Didier Rault NASA Langley Research Center

1

SAGE III, OSIRIS and beyond OMPS

• Didier Rault (NASA Langley Research
  Center)
• Svetlana Petelina (University of
  Saskatchewan)
• Doug Degenstein (University of
  Saskatchewan)
• 3rd International Atmospheric Limb Workshop
• Montréal, CanadaApril 25-28, 2006

2
Outline

• SAGE III Limb Scatter products status
• ? Ozone
• - Accuracy (5-10 from tropopause to
  45km)
• - Precision (5-10 from 20 to 40 km)
• - Height registration error on the order
  of 350m RMS.
• ? NO2
• ? Stratospheric aerosol
• SAGE III LS data set currently includes over
  10000 scans, and contains both
• level 1 (limb scatter radiance profiles) and
  level 2 (retrieved ozone profiles) data
• Working on posting level 1 and level 2 (O3, NO2,
  aerosol, albedo) data on Web
• SAGE III being used for OMPS algorithm check up
  (including correlations with SAGE II)
• SAGE III OSIRIS being used for OMPS algorithm
  check up

3

SAGE III mission status

• Spacecraft failure on March 8, 2006 End of
  SAGE III mission
• 2500 Limb Scatter events (4-10 scans) were
  collected, total
• Mission wrapping up (Occultation, Limb Scatter)
  next few months

4
SAGE III LSmeasurement modes
(1)  Precision To make repeated measurements
over same geolocations on consecutive orbits and
consecutive days. (Tropical ozonesondes stations
Lidars) (2) Vortex (3) Global

5

Validation methodology

• Coincidence criteria
• Latitude3o, Longitude 10o, Time
  same day, Delta Total Ozone 20DU
• For ensemble of coincident measurements,
  evaluate
• - Mean profiles
• - Bias between mean profiles ? measurement
  accuracy
• - Standard deviation from mean ?
  measurement precision
• Tangent height registration
• Two methods Stretch and shift
• Maximum correlation

6
Comparison SAGE III LS vs SAGE II
7
Comparison SAGE III LS vs SAGE III
8
Comparison SAGE III LS vs OSIRIS
9
Comparison SAGE III LS vs RAQMS
10
Sensitivity of Ozone retrieval to Height offset,
aerosol, NO2, albedo
11
Aerosol retrievalComparison with SAGE II
521 nm
1020 nm
12
SAGE III Ozone/NO2 retrieval
precision estimate
Ozone Precision
NO2 Precision
13

Aerosol retrieval. Precision

14
OMPS Limb profiler

• Ozone Mapping Profile Suite (OMPS) mission on NPP
  and NPOESS satellites. OMPS is comprised of 3
  sensors
• Nadir total ozone sensor (TOMS/OMI
  like)
• Nadir profiler (SBUV like)
• Limb profiler (new sensor)
• To meet CDR requirements, OMPS needs 10
  accuracy, 3 precision and 3-km vertical
  resolution over the critical 15-50 km altitude
  range.
• Major issues
• Straylight (large dynamic range across
  limb)
• Altitude registration
• Scene inhomogeneity
• Ozone horizontal gradients

15
OMPS LP algorithm

• Limited number of pixels
• 21 binned pixels from 290 nm to 1000
  nm, with bandpass 1.5nm to 40nm
• Possibility to download 82 unbinned
  pixels
• Ozone retrieval method
• Series of triplets in Chappuis
• Series of doublets in Huggins / Hartley
• Other retrieved quantities
• Aerosol (400, 525, 675, 1000nm)
• Surface reflectivities (and cloud
  fraction), using absolute calibration
• Cloud height (using long wavelength)
• Height registration
• Knee (using Nadir TC), RSAS
• Forward model
• U of Arizona with look up tables
• - Multiple scatter correction f( ozone
  profile, SZA, AZ)
• - Monochromatic effect

Coupled 1 profile
16
Using SAGE III LS radiances to check OMPS
algorithm

Concept

• Construct proxy data set
• Check individual modules
• Check ozone retrieval case-by-case, ensemble,
  SAGE II

Proxy ? FM(?) . DATA(pix) / FM(pix) .
bandpass(?) d? and SNR
OMPS 21 pix SAGE 340 pix
Compare ozone profiles
SAGE III LS OMPS
SAGE II
(triplet, MLR)
(SAGEIII LS)
17
Comparison SAGE III LS with OMPS (SAGE III)
133 profiles July 17 30, 2004
18
Statistical analysis II

• Comparison with SAGE II
  occultations

Solar Zenith angles from 40 to 81
to Scattering angles from 36 to 140 Surface
albedo from 0 to 0.4
19
Using OSIRIS radiances to check OMPS
algorithm

