Oxford MIPAS meeting#21
17 Dec 02

Present

Oxford: Alastair Burgess, Anu Dudhia, Don Grainger, Martin Juckes, Vivienne Payne, Chiara Piccolo, Clive Rodgers.
RAL: Victoria Jay

Instrument Status [Prev] [Next]

Operating normally

Software Updates [Prev] [Next]

Software in /home/crun/eodg/mipas/

L2/util/
- (VJ) update RS subroutine read_continuum.pro
- move Reinhold Spang's L2 read subroutines into /read

L1/L2 Data Delivery [Prev] [Next]

(CP) collecting L2 data via ftp from Kiruna and ESRIN, storing in /home/crun/eodg/mipas/L2/NL/, separate directories for each day.
(AB/VJ) L1B data being archived at BADC
cdproc.pro has to be adapted to extract residuals and generate profile maps from the L2 data stored on disk.

RAL Retrievals (VJ) [Prev] [Next]

2D Joint retrieval of O3 and H2O using ESA pT profiles

Adapting retrieval for more sophisticated choice of initial guess, e.g, climatology, previous retrieval, ECMWF
Problems matching 6km and 9km sweeps to ECMWF temperatures due to lack of pressure/pointing retrieval in ESA pT profiles below 12km.

Data Assimilation (MJ) [Prev] [Next]

Running simple single-level assimilation scheme for MIPAS O3, driven by ECMWF winds.

Modified advection to use cubic rather than linear interpolation
Looking at degrees of freedom for signal and noise for consistency check

Gain Oscillation (VP) [Prev] [Next]

Manfred Birk (DLR) apparently knows the origin of the oscillations and has a possible solution
OPTIMO pT retrieval oscillations only slightly reduced with large a priori correlation - might fudge L1B gain to see what happens when gain oscillation is removed.

ILS Uncertainty (AB) [Prev] [Next]

SF6 retrievals show unreasonably large values at high altitudes.

This may be due to the fact the the main SF6 peak sits close to a CO2 line so could be affected by any uncertainty in AILS width
Uncertainty in AILS width not well know, so generating error spectra corresponding to 1% spread in AILS, to be correlated with 2nd derivative signature in residuals
Aim is to quantify AILS uncertainty in terms of a % and then propagate this through the MWMAKE error analysis

Envisat Validation Workshop [Next]

Held 9-13 December, Frascati. AB,AD,VP and CP attended.

MIPAS data looks reasonable but data distribution problems, GOMOS and SCIAMACHY data less well-behaved

Main problems with MIPAS data identified as

Occasional spikes at low altitude due to cloud contamination
Profiles sometimes oscillate (notable CH4 and N2O in comparison with Geophysica aircraft measurements)
H2O apparently biassed ~20%/1ppmv too high in stratosphere

H2O Bias (AD) [Prev] [Next]

Comparison with other instruments shows MIPAS H2O generally high in stratosphere.

Also evident in plot of MIPAS 2CH4+H2O, which increases with altitude to around 8ppmv instead of remaining constant at 6ppmv.

SAG presentation - changing microwindows produces 1-2ppmv lower at stratopause, no evidence of H2O residuals in other microwindows (more statistics required).

Possible causes:

Retrieved temperature too low? no evidence in other species.
Non-LTE effects? Some evidence of day/night difference but not large enough to explain bias
Spectroscopic line strengths? Uncertainties significant but 20% is rather large
Isotopic Lines (Microwindow selection assumes isotope ratios remain constant)? HDO is much lower in stratosphere but HDO lines not present in MIPAS H2O microwindows (and wrong sign anyway).
Residual Water vapour onboard Envisat? Doppler shift would completely decorrelate these with atmospheric lines (Satellite velocity ~km/s cf atmospheric kinetic energies ~100m/s).
ORM Convergence? Not expected to produce bias, but check if OPTIMO H2O produces same mean bias as ORM using same microwindows
Channel C gain calibration? Not expected to be problematic (unlike A, AB and B which are non-linear detectors) but should be checked
Instrument Line Shape? MW selection assumes C band AILS is almost perfectly known but 2nd derivative signature in residuals shows some extra uncertainty - to be quantified

NO Retrievals (CDR) [Prev] [Next]

Using MWMAKE to determine feasibility of retrieving UTLS NO from TES

Still using MIPAS ILS, get very different microwindows if altitude range is restricted to TES values
Only very small contribution to Jacobians from UTLS region - might have to investigate using whole band rather than microwindows.

Oxford MIPAS meeting#21 17 Dec 02

Oxford MIPAS meeting#21
17 Dec 02