by Martin Setvak (CHMI) and Alois Sokol (Comenius University, Bratislava, Slovakia)Jump to images
Three mesoscale convective systems (MCS, for a detailled definition and further reading about MCS see e.g. paper from Robert A. Houze) developed over Central Europe between 9 and 13 July 201, each of the systems being significantly different from the other. Probably the most spectacular, from the satellite perspective, was the last one (see image below), which exhibited various storm-top features, well seen in rapid scan imagery from Meteosat-8 (MSG-1) shortly before sunset. However, let's start the story from the beginning, with the first two MCSs ...
Mesoscale convective systems of 9-11 July 2011
The first of the mesoscale convective systems began to form shortly after sunset on 9 July 2011, above Bavaria and southwest of the Czech Republic. Around midnight 9/10 July, its cloud shield covered most of the Czech Republic and by early morning the system moved to southwest Poland, where it dissipated before local noon. The full evolution of this system (and the one that followed) can be tracked in the Meteosat-8 rapid scan loop, covering the period from 9 July 12 UTC till 11 July 06 UTC - either in the better quality MOV Animation (H.264 encoded, 26 MB) or the lower quality MPG Animation (easier to play, 20 MB). The second of the systems became organized one day later around sunset, above Bavaria and northeast of Austria, moving to the Czech Republic. Here it attained its peak activity (as observed in the IR10.8 band) around 21-22 UTC, again covering most of the Czech Republic with its cloud shield, and then moving, close to midnight, to Poland where it quickly dissipated. Though this second system had many of the features of a mesoscale convective complex (MCC) as defined in 1980 by R. Maddox, it did not last long enough to formally meet the MCC definition criteria.
The two systems differed substantially in the appearance of their cloud tops in IR10.8 satellite imagery and their radar structures. The first of these was much less "organized", with significantly warmer and rather chaotic cloud-top structure when considering the brightness temperature (BT) minima; however these minima were located close to the radar reflectivity cores, as seen in this image (PNG, 381 KB), which compares the satellite IR10.8 BT image and the matching radar reflectivity image. The evolution of the system in radar imagery is shown in this loop: MOV (17 MB) or MPG (7 MB). The second system (10/11 July) was apparently colder, with the BT minima forming one large area at the center of the cloud shield, and the radar reflectivity image showing the high reflectivity organized into linear, curved structures (see again the comparison (PNG, 378 KB) of the MSG and radar appearance, and the radar loop, either the MOV (16 MB) or the MPG (6 MB) format). For both cases, the mesoscale cyclonic rotation was obvious in the radar imagery, however the mesoscale vortex was much better developed in the case of the second MCS. This low- to mid-level cyclonic circulation should not be confused with the upper-level anticyclonic divergence outflow patterns, best seen in the MSG WV6.2 band loops, showing the highest clouds only - see either the MOV (20 MB) or MPG (19 MB) loop for the entire period.
Mesoscale convective system of 13 July 2011
An advantage, when viewing this system (as compared to the previous two) was that it had already reached a mature state by late afternoon, and before sunset. Therefore, it was also possible to use the HRV band and, especially, its combination with the IR10.8 band, called the sandwich product, to show some of the spectacular details of its cloud-top structure. The appearance and evolution of this MCS is given in the following loops:
Animations (Meteosat-8 rapid scan service, 5-minute intervals):
Note that the IR10.8 loop shows a somewhat longer time period than the sandwich or HRV loops, which terminate at sunset. The radar loops show the distinct bow-echo structure of this storm, MOV (10 MB) or MPG (27 MB). A comparison of the appearance of the storm in the MSG sandwich image and radar image is here (PNG, 863 KB). Due to the favorable timing of the system, it was well documented by Czech storm chasers; their photos and time-lapse movies of a spectacular shelf-cloud associated with the bow-echo are available here. And a short time-lapse movie showing the passage of a gust front of the system above Prague is available here.
There are several distinct cloud-top features which are clearly seen in the sandwich product (image at the top of this page) and its loop. Probably the best known of these (and easiest to see) are the gravity waves, spreading almost concentrically from the center of the system. Another significant feature of this MCS is the "radial cirrus" - narrow bands of above-anvil cirrus (these bands cast shadows on the underlying anvil top, which means that they are vertically separated from the rest of the storm top), and slowly rotating clock-wise with the storm top. Another interesting feature in the sandwich product are the two anticyclonally-curved cold bands (red in the sandwich image), forming an "S" sign in the image above. The material of this cold "S" began to spread westward, quite rapidly, from the overshooting tops region at about 17:00 UTC - see the following short loop showing this episode, in MOV or MPG format. The eastern part of the "S" formed later on, around 18:30 UTC. The reason for the much lower temperature of the cirrus within this cold "S" remains unclear, while its "S" shape is most likely a result of the anticyclonic rotation of the cloud top, resulting from the strong upper-level divergence (and the Coriolis force).
Final comments - satellite IR imagery and radar-determined storm structure
The three MCS cases occurring in such a short time apart, nicely document the big variability as regards the link between the cloud-top brightness temperature and the storm structure as observed by weather radars. In some cases, the satellite-observed cloud-top BT minima (in most cases manifesting the overshooting tops) correspond well to the location of radar reflectivity cores, as in the 9 July case (PNG, 381 KB). In other cases, the position of the radar reflectivity cores match the locations of the overshooting tops, which may not always appear as cold - e.g. in the 13 July case (PNG, 863 KB) the overshooting top northwest of the central part of the "S" (and several others shown in the loop of the sandwich product). At the same time, the large area of the cold "S" (or other similar cold cloud-top features not directly representing the overshooting tops) may not be directly associated with increased precipitation directly below. This is even more true for the 10 July case (PNG, 378 KB) - here the coldest area is located above the center of the forming mesoscale vortex, whilst the strongest precipitation, taking the form of a distinct high-reflectivity line, is located in the south-east part of the system.
Cases like these, with a rather loose linkage between cloud-top temperature and radar-determined storm structure, or even relatively warm overshooting tops, continue to pose a real challenge for automatic satellite-based precipitation estimate techniques or overshooting tops detection algorithms.
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 to mesoscale convective systems and related weather phenomena at Wikipedia
C. Morel, S. Senesi, 2001: A climatology of mesoscale convective systems over Europe using satellite infrared imagery. Part I, Part II
K. Bedka, 2011: Overshooting cloud top detections using MSG SEVIRI Infrared brightness temperatures and their relationship to severe weather over Europe
If you have any questions or queries about EUMETSAT, its products and services, please .