Experimental 2.5-minute super rapid scans

By Jochen Kerkmann (EUMETSAT), Martin Setvak (CHMI) and Agostino Manzato (OSMER ARPA-FVG)

During the commissioning activities of the MSG-3 satellite, EUMETSAT has carried out (among other regular tests) an experimental 2.5-minute super rapid scan test. According to the original commissioning plans for the MSG-3 satellite, a 3-minute rapid scan was initially scheduled, however following suggestions by the CWG (Convection Working Group) it was decided to modify the test to shorter, 2.5-minute intervals. This interval is exactly half of the present 5-minute RSS (Rapid Scan Service), which is being operationally performed by the MSG-1 (Meteosat-8) satellite. It will also be the native rapid scan interval of the future Meteosat Third Generation Imager (MTG-I) satellites. The 2.5-minute rapid scan test started on 11 September 2012 at 09:00 UTC, and lasted exactly 24 hours. Area covered by the 2.5-minute rapid scan, and the 24-hour loops based on this scan are shown below.

Animations MSG-3 super rapid scan test*

A similar test is planned to be carried out with Meteosat-8 once it is re-located at 3.5° E, providing backup to the Meteosat-9 and Meteosat-10 satellites. The time frame for this is expected to be May/June 2013.

Despite the relatively late time of the year for convective storms, several storms occurred over various parts of Europe, and were captured by the MSG-3 satellite as part of this test. Some of these storms evolved in the morning of 12 September 2012 above north-east Italy (Friuli Venezia Giulia, FVG), being accompanied by heavy rain and hail (see images and animations below).

The 2.5-minute rapid scans capture the evolution of these morning storms from their very initiation, until full maturity. Shortly after sunrise, at the very beginning of the sequence, it is possible to observe a line of several towering cumulus clouds, casting distinct shadows onto lower cloud layers. About one hour later, these rapidly changing clouds begin to merge and evolve into a quickly growing storm system, with new cells forming at its west side (see HRV image from 6:50 UTC). This storm system, which started near Vivaro in the foothills of the Alps, will be called the 'northern' storm. As can be seen from the surface wind measurements (see Figure 4), in the morning hours there was a strong south-east wind (Scirocco) blowing in the southern part of FVG (see Grado, Lignano and Palazzolo stations). This could be the primary channel bringing the moist air of the North Adriatic into the updraft zone. The hypothesis is that the northern storm was triggered by the interaction of this flow with the orographic barrier of the Alps.

About one hour later, at around 07:20 UTC, new convective cells begin to develop in the area of Portogruaro in the southern part of the region, close to the Adriatic coast (see animation of radar images, 07:00–09:55 UTC. Credit: OSMER-ARPA FVG). These cells, which may have been triggered by the convergence of the south-easterly flow with the downdraft from the northern storm and the westerly flow over the Veneto area (see ECMWF 1000hPa wind analysis, 06:00 UTC, Credit: ECMWF), quickly develop to form a second storm system (the southern storm, see MSG-3 HRV image from 08:17 UTC, and Metop-A AVHRR Band 2 image from 8:45 UTC), which, according to the weather radar images, nearly merges with the northern storm at around 8:30 UTC (see sequence of radar images, 08:00–09:00 UTC, source: Mateja Irsic-Zibert, ARSO). After 08:50 UTC the northern storm dissipates quickly while the southern storm continues with strong intensity crossing the southern part of FVG from west to east (strong hail reported on the A4 highway close to Latisana). Most likely, the southern storm 'killed' the northern storm by cutting it off from the south-easterly feeder flow.

It should be noted that the distinction between the two storms is mainly based on weather radar imagery. This is a common problem between radar and satellite meteorology. Often, what appears from satellites as one cell (covered by a large common anvil), in radar appears as multiple cells.

Finally, looking at the Doppler radar images (see LBM animation, 07:45–09:45 UTC. Credit: OSMER-ARPA FVG), the southern storm shows the typical pattern of a large mesocyclone, at least for part of the time, with strong diverging outflow (downburst) later on. This suggests that this storm was probably a supercell, which formed on a convergence line. The convergence line, which is well visible in the ECMWF low-level wind field as shown above, extended over large part of the northern Adriatic Sea as shown in the Metop-A AVHRR sandwich product from 08:45 UTC. Along this line, several new storms formed in the following hours, each new storm developing south of the previous storm: at 08:50 UTC over the Adriatic Sea, at 09:10 UTC close to the coastline of Istra and last but not least at 09:30 UTC close the island of Cres (Croatia).

