With Jean-Marie Flaud, reanalysing MIPAS spectroscopic
database
- Minor modifications to CO2, O3, HNO3, COF2
- Currently rerunning ORM with old/new databases in line-by-line mode
(~4000s per limb scan) to see if differences are noticeable.
Non-LTE REC Analysis (AD)
[Prev]
[Next]
REC analysis shows net non-LTE signature close to zero rather than
the expected value ~1 (eg Previous Meeting)
- Use new REC LSF option to provide independent fit of non-LTE signature
to residual at each microwindow/altitude
- Also divide residuals into day/ngt and by latitude band according to
Occupation Matrix (OM 006=65-90N, 005=20-65N, 004=0-20N, 003=0-20S,
002=20-65S, 001=65-90S).
- Although fits are obtained for all combinations, only the most
statistically significant results are plotted (ie those MW/alts with
the largest expected mid-latitude day-time non-LTE signatures).
- Three sets of residuals have been averaged and analysed:
- ESTEC: Orbits 2081-2083 processed at ESTEC
- July: 1 month of ESA data received on CDs
- August: 1 month of ESA data received on CDs
- Day and night are shown on separate pages, different latitude bands
in different panels (only day-time OM002-OM004 are expected to match
the mid-lat daytime non-LTE residuals fitted here), different colours
indicate the different species microwindows, and different symbols
the different altitude ranges.
- Conclusion is that the non-LTE signal is close to expected fit (=1) at
high altitudes (circles and downward-pointing triangles) but lower
altitudes are rather random and these reduce the overall fit to
numbers near zero.
