aug 2014
Flight Hours and Sampling Statistics
Location
P-3B Hours
(Flight Days)
King Air Hours (Flight
days)
Other aircraft
participation
104 (14)
101 (14)
UMD Cessna
84 (10)
10 (2) - PODEX
70 (10)
NASA ER-2
Alpha Jet
UC-Davis Mooney
Houston
71 (9)
68 (9)
23 (3) - GEO-CAPE
NASA LaRC Falcon
Denver, Northern Front
Range
93 (16)
89 (17)
NCAR C-130
NASA LaRC Falcon
Total for all campaigns
352 (49)
328 (50)
Baltimore-Washington
California, Central
Valley
Location
P-3B Spirals
King Air Overflights
Missed Approaches
Baltimore-Washington
254
342
0
California, Central Valley
170
307
166
Houston
195
341
105
Denver, Northern Front Range
214
451
108
Total for all campaigns
833
1441
379
Aerosol Characterization
Fresno
Examples of surface PM2.5, Aeronet AOD,
P-3B in situ profiles of scattering, and HSRL
observations during the California campaign.
10000
9000
PM2.5 (ug/cm3)
8000
7000
AOD
6000
5000
4000
3000
2000
Bakersfield
1000
Day in January
0
0
100
200
300
550 nm Scattering (Mm-1)
Trace Gas Characterization
Examples of Pandora column-integrated observations NO2
and in situ profiles by the P-3B from the Colorado deployment
Comparison of ACAM and P-3B CH2O
(Maryland deployment)
Ozonesondes
In addition to traditional ozonesondes, partners
also launched tethersondes for O3 (NOAA, Bryan
Johnson) and NO2 (KNMI, Deb Stein-Zweers)
Tethersondes
28 July
0750-1030 LT
NO2
O3
Final Disposition of Data
A Digital Object Identifier has been obtained for the DISCOVER-AQ data to enable data
discovery as well as attribution and referencing in publications:
doi:10.5067/Aircraft/DISCOVER-AQ/Aerosol-TraceGas
This doi points to the Langley ASDC, the long-term archival site for the DISCOVER-AQ
data (https://eosweb.larc.nasa.gov/project/discover-aq/discover-aq_table)
Summary of Column vs. Surface
Correlations for NO2 and O3
Maryland
California
NO2
O3
NO2
O3
P-3B in situ
Low
High
P-3B in situ
Low
Low
P-3B + surface
Moderate
High
P-3B + surface
Moderate
Low
Pandora
Moderate
Low
Pandora
Low
Low
CMAQ
Moderate
Moderate
CMAQ
Moderate
Low
Texas
Colorado
NO2
O3
NO2
O3
P-3B in situ
Low
Low
P-3B in situ
Low
N.S.
P-3B + surface
Moderate
Low
P-3B + surface
Low
N.S.
Pandora
Low
N.S.
Pandora
High
High
CMAQ
High
Low
CMAQ
High
Moderate
Low correlation: R < 0.4; Moderate: R=0.4-0.8; High: R=0.8-1.0; N.S. = not significant
Clare Flynn, U. Maryland
Surface vs Column O3 at BAO
(Erie, CO): All DAQ/FRAPPE days
N = 2294
ΔO3 ≤ 10 ppbv:
66% of all cases
08 09 10 11 12 13 14 15 16 17 18
MDT
66% of 1500 m AGL column and surface O3 values agree within 10 ppbv, almost entirely after midday,
showing lower tropospheric column O3 to be a good predictor of surface conditions.
34% exhibit O3 column exceeding surface values due to shallow mixing in the morning, elevated
plumes, and thunderstorm outflows leading to significant vertical gradients of ozone in the lower
atmosphere
Christoph Senff, NOAA TOPAZ Lidar
Relationships Between AOD
and Surface PM2.5
Crumeyrolle et al. (2014) showed that AOD as measured
by vertical integration of the extinction from the P-3B
was well correlated with the product of surface PM2.5,
the height of the buffer (BuL) layer, and a hydration factor.
The BuL is the transition layer between the BL and the FT
with a pronounced gradient of aerosol concentrations from
the typically higher concentration observed in the BL and
typically cleaner conditions observed in the FT.
