Ring-shape convective storms

By HansPeter Roesli (EUMETSAT) and Humaid Al Badi (Directorate General of Civil Aviation and Meteorology, Oman)

In the late afternoon on 9 August 2008 deep convection broke out over the southeastern part of Rub al Khali. The area borders the United Arabian Emirates, the Kingdom of Saudi Arabia and the Sultanate of Oman. This tri-border region has no significant orographical features and is very sparsely populated.

Further to the east, along the Hajar mountains in Oman, severe convection is quite common, but over this flat desert area deep convection appears to be very rare — the last similar case was recorded in August 2003. For these reasons the case is quite exceptional. It is also fortunate that the case was covered by the rapid-scanning service of Meteosat-8 (all channels except HRV).

In the absence of orography the convection most probably developed for two reasons: low-level moisture convergence and the passage of a weak easterly disturbance in the upper troposphere. Low-level dry air was moving from west-northwest into the area, tagged by the light blue tongue in the Meteosat-8 Dust RGB Animation (09:00–22:30 UTC). It undercut moist monsoon air over Oman, coloured in darker blue on the same RGB composite. The moisture fields and streamlines of the ECMWF analysis at 12:00 UTC confirm the presence of the two converging air masses at lower levels.

This is also supported by an analysis of the Metop-A IASI spectra. The lower part of the image shows a brightness temperature map of the weak H₂O absorption line at 852.25cm^-1 and two data probes, a blue one in the dry and a magenta one in moist air. The upper part of the image shows the relative spectra at these points in the 800cm^-1 to 1000cm^-1 (10 to 12 microns) region.

Not considering the isolated absorption lines, the blue (dry) curve runs horizontally whereas the magenta curve slightly dips towards the left. Taken together, the diverging behaviour of the two spectra in the 10 to 12 microns IR window is characteristic for the presence of low-level moist air in the region where the cooler spectrum is coming from. The ECMWF streamlines indicate the Gulf of Aden as the moisture source.

The evolution of the convective cells can be followed best in the IR10.8 channel (see Meteosat-8 animation, 09:00–22:30 UTC). Several very cold towers develop in rapid succession, most of them showing so-called cold rings or U-shapes, signalling severe convection leading to heavy downbursts with rain and/or hail and lightning. Most prominent is the northern-most cell where the temperature difference between the cold ring and the warm core reaches values of about 20K.

In the absence of ground truth, the IR10.8 surface temperature field can be checked for cool streaks as indicator of precipitation reaching the ground. To this end the maximum temperature of the IR10.8 channel is extracted from image sequences prior and after the event as well as the maximum temperature in the hours immediately following the event. In fact, comparing the two temperature maps irregular cooler patterns can be observed, thus, it may be assumed that some precipitation did reach the desert area at places.

Later in the night a relatively fast moving arc-shaped, low-level cloud is observed moving from the UAE coast northwards (see Meteosat-8 RGB Night Microphysics animation, 18:15–19:15 UTC). Special enhancement of the Meteosat-8 IR10.8 image reveals 'ripples' of clouds behind the leading edge of this cloud arc (see 18:25 UTC image) better visible in a MODIS image from the same time.

This might be interpreted as an undular bore (see case study Outflow from convective storm over Mauritania). Its ground speed is 40km/h and it might have originated from an outflow from one of the northernmost convective towers. The MODIS image shows even more wavy structures. Their origins are difficult to track, however.

Additional content

Met-9 Dust RGB image with indication of moisture boundary (11:00 UTC)
Met-8 IR10.8 animation (Satellite projection, 09:45–19:00 UTC)