further enhance speed of processing and performance. A more
detailed description of the ASPECT Retrieval Algorithm,
illustrated in Figure 7, is the first step is to run the NCEP
atmospheric profile, sun/sensor geometry, through KLAYERS
algorithm and then kCARTA algorithm two times, this is done
to raise the simulated methane column amount, and rerun
kCARTA (once using optical depth KLAYER profile (A) and
once with perturbed KLAYER profile (B)).
From profiles A and B we have formed our Jacobian matrix.
This matrix is the change in the radiance as a function of a
change in the amount of methane; the Jacobian derived is then
used in a comparison to the measured radiometrically calibrated
data to the reflected scaled unit matched spectral radiance. This
is a spectral comparison of the difference in measured data and
the reflectance scaled unit matched radiance spectra divided by
the perturbation amount (or the change of methane in the state
Rene KG
jacobian
optical depths
, Reflectance scaled f
i matrix {model
nit matched — 4
fadiance spectra
Figure 7. Lockheed Martin Methane Retrieval Processing
Diagram
of the system). The result is a calculated spectra that is
generated applying the methane perturbation amount, and the
difference from the measured spectra and is determined. If this
value is near the order of the noise then the spectra agrees with
the measurement, if not then one must resume the loop
illustrated in Figure 7 and iterate the Jacobian calculation once
more while applying small changes in methane amount, CH,,
until the spectral difference converges (i.e., is less than noise).
This retrieval algorithm is shown in Figure 7. The calculation
from KLAYERS through kCARTA and are done, and we
manually adjust the reflectivity factor and iterations of Jacobian
calculations to calculate total column CH,. As is illustrated in
the retrieval algorithm figure, there are components of the
retrieval algorithm that are currently done manually (green
boxes in Figure 7). We propose in future efforts to automate
these to enable an algorithm for ASPECT retrieval that does not
include a ‘man in the loop.’ We have done retrievals on both
the data over the sunglint with and without the methane cell.
Data and calculations are shown in Figure 8. The final results of
running the ASPECT Retrieval Algorithm (Figure 7) upon the
sunglint collections retrieved the amount 0.903 atm-cm of CH,
in the gas cell. The gas cell was known to contain 0.908 atm-
cm of CH,to 196 accuracy.
Figure 8. This graph shows a methane retrieval spectra
collected over the sunglint. The dark blue line is the
observation, green is with no perturbation, red is the
methane amount perturbed, aqua is the reflectivity
factor and purple is the final calculation with
reflectivity applied.
3. CONCLUSIONS
The analysis demonstrated retrieval of a gas cell simulated
enhancement in the total atmospheric CH4 column of 0.908 atm
— cm with 1% accuracy. This is an equivalent sensitivity to an
enhancement of 200 ppm in the first meter of the atmosphere.
In a parallel effort Kumer et al., 2011, have calibrated and used
the data from consecutive frames with the absorption cell NOT
in the beam and showed retrieval of total atmospheric column to
1% precision. There is potential to considerably improve the
precision. For example, for the use of all spectra (in this study
only two out of the 512 taken along a given slit of a TIMS
collection) the precision is theoretically expected to improve to
the order of an 18 ppm enhancement in the first meter of the
atmosphere sensitivity. This research constitutes an important
step in demonstrating sounding CH4 enhancements over water
by the use of high-resolution spectral measurement of sun glint.
REFERENCES
Desouza-Machado, S., Strow, L., Hannon, S., Dec. 1997.
kCompressed Atmospheric Radiative Transfer Algorithm
(kCARTA). Proceedings SPIE, 3220, Satellite Remote Sensing
of Clouds and Atmosphere II, Joanna D. Haigh; Ed., pp. 156-
162.
Kumer, J.B., Roche, A.E., Rairden, R.L., Mergenthaler, J.L.,
Doolittle, J.H., Blatherwick, R.D., Hawat, T., Chatfield, R.B.,
July 2008. Tropospheric Infrared Mapping Spectrometers
(TIMS) to Provide Measurements with much Improved
Vertical, Temporal, and Spatial Resolution in the Lower
Troposphere. Geoscience and Remote Sensing Symposium,
2008. IEEE International. 111-522 — III-525.
Kumer, J.B., Rairden, R., Roche, A., Chatfield, R., 2011.
NASA ESTO IIP Tropospheric Infrared Mapping
Spectrometers (TIMS) Demonstration First Deployment on an
Airship: Preliminary Results. Hyperspectral Imaging and
Sounding of the Environment, Optical Society of America
Technical Digest (CD) (Optical Society of America), paper
HTuD2.
Larsen, N
from Spa
016202.1-
We thank
Departmer
his deve
Radiative
using this