Concept

• Construct proxy data set
• Consider LS events with OSIRIS SAGE III LS
  collocations
• Check ozone retrieval case-by-case, ensemble

Proxy ? DATA(?) . bandpass(?) d? and SNR
OMPS 21 pix OSIRIS 1353 pix
Compare ozone profiles
SAGE III LS OSIRIS
OMPS OMPS
(triplet, MLR)
(triplet, global)
(OSIRIS data)
(SAGEIII LS data)
20
Absolute calibrationSAGE III LS vs OSIRIS

Variation as function of Zenith and
azimuth angles
21
SAGEIII/OSIRIS

Triplets
Degenstein
22
Comparison OSIRIS (triplets) with OMPS (OSIRIS)
No UV

23
Comparison OSIRIS (Degenstein) with OMPS (OSIRIS)
37 profiles (July 17-30, 2004) Assumed albedo
0.05
Ozone mean profiles
Standard deviation
Ozone mean bias
24
OSIRIS Retrievals Triplet vs
Degenstein
  .  
25
Conclusion

• SAGE III LS has demonstrated capability
• ? Ozone
• - Accuracy (5-10 from tropopause to
  45km)
• - Precision (5-10 from 20 to 40 km)
• - Height registration error on the order
  of 350m RMS
• ? NO2 (25 precision)
• ? Stratospheric aerosol (15-20 for ? lt 800nm,
  30-40 above)
• More to be done, using LS data collected in past
  year (precision,
• vortex, global)
• SAGE III LS and OSIRIS data set ( level 1
  radiances and products) are being used to tune up
  OMPS LS algorithm

26
Comparison of SAGE II solar occultation ozone
with other instruments
Relative ozone differences between SAGE II and
correlative data set in N.H.
  .  
27
Vertical resolution
40
  Individual ozone profiles from SAGE III LS
(black) and ozonesonde (red)
 
30
FWHM of averaging kernel A Cov . KT . Se . K
20
  .  
  .  
10
Accounts for FOV 0.5km and smearing (1/16s)
28
Aerosol, albedo, cloud from SAGE III LS

• Aerosol residual data over model
  (520,600,670,750,870,1020nm)
• Surface albedo matching model to data above 35km
  (520,600,670,750,870,1020nm)
• Cloud height Structure in radiance profiles
  (750,870,1020nm)

Vertical structure
29
Ozone retrieval methodologies
MLR method
Triplet / Singlet method .Series of
Triplets in Chappuis band and Singlets in UV
.Forward model U of Arizona code, modified to
perform fast high spectral resolution simulations
(no look up tables) .SAGE III instrument
model .C. Rodgers optimal estimation
pseudo
Retrieved with UV
Retrieved with visible
Ozone retrieved only above cloud, no lower than
10 km
MLR advantage residuals diagnosis
30
LaRC/UW-Madison Regional Air Quality Modeling
System (RAQMS)
Pierce, R. B. et al., J. Geophys. Res. 108,
8825, 2003

• Multi-scale chemical/dynamical modeling and data
  assimilation system to provide physically
  consistent links between regional air quality and
  global chemical composition
• Combines two meteorological forecast models
  (regional domain nested within a global domain)
  with a stratospheric-tropospheric chemical
  prediction scheme
• Simulates processes involving the long-range
  transport of trace constituents
• Resolves atmospheric structure on spatial scales
  as low as 5 km in horizontal and 200-400 m in
  vertical
• Chemical predictions are conducted online using
  instantaneous meteorological conditions to
  accurately account for large-scale advective
  processes and the exchange of trace gases via
  moist convection and boundary layer turbulence
• The unified stratosphere/troposphere chemistry
  module includes standard Ox-HOx-NOx-ClOx-BrOx
  reactions, oxidation of CH4 and CO, explicit
  isoprene chemistry, photolysis calculations and
  hydrocarbon chemistry
• Assimilation of satellite based chemical
  measurements is conducted using the statistical
  digital filter (SDF) analysis system to perform a
  univariate assimilation of stratospheric profile
  (Solar occultation SAGE, HALOE) and total column
  observations of ozone (TOMS). The SDF formalism
  is based on optimal interpolation (OI)  

31
Height registration

Radiance ratio I350(H5km) / I350(H)
Captions to be set in Times or Times New Roman or
equivalent, italic, between 18 and 24 points.
Right aligned if it refers to a figure on its
right. Caption starts right at the top edge of
the picture (graph or photo).
Data(?,H) M.simul(?,H) ?H.Dsimul(?,H)
32
OMPS Focal plane
Issues - Straylight - a/? registration -
spectral smile - dark current