Unfortunately, the MSG-3 2.5-minute super rapid scans end on 12 September at 09:00 UTC. However, the southern storm lasted for about two more hours, till about 11:00 UTC, when it began to weaken, and ultimately dissipated around 12:00 UTC (see MSG-1 five-minute rapid scan imagery below). From 09:00 till 09:45 UTC the storm exhibited a distinct cold ring, which has later (~ 09:50 UTC) transformed itself into a cold-U shape, disappearing at about 10:25-10:30 UTC.

Animations MSG-1 rapid scans

The 2.5-minute super rapid scan series shows (among other) the big variability of the storm's cloud top, namely of its overshooting tops. Some of these appear in one or two consecutive images only, while some of the others last somewhat longer, or re-appear in the same place as previous overshooting tops. Besides the overshooting tops, the loops also show some other features, such as gravity waves spreading from the center of the storm to its outer parts, or at the end of the sequence a formation of a smaller embedded warm area, which later (after the end of the 2.5-minute super rapid scan) further evolved, forming a cold ring and cold-U shapes later on.

The overshooting tops were one of the main reasons for the 2.5-minute super rapid scan test (and data) request. The 5-minute rapid scan data (RSS) of Meteosat-8 in many cases shows an overshooting top in one single image scan only, with no trace before or after. This poses a certain ambiguity about a typical life cycle of the overshooting tops, namely their duration. The shorter periodicity data such as the 2.5-minute rapid scan by MSG-3, or even shorter-periodicity ones such as those from the US GOES-14 satellite can help to improve our understanding of the overshooting tops, which are frequently used as one of the satellite-based indicators of possible storm severity.

*Note about the speed of the animations: the faster animations (8fps) show better the dynamics of the convective development, while the slower ones (2fps) show better the cloud-top details and their variability.

*Note about the weather radars: the OSMER - ARPA FVG (Fossalon) weather radar is a C-band radar that works in the 5.42-5.63GhZ frequency and has a domain radius of 125km (high spatial resolution). It is located at 25m above mean sea level (msl). The ARSO-Slovenia (Lisca) C-band (5cm) radar has a radius of about 200km and is located at 948m above msl.

Additional content

MSG-3 HRV Animations (12 September, 06:00-08:55 UTC) 8 Frames per second (8fps)
MSG-3 HRV Animations (12 September, 06:00-08:55 UTC) 24 Frames per second (24fps)
MSG-3 IR10.8 animations (12 September, 06:00-08:55 UTC) 2 Frames per second (2fps)
MSG-3 IR10.8 animations (12 September, 06:00-08:55 UTC) 8 Frames per second (8fps)
MSG-3 sandwich product animations (06:00-08:55 UTC) 2 Frames per second (2fps)
MSG-3 sandwich product animations (06:00-08:55 UTC) 8 Frames per second (8fps)
MSG-3 storm sandwich product animations (06:00-08:55 UTC) 2 Frames per second (2fps)
MSG-3 storm sandwich product animations (06:00-08:55 UTC) 8 Frames per second (8fps)

Met-9 Ice Particle Size (Convection) RGB image & cloud droplet effective radius product (09:00 UTC, Credit: KNMI)
Metop-A, AVHRR, Band 2 image (08:45 UTC)
Metop-A, AVHRR, Band 4 image (08:45 UTC)
Metop-A, AVHRR, sandwich product (08:45 UTC)

Convection Working Group - a web page devoted to atmospheric convection from many aspects, from case studies to technical documents and scientific articles (a joint initiative of EUMETSAT and ESSL).

References

K.M. Bedka, J.C. Brunner, R. Dworak, W.F. Feltz, J. Otkin, T. Greenwald, 2010: Objective satellite-based overshooting top detection using infrared window channel brightness temperature gradient. J. Appl. Meteorol. Climatol., 49, pp. 181-202.
Kristopher M. Bedka, 2011: Overshooting cloud top detections using MSG SEVIRI Infrared brightness temperatures and their relationship to severe weather over Europe. Atmos. Research 99, pp 175-189,
http://dx.doi.org/10.1016/j.atmosres.2010.10.001