Oxford MIPAS meeting#43
11 Nov 03

Present

Oxford: Alastair Burgess, Anu Dudhia, Vivienne Payne, Chiara Piccolo

Instrument Status [Prev] [Next] (Some details on the Envisat Web-Site)

Functioning normally

Changes in ESA L1B data in NRT processing from 04 Nov 03 onwards. These are described in a Tech.Note from ABB (03 Nov 03), summarised as follows:

Non-linearity correction for forward/reverse sweeps
Spectral calibration
ILS fitting

L2 Data [Prev] [Next]

In /home/crun/eodg/mipas/L2/ [Plots of profile locations]

Data received up until 10th November, plotted up until 31st October.
Some L2 data has now been retrieved for every day since 20th September, 52 days of continuous operation.
Software version 4.59 in use since 23 Jul 2003.

ESA Processing [Prev]

From Rob Koopman (10 Nov 03)

Maybe its useful that I recapitulate the latest important anomalies that we are investigating.

1. October data display increased spread in temperature values
In progress within the QWG, but it cannot be excluded that it is linked to item 6 (LOS calibration). Once the LOS calibration has been performed, analysis will be resumed.
2. O3 and H2O discontinuity has been observed towards the end of October
The final conclusion is to be drawn at the QWG-2 meeting, but consensus seems to be that the solar flare is at the heart of the problem [see 52km Zonal Mean Traces below]
3. Some corrupt source packets have been received form the instrument towards the end of October
Also for item 3 this is a suspected cause. This issue is currently under investigation at ASTRIUM.
4. Data production has stopped on 4 November
The investigation has shown the problem to be due to installation of a new PFHS software, which was incorrectly configured. This configuration has been corrected in the mean time, and production has resumed.
5. Cloud retrievals from the operational system seem to deviate from specifications
In progress with currently a verification action ongoing at ESRIN [see comment from John Remedios]
6. Absence of star signal during LOS calibrations in October.
Due to evolution of the mispointing. This evolution has been characterised in the mean time, and commanding for LOS pointing has been adjusted during an operations plan upload in early October. The LOS measurements with modified commanding have been executed on 8 and 9 November. Data from 8 November have been received today and LOS calibration analysis is underway. Historic pointing information will be derived, in order to not only improve forward processing in November, but to also improve October data quality
In addition to anomaly investigation, recent progress has been achieved on the product improvement: on 4 November the version 3.6 auxiliary data have been activated in the operational chain, please see my separate message to the QWG. [see L1 Data above]
A third item with progress is the reconstruction of the calibration history. The full MIP_CL1_AX history has been completed until September 2003. Reconfiguration is underway to reconstruct the history of MIP_SP1_AX, MIP_CO1_AX and MIP_GC1_AX with the latest version 3.6 ADF configuration.
A fourth area of work are the reprocessing preparations. The processing of the MIPAS Validation Data Set will now be performed at the D-PAC in stead of on a stand-alone chain. This offers large advantages in terms of processing time.
Obviously there is more work ongoing (instrument and product monitoring, cooler performance analysis, interferometer analysis, etc), but these 'almost routine' investigations will be discussed at the QWG meeting.
The agenda for this meeting will be issued at the beginning of this week.

Zonal Mean Traces [Prev] [Next]

30km plots of day-by-day zonally-averaged NRT data 30km nominal tangent height

Large deviations from day-to-day (which appeared to start in early October) continue. As noted by Rob, these may be associated with mis-pointing given the large altitude variations.

52km plots now also generated. Antje Dethof (ECMWF) noted low retrieved mesospheric O3 and H2O values for N.Polar region at the end of October - related to increased solar flare activity?

Joint Retrieval Microwindows (AD) [Prev] [Next]

Selecting microwindows for joint retrieval of pT, H2O and O3.

Plot shows information content as a function of number of measurements for different microwindow selections, with existing independent retrieval microwindows for comparison.

Solid symbols are total number of measurements within microwindows, open symbols are number of `unmasked' measurements which are actually used in the retrieval

Green: 10 joint MWs, each of max width 1cm^-1
Brown: 10 joint MWs, each of max width 2cm^-1
Dark Blue: 5 joint MWs, each of max width 3cm^-1
Black: 7 operational pT microwindows
Light Blue: 3 operational O3 microwindows
Blue: 4 operational H2O microwindows

Results emphasise `efficiency' of joint retrievals.

OPTIMO Retrievals (CP) [Prev] [Next]

Reprocessing Sep/Oct 2002 data using the Oxford processor, OPTIMO. pT, O3, H2O and CH4

Results compared to ESA retrievals are much improved if climatological profiles (rather than previous retrieval) is used as the first guess and the a priori.

CO Microwindows (AD) [Prev] [Next]

Look-Up Tables and irregular grids now being generated for `LTE' microwindows suitable for retrieval of CO at nighttime

Plot showing predicted retrieval error as a function of altitude, solid=accuracy, open=precision

Black - LTE microwindows, nighttime
Grey - LTE microwindows, daytime (note large errors)
Light Blue - non-LTE microwindows daytime (assuming a joint retrieval of CO and vibrational temperature)

REC Analyses [Prev] [Next]

Results for October. Summary:

Large changes in the spectral signatures of the target species associated with the 65S-90S region
Main changes in the spectral derivative signatures are associated with the NO2 microwindows.

