NRT Notification Detail

Dissemination of ROM SAF Metop-B products resumed
DATE:   2013-04-08 11:50
Dear ROM SAF GTS Users

Issued: 2013-04-08 11:45 UTC

The dissemination of the ROM SAF demonstration NRT refractivity product from Metop-B/GRAS (product id GRM-40) resumed with instrument sensing time 2013-04-08 09:47:17 UTC.

EUMETSAT CAF has implemented a change in the Metop-B/GRAS instrument configuration and enabled the so-called extrapolation of the L2 signal. This means that we are able to process more occultations to lower altitudes, but also that the quality of the retrievals are different than for Metop-A. In the lower troposphere (below 8 km), Metop-B/GRAS refractivity, when compared to ECMWF forecasts, shows a significant negative bias (larger for rising than for setting), and also a larger standard deviation. For rising occultations we have also observed a larger standard deviation between 10 and 20 km (up to 100% larger), which is currently under investigation. The quality of setting occultations above 10 km is unaltered. It should be recalled that the current Metop-B/GRAS produts are derived from a geometric optics algorithm (and using closed loop data only) which means that the data are of limited use below a height of about 10 km.

For more information see also the information from EUMETSAT CAF included below.

Please do not hesitate to contact us if you have any questions related to the data. We would also be very interested in learning about your experience and analysis of these Metop-B/GRAS data.

With kind regards
The ROM SAF Team

Forwarded Message:

Subject: GRAS/Metop-B data being disseminated again
From: Christian Marquardt
Date: 26. marts 2013 11:48

Dear all,

since yesterday, 25th March 2013 08:38 UTC sensing time, we are again disseminating level 1b data from the GRAS instrument on Metop-B.

As you know, we ran several parameter test campaigns using the GRAS instrument on Metop-B over the last few weeks; GRAS/Metop-B is now being operated using the future parameter settings. The characteristics of GRAS/Metop-B data is therefore slightly different from GRAS/Metop-A.

For users of operational GRAS level 1b data products, the main difference in the GRAS/Metop-B data characteristics is that we have enabled the extrapolation of the ionospheric correction into the troposphere beyond altitudes where L2 data is available. As a consequence, bending angle retrievals will reach further down into the troposphere, in particular for rising occultations. As the measurements see more atmospheric multipath below about 8 km impact height, the statistics against, e.g., ECMWF data exhibits larger standard deviations than before, reflecting the limitations of the geometrical retrieval algorithm in the operational GRAS processing at Eumetsat. For upper tropospheric and stratospheric measurements, we have not seen changes in the validation statistics against ECMWF.

The modified instrument parameters have also reduced the number of occultations exhibiting low L2 SNRs, the latter causing large bending angle spikes in the mid and upper stratosphere. We therefore hope that the percentage of occultations rejected for this reason in your quality control schemes will be reduced. I provide more details on the results of the parameter updates below for those of you who are interested.

Note that the ROMS SAF will start to disseminate level 2 products for GRAS/Metop-B soon.

Please let us know if you face any difficulties or problems with the GRAS/Metop-B data. Unless we don’t hear about major difficulties on your side, we intend to update GRAS/Metop-A to the new settings in late April or early May.

With kind regards,


More details:

The parameter test campaigns were addressing the three major issues of the GRAS instrument:

* About 10% of all occultations suffer from poor / low SNR L2 closed loop tracking;
* About 35% of rising occultations suffer from data gaps in the C/A closed loop tracking which are not covered by open loop data;
* A large part in particular rising occultations suffers from data gaps in open loop data.

The first issue is responsible for the majority of GRAS occultations being rejected in NWP applications, usually due to the occurrence of large bending angle spikes at high altitudes.

The figure below demonstrates the improvements on all three issues; “Metop-A” refers to the previous, “Metop-B” to the future settings:

* The number of occultations suffering from low L2 SNR has been reduced from ~10% to less than 2% (overall; top left);
* Uncovered data gaps in C/A closed loop tracking has been reduced from ~35% to < 2% (rising occultations; setting ones were not much affected; top right)
* Open loop data gaps still exist, but the penetration of the highest (contiguous) data segment into the troposphere has been increased by ~ 15km in Straight Line Tangent Altitude (SLTA) for both rising and setting occultations (bottom figure, showing the 10%, 25% and 50% penetration depths in SLTA). At the same time, false tracking events in rising occultations were reduced.

In order to fix the uncovered C/A closed loop data gaps in rising occultations, the upper limit of open loop measurements was raised compared to the previous parameter settings in order to cover the data gaps in the closed loop tracking. Due to the design of the GRAS receiver, this implies that L2 data is only available above ~10 km SLTA / 20-25 km in impact height for the new parameter settings. In order to compensate for this, the operational processing of GRAS/Metop-B data has been reconfigured to extrapolate the ionospheric correction term based on the combined L1 and L2 closed loop bending angles. Comparisons with ECMWF (not shown, but available on request) showed no difference in the statistics against short term forecasts at altitudes above 8 km, suggesting that the ionospheric extrapolation does not cause systematic or random degradations. Below 8 km, standard deviations are slightly increased due to more profiles affected by atmospheric multipath are part of the data set due to the extrapolation of L2 data.

** Note , please email to get this figure **

Thanks for reading till the end;-)


Dr. Christian Marquardt, Remote Sensing Expert, EUMETSAT Eumetsat-Allee 1, D-64295 Darmstadt, Germany
Tel: +49-6151-807-4560, Fax: +49-6151-807-8380
Back to NRT Notifications List

Newer     Older

DMI's logo IEEC's logo Met Office's logo ECMWF's logo WEGC's logo