Chu et al., 2015 proposed the following relationships for predicting surface PM2.5 given AOD observations
from satellite, HSRL, AERONET, etc.:
ta :AOD at 0.55 mm
f(RH): hydration factor (function of relative humidity)
Aerosol extinction cross section per
unit dry mass
HLH: Haze layer height = top of BuL
MPE: mean PBL extinction
Errors in AMFs/retrievals from a-priori NO2 profiles
Lok Lamsal
8 km
AMF 
w  f
i
i
surface
wi scattering weights
fi NO2 shape factors
► AMF calculated for ACAM (air-borne spectrometer located at
~8km) but is also relevant for tropospheric NO2 column
retrievals
Padonia, MD Retrieval errors based on 11 AM campaign mean profiles
Surface reflectivities: 0.1 to 0.14 at 0.01 steps
Solar zenith angles: 10° to 80 at 10°steps
Aerosol optical depths: 0.1 to 0.9 at 0.1 steps
NO2 profiles and AMFs: Summary plot
NO2 tropospheric column retrievals are highly sensitive to
diurnal variation in a-priori profile shapes.
L. Lamsal
Spatial and Temporal Variability
Follette-Cook et al. (2015) Atmos. Environ.
• Quantified and compared the spatial and temporal variability
seen in the Maryland/DC DISCOVER-AQ P-3B trace gas data
and in a regional WRF/Chem simulation
• Compared variability with TEMPO/GEO-CAPE precision
requirements
• Found that the variability simulated for the July 2011
campaign period was large enough to be meaningfully
resolved by a TEMPO-type geostationary instrument
Additional analyses:
• Comparing the variability observed in MD campaign with the
variability observed during the other DISCOVER-AQ campaigns
Geographic coverage of differences that fall
above the PRs at several distances
O3 – 0-2 km
CO - 0 – 2 km
These results indicate that the TEMPO
instrument would be able to observe O3
air quality events over the Mid-Atlantic
area, even on days when the violations of
the air quality standard are not
widespread.
PR = 1.7 DU
PR = 0.91 e17 molec/cm2
7/29/2011 2 pm EDT
M. Follette-Cook
Geographic coverage of differences that fall
above the PRs at several distances
NO2 - Tropospheric column
PR = 1x1015 molec/cm2
7/29/2011 2 pm EDT
• The maximum differences in NO2 are in
the morning and early evening, so 2
pm could be considered a minimum in
what would be observable
• Despite that, the major features in the
tropospheric column field can be seen
in the plot of the visible differences at
4 – 8 km distances
• The entire field is almost visible at
distances greater than 20 km
• This suggests that TEMPO will be able
to satisfactorily observe the spatial
variability of this species
Percentiles – O3
TEMPO PR
MD, TX, and CO 75th percentile
curves above the PR at distances of
8 (TX), 12 (CO), and 16 km (MD)
All campaigns 95th percentile
curves above the PR
MD
CA
TX
CO
Percentiles – NO2
• MD shows the lowest NO2
variability of any campaign
• Follette-Cook et al. (2015)
concluded that the variability
during the MD campaign was
large enough to be wellresolved by TEMPO
• Therefore, with respect to
variability, TEMPO’s NO2 PR is
more than sufficient for these
other regions as well
MD
CA
TX
CO
Surface Ozone over Bay vs Land
Possible reasons:
D. Goldberg, UMD
Stratospheric Intrusions during DISCOVER-AQ
Lesley Ott, Bryan Duncan, Anne Thompson
21
Profiles of Aerosol Absorption Suggest
Model/AERONET Over-predictions
LARGE and AERONET Teams
• Absorption aerosol optical depth
(AAOD) was calculated from in-situ
measurements for each vertical profile
• Comparisons with AERONET yielded
significantly lower in-situ AAOD values at
each of the DISCOVER-AQ urban sites
• This discrepancy is compounded by model
overestimates in remote regions by factors
of 2-5 (Schwarz et al., 2013)
• No evidence was observed to support
model up-scaling by Bond et al. (2013)
Key Scientific Outcomes
•
Surface-Column correlations are a mixed bag, but there is indication that lower
tropospheric ozone (especially in the afternoon) can sometimes be trusted to infer
surface values.
•
Spatial variability analyses show that TEMPO precision requirements are adequate to
detect air quality violations.
•
Model representation of diurnal evolution in profile structure is expected to be an
important source of error for TEMPO retrievals.
•
Pandora validation of TEMPO diurnal variability would appear to require no more than
a couple of instruments in a given metropolitan area.
•
Formaldehyde shows promise as a proxy for ozone production.
•
Validation of AERONET AAOD reveals a high bias that may be responsible for
overprediction of black carbon by global models.
•
Strong correlations are observed between AOD and PM2.5, but the scaling factors vary
widely based on BL depth, humidity, and aerosol composition (hygroscopicity).
•
Reactive nitrogen chemistry in air quality models needs to be improved.
•
Ozone can be highest over unmonitored areas downwind (e.g., Chesapeake Bay).
•
A single lidar may be sufficient to characterize the vertical distribution of aerosols
over distances of up to 100 km.