Doug Degenstein

1
Yet Another NO2 Retrieval from OSIRIS
Doug Degenstein Adam Bourassa, Chris Roth, Nick
Lloyd, Dick Gattinger and Ted Llewellyn Universit
y of Saskatchewan
2
Opening Remarks There is no need to do yet
another NO2 retrieval with the OSIRIS data set.
The current, official version of the NO2
retrieval works just fine. We only wanted to try
out an idea and to see if it worked when we
compared the results to the current OSIRIS
product.
3
Opening Remarks The goal was to use the existing
SaskTran and SaskMART infrastructure to perform
NO2 retrievals. We do no real spectral
fitting. We use three triplets similar to those
used to retrieve ozone with the Chappuis
absorption feature. We use a Multiplicative
Algebraic Reconstruction Technique.
4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7
ZOOM IN
8
(No Transcript)
9
too noisy
1) 432.79 nm 435.17 nm 437.94 nm 2) 437.94
nm 439.52 nm 441.89 nm 3) 446.64 nm 448.23
nm 449.81 nm
10

• normalized at 40 km
• the vector is upside down from the usual
• one vector each triplet k
• one element for each line of sight j

11
not a bad match
minor perturbations from zero
There is plenty of sensitivity and the vectors
match not too bad
12
NO2 from a different OSIRIS scan was used to
create vectors with the model and these were
compare to the measured vectors
13
SaskMART a Multiplicative Algebraic
Reconstruction Technique based on the Maximum
Likelihood Expectation Maximization technique
used in medical imaging

• three different vectors indexed by k
• it is not just j i but lines of sight tangent
  below a grid cell are allowed to affect the
  retrieved value in that cell

14
SaskMART a Multiplicative Algebraic
Reconstruction Technique based on the Maximum
Likelihood Expectation Maximization technique
used in medical imaging
- if j i then Wji 0.7 - if j i-1 then
Wji 0.2 - if j i-2 then Wji 0.1 - all Wki
values are 1/3
15
Vector 1
Vector 3
Vector 2

• not all the structure in the OSIRIS vectors is
  captured in the modelled vectors
• all vectors dont always converge
• all in all not too bad

16

• As we go equatorward or later in local time the
  peak rises appropriately

17
Summary of NO2 Results There is no need to do
yet another NO2 retrieval with the OSIRIS data
set. The current, official version of the NO2
retrieval works just fine. We only wanted to try
out an idea and to see if it worked when we
compared the results to the current OSIRIS
product.
18

• Summary of All USask Results
• What have we done so far?
• Built a radiative transfer model
• Developed an aerosol retrieval
• Developed an ozone retrieval
• Developed an NO2 retrieval

19

• Why Build The Radiative Transfer Model?
• Our goal was to retrieve stratospheric sulphate
  aerosols
• We needed complete control of the model
• We wanted both operational and retrieval grade
  models that were consistent
• Our solar geometry is complicated
• It had to be as fast as possible
• We need to modify the model until it agreed with
  the measurements
• The model needed to be expandable to do
  two-dimensional tomographic retrievals

20

• we now have complete control over a model that
  very acurately predicts the OSIRIS signals

21

• Why Do The Aerosol Retrievals?
• UT/LS Scientists are very interested in maps of
  aerosols just above and below the tropopause

22

• Why Do The Aerosol Retrievals?
• UT/LS Scientists are very interested in maps of
  aerosols just above and below the tropopause

23

• Why Do The Ozone Retrievals?
• We wanted to see if we could extend the OSIRIS
  ozone up to 60 km with a combination of
  information from both the Hartley and Chappuis
  bands
• We wanted to see if this could be done in one
  retrieval using the MART technique developed for
  two-dimensional volume emission rate tomography

24

• Why Do The Ozone Retrievals?
• We wanted to see if we could extend the OSIRIS
  ozone up to 60 km with a combination of
  information from both the Hartley and Chappuis
  bands
• We wanted to see if this could be done in one
  retrieval using the MART technique developed for
  two-dimensional volume emission rate tomography

25

• Why Do The NO2 Retrievals?
• We wanted to see if we could do NO2 retrievals
  with only a few pieces of information
• Because Adam and Chris did such a good job of
  building the infrastructure we could set up the
  retrievals with less than one days work

26

• Why Do The NO2 Retrievals?
• We wanted to see if we could do NO2 retrievals
  with only a few pieces of information
• Because Adam and Chris did such a good job of
  building the infrastructure we could set up the
  retrievals with less than one days work

27
We believe that using lessons learned during the
design, build, characterization, commissioning
and operation of OSIRIS a conceptually simple
instrument can be built to provide unprecedented,
extremely high spatial resolution measurements of
species related to the exchange processes that
occur within the Upper Troposphere Lower
Stratosphere region.
28
CATScan The Canadian Atmospheric Tomography System
29
Optical Spectrograph and InfraRed Imaging System
Two separate technologies the OS and the IRI
30
InfraRed Imagers (IRI)
- the three InfraRed Imager channels are very
simple - they consist of a detector, narrow
passband filter, lens and baffle - each detector
has 128 pixels that sample the line of sight
brightness at about 1 km vertical resolution at
the tangent point
31
the spacecraft nods
The 2900 images collected during an orbit
32