Occupation Matrix Statistics [Prev] [Next]

Results for October.

Long-standing problem in L2 data counting the number of profiles used in averaging residuals and occupation matrices now seems to extend to pT profiles (since early October) showing, for example, many more profiles in southern polar region than actually the case - statistics using these numbers are no longer useful
Pending correction to the ESA L2 data, it will be necessary to modify the OM statistics program to calculate its own profile counter from other information in the L2 data.

Monthly Mean Data [Prev] [Next]

Results for October

Altitude (AD)
- Altitudes corrected by L2 pT retrieval are usually higher than the nominal altitudes, particularly for the southern hemisphere, reaching a maximum difference of 2.5km for high altitude, S.Polar region
- No significant changes from the previous month
Temperature (AD)
- S.Polar temperatures up to 40K warmer than the `initial guess' (IG) climatology in the stratosphere, colder in the mesosphere
- N.Polar temperatures up to 70K cooler than climatology at 40km, warmer in the mesosphere
- Other latitudes generally 0-10K cooler than climatology in the stratosphere
- Compared to previous month, N.Hemisphere temperatures getting cooler, S.Hemisphere temperatures getting warmer, as expected.
Pressure (AD)
- Retrieved pressures up to 50% lower than climatology in the N.Polar region, 100% higher in the S.Polar
- Retrieved pressure within 20% of climatology for other latitudes
- Compared to previous month, main change is the increase in S.Polar pressures - deviation from climatology has doubled.
CH4 (VP)
- MIPAS values are generally higher than climatology by around 0.1-0.3ppmv.
- The retrieved values for equatorial regions at 12 km are significantly higher than the climatology, with differences of up to 1.5ppmv, and also show a marked local minimum at 15km (same for N2O). It could be that this feature arises due to the vertical resolution of MIPAS being unable to accurately represent the temperature structure at the tropopause. [see comment from Thomas von Clarmann].
- No significant change since previous month
H2O (VP)
- Equatorial values at 1km much larger than would be expected from climatological profiles.
- Aside from this, monthly mean H2O profiles for all latitudes generally agree with climatology to within 1ppmv throughout most of the stratosphere.
- From 33km to 52km, MIPAS water vapour is generally higher than climatology. Since the main source of water vapour in the stratosphere is methane oxidation, this would appear to be consistent with the observed methane values.
- MIPAS H2O values at 60km appear rather low, being up to 3.5ppmv lower than climatological values at this altitude.
- The southern polar mean profile is for October is signficantly drier than the September profile, but the September profile showed much higher values than expected from climatology.
2CH4+H2O (`Total Hydrogen') (VP)
- H2O+2CH4 values are reasonably constant throughout the stratosphere, as would be expected.
- Values at 60km are somewhat low, due to the low water vapour values at this altitude.
- With the exception of 60km, where the water vapour is low, and at 15km, where the methane is low, the profiles of H2O+2CH4 for are around 1ppmv higher than climatology.
- Approximately half of this difference can be attributed to water vapour and the other half to methane.
HNO3 (AB)
- The basic structure at all latitudes is similar to previous months, with lower altitude equatorial regions being enhanced when compared to climatology.
- A deficit at high altitudes over the southern pole still exists.
- There has been an increase in the northern polar mid-stratospheric concentration in the last month.
N2O (AB)
- The month-on-month differences remain very small, with only a slight increase in Southern polar VMR over last month.
- There is a persistent enhancement at 12km for equatorial regions with a reduction at 15km (same as for CH4).
- The standard deviation of higher altitudes has reduced over last month.
NO2 (CP)
- N.Polar values high at 47 km (increase of 6ppbv compared with previous month)
- S.Polar has peak value at 30km, other latitudes have peak values between 33-36km
- S.Polar profile decreased with respect to previous month
O3 (CP)
- Shift of the peak values with respect to initial guess ((IG) climatology for all latitudes
- Values continue to exceed IG climatology by around 1ppmv at 40km at all latitudes
- Equatorial latitudes show enhancements at 12km
- No significant changes compared with previous month

ASSFTS-11 Workshop (8-10th Oct, Bad Wildbad, Germany) [Prev] [Next]

Proceedings available on http://www-imk.fzk.de/asf/ame/assfts/

Envisat Validation Workshop (9-13th Dec 2002, ESRIN) [Prev] [Next]

Proceedings available on http://envisat.esa.int/pub/ESA_DOC/envisat_val_1202/proceedings/

Email from Thomas von Clarmann, 12 Nov 2003

We also see high CH4 and N2O in our retrievals and also suspect that insufficient altitude resolution of T could cause this trouble. More realistic N2O and CH4 values were retrieved from MIPAS S2 mode measurements (which hints in the same direction, because also T is better resolved there). Some further tests on this are planned (starting next week). I will keep you posted.

Email from John Remedios, 12 Nov 2003

I think it is now clear that the operational processor is not flagging all clouds and removing them from the retrieval. The problem with the processor has not yet been identified but an IFAC retrieval shows that the prototype processor and auxiliary data file work fine. In particular we have seen that many of the PSCs are not removed in the Antarctic spring leading to anomalous zonal means for gases. Also, some tropical clouds are missed which can also lead to high values of CH4 etc. in the mean values at altitudes below 20 km.

Oxford MIPAS meeting#43 11 Nov 03

Oxford MIPAS meeting#43
11 Nov 03