FLOODING IN SADC:

AN OVERVIEW

Eugene Poolman
South African Weather Service
 OUTLINE OF PRESENTATION

• Global perspective of floods

• Flooding in SADC - Historical patterns
• Flooding in SADC – Future patterns
• Status of flood forecasting in SADC
• Way forward
Global Perspective of Floods
The number of flood disasters by country
from 1974 to 2003
• Floods are worldwide a major hazard
• Cause death of on average 7000 people every year
• WMO – of all natural hazards:
– Floods has greatest damage potential
(over 1/3 of total damage)
– Affect the greatest number of people (over
2/3 of people affected)
• Impact is increasing rapidly
Frequency of worldwide flood disasters
 Impact on Developing Countries

• Developing countries are more vulnerable:

– Social and economic vulnerability
– Reduced coping mechanisms
• Rural people are most vulnerable
• 95% of deaths occurred in least developing countries
• 8 SADC countries are least developing countries
Flooding in SADC –
Historical Patterns
 Flooding events between 1985 and 2002

• Occur mostly along

southern and
eastern parts of
the subcontinent
• Caused by
– tropical weather
systems moving
westwards in
summer
– Cold fronts and cut-
off lows in winter
• In SADC Floods follows
second to epidemics in
frequency
• Third to drought and wind
storms on damage caused
– Wind storms mostly tropical
cyclones, and winter frontal
systems that also cause floods
• R7,000 million flood damage
caused by tropical cyclone
Eline and Gloria in 2000 in
Mozambique
– 3 times their 1999 export
 Impact on SA, Mozambique, Zambia

• Frequency of natural disasters:

– Most frequent natural disaster in SA and Mozambique,
second to epidemics in Zambia
• Deaths from natural disasters:
– Mozambique 37% of deaths, SA 62%, Zambia unclear
• Affecting people (injured, displaced):
– Drought affects most people (slow and long-lasting disaster)
– Floods second in Zambia 50% (about 3 million people) and in
Mozambique 33% (about 8 million people)
• In Mozambique about 30 people per square km stay in
flood prone river basins
 Trends of flood occurrences in SADC

• Following global trends, floods have increased in SADC

– 1984 - 1988: 9
– 1994 – 1998: 26
– 1999 – 2003: 59
• Dramatic increase in number of people affected and died:
– 1984 – 1988: 4000
– 1998 – 2003: 13000
• Particularly elderly, women and children are affected
• This upward trend is due to
– Growth in urban population
– Population growth in flood prone river basins (nutritious and water
rich)
Flooding in SADC –
Future Patterns
 From reports by the IPCC Assessment Report 4,
and WMO:

• Great deal of uncertainty on the impact of climate

change on flooding in SADC
• Flood risk is also controlled by other factors:
– Changes in economic and social systems
– Changes of hydrological aspects such as vegetation and soil
moisture in basins
– Land-use changes
– Deforestation
– Reduction in wetlands

IPCC: Intergovernmental Panel on Climate Change

• IPCC: Precipitation intensity expected to increase
virtually everywhere (though parts of SA
expected to be drier and warmer)
– “A warmer climate with increased variability will increase
the risk of both floods and droughts”
– “Increase in precipitation intensity and frequency will
lead to increase in risk of flooding and droughts in many
areas”
– Flood magnitudes and frequencies are likely to increase in
most regions
Status of Flood Forecasting
in SADC
• Distinguish broadly between two kinds of floods: Large
scale river floods, and Flash floods
• Large scale river floods:
– Occur in large river basins from large scale heavy rain (Orange,
Vaal, Limpopo, Zambezi Rivers)
– Long response time (6 hours to many days) between heavy rain
and flooding
– Complex hydrologic models calculating river levels as the flood
moves down the river over the next few days
• Situation in SADC:
– Outside SA there is a general lack of river forecast systems
due to limited rainfall data and river monitoring data
– SA is implementing a river forecasting system in the Vaal River
– USA developed a crude satellite based system for Limpopo
following 2000 floods
– Some effort in Zambezi
Limpopo River Basin
• Flash floods:
– Occur in small river basins (50 – 200 square km)
– Quick response (<6 hours) between heavy rain and flooding
– Traditional hydrologic models ineffective due to small lead time
– Rainfall measurements on hourly basis using rain gauges and
weather radars and satellites are very important
– Promising system implemented in Central America
• Situation in SADC
– There is currently no flash flood warning system operational
– Real time rainfall measurement only in South Africa
– Very few weather radars outside SA (Mozambique has 2,
Botswana has 1) and their coverage not sufficient
– SA is investigating implementation of Central American system
under radar coverage in flood prone regions
Example of the Central American flash flood
warning system
Way Forward
 1. Integrated Flood Management as a
comprehensive solution

• Munich Re: “It is not possible to prevent disasters, we

rather need to learn to live with them”
• WMO developed an Integrated Flood Management
program as part of a holistic disaster risk management
approach
– Early warning systems
– Socio-economic activities to balance needs and risks
– Structural solutions such as levees and dams
– Non-structural measures such as land-use regulation
– Awareness raising of population and governments
• This approach should be strongly supported
 2. Early Warning through Flood
Forecasting
• WMO emphasizes that the most critical aspect of
a flood-loss reduction strategy, lies in emergency
preparedness and response, particularly a flood
forecasting and warning system
• This requires collaboration between
meteorologists, hydrologists and disaster
managers, and between countries
• Forecasting for large scale river floods:
1. Development of comprehensive hydrologic river forecasting
models for the large basins (Limpopo, Zambesi, etc)
• Long-term solution
• Depends heavily on the improvement of rainfall measurement
and river level monitoring in all riparian countries of river
basins
• An example is the efforts by the Limpopo Basin Commission

2. Investigate more crude interim measure such as the US

satellite based system to provide some kind of guidance
• Easier to implement since it does not depend on surface data
• But, it is less accurate because it uses only remote
observations by satellite and not real data
• Forecasting for flash floods:
1. South Africa is investigating the implementation of a
flash flood warning system based on the Central
American system
• Can operate under the radar coverage in flash flood prone
basins of 50 to 100 square km
• Complement the SAWS current early warning system

2. WMO is planning the development of a global flash

flood warning system (FFGS) based on the Central
American system, but using only satellite information
• Can be rolled out where no surface data exists
• Can only work for larger basins of about 200 square km
• Will be implemented on a regional basis (like SADC) in
2010
• SAWS as a RSMC can play a pivotal role in this activity
 3. Enhancing data measuring

• Reliable weather and hydrological data is an

essential system for flood forecasting systems
– Particularly rainfall and river flow and water level
monitoring
• WMO operates the SADC-HYCOS program aimed
at improving hydrological monitoring.
• National meteorological services must continuously
improve their existing rainfall networks
substantially
• This depends heavily on funding levels
FOR THERE IS NO NOAH AMONG US
For there is no Noah among us, let us do it ourselves.
Let us not attempt to cross flooded rivers.
Let us not build our homes on river banks.
Let us not build our homes on low land.
Let us fill gullies and stop them from becoming rivers.
For there is no Noah among us, let us protect
ourselves from floods

Poem by Nokuthula Ndou

Grade 6 Pupil
Shashi Primary School
Maramani Communal Areas
Beit Bridge
Zimbabwe