Geoscience Reference
In-Depth Information
Case study
MOZAMBIQUE FLOODS OF 2000
During the early months of 2000 world news
was dominated by the catastrophic flooding that
occurred in southern Africa and Mozambique in
particular. The most poignant image from this
time was the rescuing of a young mother, Sophia
Pedro, with her baby Rosita, born up a tree while
they sought refuge from the flood waters. The
international media coverage of the devastating
flood damage and the rescue operation that fol-
lowed has ensured that this flood will be
remembered for a long time to come. It has given
people the world over a reminder that flooding is
a hydrological hazard capable of spreading
devastation on a huge scale.
The floods of Mozambique were caused by four
storms in succession from January through to
March 2000. The first three months of the year
are the rainy season (or monsoon) for south-eastern
Africa and it is usual for flooding to occur,
although not to the scale witnessed in 2000. The
monsoon started early in southern Mozambique;
the rainfall in Maputo was 70 per cent above
normal for October-November 1999. This meant
that any heavy rainfall later in the rainy season
would be more likely to cause a flood.
The first flood occurred during January 2000
when the Incomáti and Maputo rivers (see Figure
5.14) both burst their banks causing widespread
disruption. The second flood occurred in early
February, as the waters started to recede, except
that now Cyclone Connie brought record rainfall
to southern Mozambique and northern South
Africa. The Limpopo river was as high as ever
recorded (previous high was in 1977) and major
communication lines were cut. The third flood,
21 February until the end of February, occurred
when Cyclone Eline moved inland giving record
rainfall in Zimbabwe and northern South Africa,
causing record-breaking floods. The Limpopo
was 3 m higher than any recorded flood and for
the first time in recorded history the Limpopo and
Incomáti rivers joined together in a huge inunda-
tion. The extent of the flooding can be seen in the
satellite images (see Plates 7 and 8). The fourth
flood was similar in size to the second and occurred
following Cyclone Glória in early March (Christie
and Hanlon, 2001).
There is no doubt that the Mozambique floods
were large and catastrophic. How large they are,
in terms of return periods or average recurrence
intervals (see Chapter 6) is difficult to assess. The
major difficulty is to do with paucity of streamflow
records and problems with measuring flows dur-
ing flood events. On the Incomáti river the flow
records go back to 1937, and this was the largest
flood recorded. For the Limpopo there is some data
back to the 1890s, and again this was the largest
recorded flow event. On the Maputo river to the
south the flood levels were slightly lower than a
1984 event. The difficulties in measuring river-
flow during large flood events is well illustrated
by the failure of many gauging stations to function
properly, either through complete inundation
or being washed away. Christie and Hanlon (2001)
quote an estimate of the flood on the Limpopo
having a 100-year average recurrence interval,
although this is difficult to verify as most gauges
failed. Smithers
et al
. (2001) quote an unpublished
report by Van Bladeren and Van der Spuy (2000)
suggesting that upstream tributaries of the
Incomáti river exceeded the 100-year return
period. Smithers
et al
. (2001) provide an analysis
of the 1-7 day rainfall for the Sabie catchment
(a tributary of the Incomáti) which shows that in
places the 200-year return period was exceeded.
(NB this is an analysis of rainfall records not
riverflow.)
The reasons for the flooding were simple, as
they are in most cases: there was too much rainfall
for the river systems to cope with the resultant
