Oxford MIPAS meeting#99
18 July 06

Present

Oxford: Anu Dudhia, Jane Hurley, Joanne Walker, Claire Waymark, Chiara Piccolo
RAL: Alison Waterfall

Just completed 3 days of nominal mode measurements (12-15 July) and about to start lengthy period of UTLS1 measurements in support of AMMA campaign (W.Africa): 21 Jul - 21 Aug.

Current version of Mission Planning Document (2006 up to Aug22) [XLS] [HTML].

June Monthly Report [Prev] [Next]

MIPAS monthly report for June is now available

[Summary].
Full report downloadable from http://earth.esa.int/pcs/envisat/mipas/reports/monthly/

L1B Data (JH) [Prev] [Next]

L1B coverage maps showing measurements taken since Jan 2005 in reduced resolution mode and processed as far as L1B (ie spectra).

Check for new 2005 data now complete - new files for Jun/Jul/Aug 05 but these might simply be the reprocessed L1B mentioned in the June Report.
Will try to update 2006 coverage maps this week.
ESRIN have today issued a list [XLS] [HTML] of all Reduced Resolution L1B files currently processed at DPAC (which should eventually find its way on to our coverage maps). One file per orbit, ordered by date (part of filename ...1PPDPA2006mmyy_...)

HiRDLS Comparisons (CW) [Prev] [Next]

Currently processed MIPAS RR27 data from 28Jan05 and 4May06 (using Oxford L2 processing scheme) for HIRDLS comparisons. Now looking for an extra day, preferably MIPAS nominal mode coinciding with HIRDLS validation days, for further comparisons.

MIPAS Precision (CP) [Prev] [Next]

Investigating MIPAS L2 precision values by comparing scatter between retrieved profiles where orbital tangent tracks intersect with scatter predicted by retrieval SD.

[Plot] shows temperature comparison at 89.2N during Aug 2003 - red bars are predicted scatter and blue line is observed scatter
Good agreement at most altitudes, although observed scatter much larger than predicted below 100mb surface
This is polar summer, corresponding polar winter plots show significantly larger observed scatter, probably due to horizontal gradients.
Most of the VMR scatter is consistent with predicted errors, but can't use this test for NO2 since there is a large diurnal difference at the intersections of any two orbits.

ACE/ODIN Comparisons (CP) [Prev] [Next]

Using ACE and ODIN measurements for validation of ESA MIPAS L2 data

ACE & ODIN O3 comparisons with MIPAS now completed, including systematic errors
Now working on other species.

RR27 pT Retrieval (AD) [Prev] [Next]

pT microwindows for reduced resolution, 27 sweeps (PT__0331-335) have been observed to produce a spike in the temperature retrieval at the 21km nominal altitude (ie, 11th tangent point from bottom of scan) when using the IFAC processor

[Plot] from IFAC showing average temperature profile for Orbit 17540, 8Jul05.
This is not evident in the Oxford processor using the same microwindows [Plot - diamonds] but if the correlation in the a priori temperature covariance is removed some increased instability at 21-23km is evident (crosses, line).

It is thought this may be due to a cut off in the occupation matrix for the longer wavelength microwindows (331,335) which only extend down to 21km.

Microwindow	Wavenumber Range		Altitude Range
PT__0331	739.8125	742.8125	23.0	70.0
PT__0332	942.3125	945.2500	6.0	31.0
PT__0333	948.6875	951.6875	6.0	29.0
PT__0334	719.3125	722.3125	27.0	70.0
PT__0335	698.9375	700.8125	23.0	37.0

Currently looking for an additional microwindow at the longer wavelengths spanning the 21km altitude to see if this produces a smoother profile

Cloud Top Height (JH) [Prev] [Next]

Generating cirrus climatology using MIPAS spectra, trying to fit cloud-top-height within 3km vertical box represented by MIPAS FOV

Inconsistency between CTH obtained using A and B band microwindows may be explainable in terms of having an inhomogeneous cloud top in the FOV - modelling this as a homogeneous cloud top will always produce a higher equivalent brightness temperature in the B band than in the A band
Conversely, where the A and B bands produce similar CTH answers, then the modelling assumption of a homogeneous cloud partly filling the FOV is more likely to be valid.

N2O Anomaly Mar03 (AD) [Prev] [Next]

Anomalously high N2O concentrations reported in ESA MIPAS L2 data around 7-8 March and 13-15 March 2003

Anomalous values also apparent in other species at these times
ESRIN Report [DOC] [PDF] (from F.Niro) concludes the following
- It is a problem in the L1 ADFs used for the radiometric calibration
- In the L1 ADF database there was a validity gap, as a result the processor did not find any available L1 ADF, then it used a very old set (from June 2002!). This causes a completely wrong spectra calibration which reflects in the anomalous retrieval results.

Impact of Reduced Resolution (JW) [Prev] [Next]

Simple theory predicts that, when the spectral resolution is reduced by factor M

Molecules with individual line features will have S/N degraded by root(M)
Molecules with broad spectral features will have S/N unchanged

Coupled with a factor M increase in number of measurements (corresponding to the reduced time required to acquire a single spectrum at lower resolution), giving an extra factor root(M) increase in S/N, so net effect on retrievals is

Molecules with individual line features will have net S/N unchanged
Molecules with broad spectral features will have net S/N improved by root(M)

Currently running some simulations for CO and N2O5 as examples of the former and latter types of molecule to see if this is actually how it works for MIPAS in the FR17, RR17 and RR27 modes of operation.

