Tropical low over southern Africa

Tropical low over southern Africa

17 January 2013 00:00 UTC

Tropical low over southern Africa
Tropical low over southern Africa

A tropical low began forming over the northern parts of Mozambique on 6 January.

Last Updated

22 October 2020

Published on

16 January 2013

by Cassandra Pringle, Bathobile Maseko (South African Weather Service (SAWS) ), HansPeter Roesli and Jochen Kerkmann (EUMETSAT)

The tropical low is the product of intense heating over the land surface and an influx of moisture. The system remained quasi-stationary over northern Mozambique. It appeared to strengthen during the day, with some explosively developing convective cells appearing in the afternoon, a definite indication that diurnal temperature changes were a direct influence on the development and sustainability of the system. From 14 January, the system slowly started moving westward.

On 17–18 January it stopped tracking westward and remained quasi-stationary over the north-western parts of Botswana. On 19 January it started tracking eastwards and by 21 January it left the continent where it eventually dissipated over Madagascar (see sequence of daily (6 UTC) Meteosat-9 Airmass RGB images ).

On 17 January, an extra-tropical cyclone started to move over South Africa. In the late evening the trough associated with this system came into contact with the tropical low pressure system over Botswana. As they got closer, the amount of cloud increased dramatically, due the merging of these two systems and the instability associated with both systems. As the extra-tropical cyclone tracked further eastward, the tropical low pressure system also started tracking eastwards. This brought widespread rain to eastern parts of South Africa and Botswana, Zimbabwe, Swaziland and Mozambique. This rain fell from 18–20 of January, with much flooding occurring in the Limpopo and Mpumalanga provinces of South Africa. On 20 January, another extra-tropical system came into contact with the tropical low causing the heaviest rainfall in South Africa. Several people died due to flooding and houses caving in. Many people were stranded on rooftops and had to be airlifted to safety (SAPA, 2013a; 2013b).

This analysis of this tropical low pressure system involves satellite imagery and the Hydro-Estimator (H-E). The Hydro-Estimator was developed to be independent of radar input and is useful for convective events with cold cloud tops (Vasiliff et al., 2007). The principle of the HE is to determine if a pixel is colder than the surroundings. If it is, then the pixel is classified as rainy cloud (Scofield & Kuligowski, 2003). The Hydro-Estimator uses the IR10.8 channel, together with numerical weather prediction model fields from the local version of the Unified Model. The assumption that the cloud height is related to cloud thickness holds well for active convective clouds but is less effective when applied to stratiform clouds. The H-E does tend to overestimate rainfall intensity for slow-moving storms with cold tops whilst it underestimates warm stratiform cloud systems.

On 17 January at 15:00 UTC, the day natural colours RGB (Figure 1) shows a defined spiral structure to the tropical low. There are many convective cells associated with the system as well. Some of the cells have large cirrus outflows which can cause overestimations of the rainfall values of the H-E. It is also evident that the extensive cyclonic banding associated with this system has influenced the airflow pattern over the majority of southern Africa.

In the airmass RGB imagery for 20 January at 12:00 UTC (Figure 2), the structure of the tropical low is no longer symmetrical. This is due to the eastward tracking extra-tropical low pressure system forcing the tropical low to be similarly displaced eastwards. The disruption to its structure is causing the system to lose energy and is reducing its ability to sustain itself.

From early afternoon of 20 January, there was explosive development of convective cells north of the tropical low. The storms were drawn towards the centre of the tropical low pressure system. By 15:00 UTC, there was an MCC at the centre of the tropical low (Figure 3) with cloud top temperatures less than -75 °C. The highest rainfall recordings for the Limpopo province occurred on this day, with Thohoyandou recording 282 mm and Makhado recording 294 mm in a 24 hour period. Many other stations for the province recorded more than 100 mm for the day. These high values in addition to the rainfall that fell on 15, 16 and 19 January, combined to cause river banks to burst and there was widespread flooding.

The Hydro-Estimator (Figure 4) correctly identified the very large rainfall values over the Limpopo province for 20 January. The rainfall values at many stations far exceeded the values of the Hydro-Estimator scale. This indicates that the Hydro-Estimator performed well on this particular day in estimating rainfall values.


Figure 1: Meteosat-9 Natural Colour RGB Image

Met-9, 17 January 2013, 15:00 UTC
RGB Composite NIR1.6, VIS0.8, VIS0.6
Full Resolution
Animation HRV channel (06:00-13:00 UTC)

Related Content

Sequence of daily (6 UTC) Meteosat-9 Airmass RGB images
Tropical Low over Namibia-Botswana border area (24 January 2010)


Figure 2: Meteosat-9 Airmass RGB Image

Met-9, 20 January 2013, 12:00 UTC
RGB Composite WV6.2-WV7.3, IR9.7-IR10.8, WV6.2
Full Resolution
Animation (16 Jan 00:00 - 17 Jan 09:00 UTC)
Animation (Airmass RGB blended with IR10.8)
(15 Jan 00:00 - 22 Jan 23:00 UTC)


Figure 3: Meteosat-9 IR10.8 Image

Met-9, 20 January 2013, 17:00 UTC
Channel 09 (IR10.8, colour enhanced)
Full Resolution


Figure 4: Hydro-Estimator Product

Hydro-estimator image of total 24 hour rainfall estimates for 20 January 2013
Full Resolution



SAPA, 2013a: Limpopo flood rescue resumes, 21 January, accessed on 21 January 2013
SAPA, 2013b: 6 dead after Limpopo River bursts banks, 21 January, Johannesburg, accessed on 21 January 2013

Scofield, R. A., and R. J. Kuligowski, 2003: Status and outlook of operational satellite precipitation algorithms for extreme-precipitation events, National Environmental Satellite, Data and Information Service, Camp Springs, Maryland.

Vasiloff, S.V., D. Seo, K. W. Howard, J. Zhang, D. H. Kitzmiller, M. G. Mullusky, W. F. Krajewski, E. A. Brandes, R. M. Rabin, D. S. Berkowitz, H. E. Brooks., J. A. McGinley, R. J. Kuligowski, and B. G. Brown, 2007: Improving QPR and very short term QPF, American Meteorological Society.