• each line of sight integrates the volume emission
  rate through the atmospheric limb
• the atmosphere can be subdivided by angle and
  distance from the centre of the earth

33
(No Transcript)
34
(No Transcript)
35
OSIRIS A Canadian Effort to Study Ozone from
Space
are used in a two-dimensional tomographic
retrieval to determine the volume emission rate
profile as a function of angle along the
satellite track and altitude (cell size of 0.2
degrees by 1 km)
36
aerosols are measured by looking at the change in
the ratio of the intensity at 470 nm to that at
750 nm Information at only two wavelengths is
required
37
high altitude ozone is measured by looking at
absorption in the blue end of the
spectrum Information at only 7 wavelengths is
required
38
low altitude ozone is measured by looking at
absorption in the red end of the
spectrum Information at only three wavelengths is
required
39
too noisy
1) 432.79 nm 435.17 nm 437.94 nm 2) 437.94
nm 439.52 nm 441.89 nm 3) 446.64 nm 448.23
nm 449.81 nm
40
SaskMART
Multiplicative Algebraic Reconstruction Technique
similar to the Maximum Likelihood Expectation
Maximization technique commonly used in medical
imaging
41
UTLS Exchange
at most 60 scans per orbit so at most 60 vertical
profiles per orbit or 600 km separation between
profiles
42
Now recall that with the IRI technology the
spatial grid cell size is 20 km by 1 km even with
low spatial resolution sampling
43
What have we learned from OSIRIS?
44
What have we learned from OSIRIS?

• High spatial resolution retrievals can be made
  with data collected by vertical imagers

45
What have we learned from OSIRIS?

• High spatial resolution retrievals can be made
  with data collected by vertical imagers
• Information at only 18 discrete wavelengths is
  required to retrieve ozone, nitrogen dioxide and
  stratospheric sulphate aerosols from the cloud
  tops through into the lower mesosphere

46
What have we learned from OSIRIS?

• High spatial resolution retrievals can be made
  with data collected by vertical imagers
• Information at only 18 discrete wavelengths is
  required to retrieve ozone, nitrogen dioxide and
  stratospheric sulphate aerosols from the cloud
  tops through into the lower mesosphere
• The altitude range for each wavelength is
  different so there is no need to image the entire
  atmosphere in each spectral region

47
CATScan The Canadian Atmospheric Tomography System
the technology is extremely simple
CATScan will be made up of 18 almost identical
imagers 10 for ozone, 6 for nitrogen dioxide and
2 for aerosols
48
CATScan The Canadian Atmospheric Tomography System
the technology is extremely simple
figure courtesy of Routes AstroEngineering
configured to look within the plane of a
satellite in low earth, sun-synchronous, near
dawn-dusk orbit
49
CATScan The Canadian Atmospheric Tomography System
the technology is extremely simple
figure courtesy of Routes AstroEngineering
There are no serious developmental risks the
baffles must be as long as possible to minimize
contamination from off axis light
50
CATScan The Canadian Atmospheric Tomography System
the technology is extremely simple
figure courtesy of Routes AstroEngineering
There are no serious developmental risks the
optics must be kept very clean to minimize
contamination from off axis light
51
CATScan The Canadian Atmospheric Tomography System
the technology is extremely simple
figure courtesy of Routes AstroEngineering
There are no serious developmental risks the
vertical field of view of each channel is kept as
narrow as possible to minimize contamination from
off axis light
52
CATScan The Canadian Atmospheric Tomography System
The ozone channels the detectors have either
256, 512 or 1024 pixels depending on the required
vertical resolution
53
CATScan The Canadian Atmospheric Tomography System
the technology is extremely simple
figure courtesy of Routes AstroEngineering
There are no serious developmental risks the
calibration and characterization must be very,
very thorough
54
CATScan The Canadian Atmospheric Tomography System
if we sample with a temporal resolution of 4 Hz
55
CATScan The Canadian Atmospheric Tomography System
and use the CATScan equation
56
CATScan The Canadian Atmospheric Tomography System
the spatial resolution that can be achieved is
unprecedented for measurements of these species
57
CATScan The Canadian Atmospheric Tomography System
Conclusion Using lessons learned during the
design, build, characterization, commissioning
and operation of OSIRIS a conceptually simple
instrument can be built to provide unprecedented,
extremely high spatial resolution measurements of
species related to the exchange processes that
occur within the Upper Troposphere Lower
Stratosphere region.
58
More About the Ozone Retrieval
59
(No Transcript)
60
(No Transcript)
61
(No Transcript)
62
(No Transcript)
63
(No Transcript)
64
(No Transcript)
65
(No Transcript)
66
(No Transcript)
67
(No Transcript)
68
(No Transcript)
69
(No Transcript)