The Meteosat-9 image below shows the case of a tropical low over the Namibia-Botswana border area that occurred in January 2010.
More information and detailed analysis of the feature can be found in the In Depth section.
by Jochen Kerkmann, Jose Prieto (EUMETSAT) and Lee-ann Simpson (South African Weather Service)
People don't normally use the words 'Namibia' and 'Tropical Cyclone' in the same breath. The south Atlantic with its cooler waters (when compared to the North Atlantic) is generally not thought of by meteorologists as a place where tropical cyclones can form (see locations where hurricanes began their development, data is for the period 1950–2000, JPG, 43 KB, source PhysicalGeography.net).
In the south Atlantic, cool ocean temperatures or the presence of temperature inversions generally restrict the formation of Tropical Cyclones. The only exceptions observed so far were Hurricane Catarina, which made landfall in southeastern Brazil on 28 March 2004, and Tropical Storm Anita in March 2010.
However, tropical cyclones do frequently form over the southwest Indian Ocean. Most of them recurve near the eastern African coast and move away to the southeast as extra-tropical cyclones, but a good number of them are known to cross the African coast (often in Mozambique) and weaken rapidly due to friction and insufficient moisture over land.
A few have been observed to travel a considerable distance inland bringing substantial rainfall over arid to semiarid areas like Namibia and Botswana (e.g. tropical cyclone Eline in February 2000, Reason and Keibel, 2004, PDF, 6 MB), and in exceptional cases they can traverse the entire breadth of the southern African subcontinent, from the Indian Ocean to the Atlantic coast (e.g. tropical cyclone Bonita in January 1996, Mudenda and Mumba, 2004, PDF, 395 KB).
In addition to Indian Ocean tropical cyclones that travel far inland, tropical lows (depressions) can also form inland over Zambia-Angola or Botswana-Namibia along the ITCZ when there is sufficient moisture available, i.e. during the rainy season from December to March. Strong, persistent convection with latent heat release at mid levels can lead to a warm-core tropical low with relatively strong winds at the surface (see diagram, JPG, 157 KB, source: B. Rosting (met.no)).
Depending on the wind shear situation and upper level divergence, such tropical lows can have the appearance of a Tropical Storm in satellite imagery, with a strong low-level circulation centre and spiral upper-level cloud bands. Due to friction, the surface winds do not reach hurricane intensity, but precipitation amounts can be very high, which is why such systems play a major role in the Severe Weather Forecasting Demonstration Project (SWFDP) for Southern Africa.
The tropical low seen by Meteosat-9 had a distinct spiral structure with embedded convective cells, but no eye. On the following day, as the low progressed through the central parts of South Africa, the intense and prolonged convective activity resulted in heavy rainfall being recorded at many stations across the region.
Meteosat-9 RGB Composite
Met-9, 24 January 2010, 10:30 UTC
RGB Composite VIS0.8, NIR1.6, IR3.9
Full Resolution (JPG, 163 KB)
Animation RGB Natural Colours (07:00–16:00 UTC, MPG, 3 MB)
Sequence of daily (12 UTC) Airmass RGB images (PDF, 2 MB)
In South Africa, heavy rainfall is defined as more than 50 mm rainfall being recorded at a station in a 24-hour period. The central and western interior of South Africa is a very dry part of the country, and on the 25th some stations within this region were reporting over 80 mm of rain for the day, with the majority of the rainfall figures being above 20 mm for the northern and central provinces of South Africa.
On 26 January, the low had moved to the eastern interior of South Africa, allowing the western parts of the country to start drying out, while the central regions continued to receive heavy rainfall. A rain gauge situated just north of Lesotho measured 120 mm of rain on the 26th, while the rest of central and eastern South Africa received in excess of 20–30 mm of rain that day. The rainfall continued through the eastern interior overnight on the 26th, into 27 January, when the low moved away from South Africa.
Besides the immediate impact of so much rainfall over the western, central and eastern interior of South Africa, which can cause localised flooding, a larger scale effect was the increasing levels of the main storage dams across the eastern part of South Africa. By the end of January 2010, the Vaal Dam and the Sterkfontein Dam, were both at maximum capacity. Both of these dams can be classified as two of the biggest dams in the country. Due to a generally good summer rainfall season and persistent rainfall, the dams were operating at near full capacity, and the heavy rain towards the end of January could have contributed to the eventual flooding of the Vaal Dam on 29/30 January 2010.
A sequence of daily (12 UTC) Meteosat-9 Airmass RGB images (PDF, 2 MB) suggests that the tropical low, after crossing South Africa and Mozambique, regained strength over the warm waters of the Mozambique Channel growing into a Tropical Storm named Fami, which dissipated soon after making landfall in Madagascar. This track, i.e. the connection between the tropical low that started over Namibia-Botswana and Tropical Storm Fami that hit Madagascar, has yet to be confirmed using more frequent Meteosat imagery.
Finally, other interesting cases of tropical lows over Southern Africa have been observed by Meteosat on 8 January 2008 (see animation, MPG, 3 MB) and 5 February 2010 (see animation, MPG, 5 MB). In the latter case, the tropical low crossed northern Namibia bringing unusual rainfall to the coastal areas before dissipating over the cold waters of the South Atlantic (see Metop-A AVHRR image, 9 Feb 2010, 08:37 UTC, JPG, 4 MB).