GOME Comparisons (AW) [Prev] [Next]

RAL GOME O3 profile retrieval algorithm as part of MIPAS O3 validation, apply GOME averaging kernel to reduce MIPAS vertical resolution.

Summary of June Monthly Report (F. Niro, ESRIN).

The MIPAS instrument performs really well during the reporting month; in fact only 7 INT velocity errors were registered (as in the previous month). This new operation stability lead the Science Team meeting to envisage an increasing duty cycle during campaign mode; this strategy will be implemented starting from next month by planning 6 orbits per day (so far campaign mode were planned with 4 orbits per day).
During June 2006 the instrument operated with a duty cycle of about 42%, since it was planned always in the 3 days-on / 4 days-off configuration. This value is higher than the previous month, when the instrument was operating prevalently on campaign mode (4 orbits /day). As already cited in the previous bullet the instrument performances were really good during the reporting month; the instrument availability is indeed equal to 95.6 % of the planned measurement time. The missing intervals due to PDS unknown failure were about 1.6% of the planned time, giving a total L0 availability of about 94.1%.
The instrument operated during June with the following plan:

NOM modes were planned during 13 – 16 June
UTLS-1 modes were planned during: 6 – 8 and 27 – 30 June,
In-Flight calibrations (IF-9, IF-11 and IF-16) were planned the 5th of June
MA/UA modes were planned during the solstice: 21 – 24 June

All the planned measurement were acquired unless some data gaps introduced by instrument anomalies or acquisition failure at PDS (see §2.2.2 and §2.2.3). The planned IF16 RAW mode measurement was not acquired due to an anomaly in the RAW data processing. Note that such RAW mode acquisition measurement failed already due to the same anomaly during October 2005 and March 2006. An anomaly report has been raised against PDS (OAR-2015), and the investigations are on-going.

The monitoring of the instrument temperatures continued nominally this month. We observe an overall stability of all the devices temperatures, the variation over the entire month being less than 1.5K (see § 2.3.1).

The cooler performance was closely monitored on a daily and monthly basis. The cooler seems to perform really well, even though we observed some orbit-dependent spikes that reach the limit of 8 mg at their maxima. These spikes are due to the warmer environment conditions, typical of this part of the year, they are not critical, but they should be monitored carefully (see § 2.3.3.1).

The spectral correction factor long term monitoring of the RR mission (since Aug 2004) shows a stable situation since the variations are of the order of 3 ppm over almost two years of operations (see § 2.4.2).

The gain calibrations were carried out nominally during the reporting period, including the dissemination of the related ADF. The gain weekly increase remains stable, the maximum of gain increase in the band A between two consecutive disseminated gains remains below the acceptance criterion of 1% (see § 2.4.3.1).

The long term gain increase monitoring since the last decontamination (June 2005) is presented in this MR. This monitoring will help the planning of the future decontamination activity (to be expected at least twice per year). We can observe that the gain increase looks linear with time after the decontamination induced by the PLSOL of April. Nowadays the gain in the band A is about 15% higher (at its maximum) with respect to the first gain disseminated after the decontamination of June 2005 (see § 2.4.3.2).

The mispointing long term monitoring show also a steady situation; in fact during the last year the relative bias seems to be stable around a value of few mdeg. To be noted that the last disseminated LOS ADF was on March 2005. (see § 2.4.4).

The level 1 quality monitoring of the OFL products created at D-PAC is on going in parallel with the mission. The check of the considered L1 products has demonstrated an overall good quality of the D-PAC results (see § 2.4.5). The level 1b daily reports can be accessed at the following address: http://earth.esa.int/pcs/envisat/mipas/reports/daily/Level_1_OFL/
A major anomaly was observed on L1 data from 3 – 23 June 2005 and 29 July – 11 Aug 2005. These data have very poor quality due to a wrong offset calibration. This problem is now closed since a new set of ADFs were generated and the affected L1 products were re-processed, the list of the reprocessed orbits can be found on Uranus ftp server (§3.6.5).
The long term monitoring of fringe count errors (FCE) for the RR mission is shown in this MR; this analysis is based on the L1b products generated OFL at D-PAC. This long term monitoring aims at the verification of the stability of FCE over time; furthermore this analysis is useful in order to verify if any correlation exists with the INT performances degradation. No evident trend can be observed over more than one year of RR mission, nevertheless in the last month the maxima of fringe count seem to decrease.
A quality control of L2 RR OFL products processed with the latest IPF 4.65 (Aug – Sept 2004) was carried out at ESRIN, showing an overall good quality of the level 2 data. Only one major problem was found in the L2 RR data processed with IPF 4.65 for some orbits recorded during 21 – 22 Aug 2004. The investigation of this problem showed that a corruption in the band D was verified for these orbits. As reported by Astrium the processor flags as corrupted one sweep even though only one band is corrupted. This processor specification seems excessively restrictive and it should be modified. To see all the L2 OFL daily reports of RR mission follow the link below: http://earth.esa.int/pcs/envisat/mipas/reports/daily/Level_2_OFL/

Oxford MIPAS meeting#99 18 July 06

Oxford MIPAS meeting#99
18 July 06