A `feature' has been noticed in our retrievals of NO2 from orbits
15237-15239 (29 Jan 05): both day and night parts of orbit have a sharp
dip in NO2 values at high altitude with +/- 10 deg of the equator
[Plots]
- As expected, it turned out to be related to the interpolation of
the NO2 IG data which is used as a priori:
- IG profiles are supplied for latitudes +/- 10, 45 and 75 deg
[Plot]
- MORSE uses a linear interpolation in latitude
- The IG profiles for +/-10 deg (blue curve, overlying green curve)
have extremely small concentrations above 60km (10-6 ppmv) compared
to mid-latitudes (yellow curve, overlying red, 10-3 ppmv) that
any linear interpolation poleward of 10 deg N will be heavily biassed
in log space towards the 45 deg value, while equatorward will be
constant at the low value.
- Effect in MORSE retrievals could be reduced by log interpolation in
latitude and/or higher a priori uncertainty assigned to IG profile (100%
assumed by default), but IG profile latitude variation looks unrealistic
in any case (note that these are nominally averages of the day/night values
since there is no mechanism in the ESA processor for using day/night dependent
IG profiles).
HNO3 Retrievals (CP)
[Prev]
IFAC have been investigating impact of updated HNO3 line parameters
but notice a larger discrepancy (~4% in peak HNO3)
between their retrievals using a line-by-line
model and the HNO3 LUTs than between the two sets of line parameters.
- Difference turns out to be due to different TIPS coefficients used in
the Oxford and IFAC line-by-line models - Oxford uses more up-to-date
coefficients (supplied by Bianca Dinelli and Elisa Castelli
from Bologna) so no problem with current LUTs.
RAL Work (AW)
[Prev]
[Next]
Alison will be looking at MIPAS and, later, HiRDLS data.
Currently setting up MORSE - has some strange features in H2O retrieval
to be investigated.
