IRI’s Role in South Africa’s Seasonal Climate Forecast Operations

Several decades of climate research have shown seasonal temperature and rainfall patterns over southern Africa to be predictable months in advance. While scientists recognized the importance of the El Niño Southern Oscillation on seasonal climate variability in this region during the 1980s, South Africa first began issuing regular seasonal forecasts in the early 1990s. Over the past twenty years, enhanced modeling systems have greatly improved the accuracy and lead-times of these climate predictions – in large part due to IRI’s partnership with the South African Weather Service and affiliated institutions.

In a paper in the journal Earth Perspectives, Willem Landman – chief researcher at the Council of Scientific and Industrial Research and adjunct research scientist at the IRI – describes the evolution of seasonal climate prediction in South Africa.

In the early 1990s, Landman participated in a nine-month training course in seasonal climate prediction at the Lamont-Doherty Earth Observatory. This statistical training formed the basis of new modeling techniques that used global sea-surface temperature anomalies to predict seasonal rainfall outcomes in South Africa.

Subsequent research published by IRI helped the South African Weather Service develop a multi-tiered forecast system for seasonal climate prediction. This enhanced forecasting system was based on both statistical and dynamical modeling approaches. One of IRI’s atmospheric general circulation models, the ECHAM4.5, showed such predictive skill that the South African Weather Service began implementing and using the model in its seasonal climate forecasts. Today, South Africa is certified by the World Meteorological Organization as a Global Producing Centre for Long-Range Forecasting, a recognition partly obtained through collaboration with IRI climate modelers.

Among IRI’s contributions to climate prediction over southern Africa since the 1990s, the Climate Predictability Tool (CPT) has been at the forefront of seasonal forecasting efforts. This software package, developed by IRI, combines downscaled global circulation models with statistical forecasting methods, and has been used extensively by the South African Weather Service and the Council for Scientific and Industrial Research.

In an interview with IRI, Landman describes how CPT’s seasonal rainfall forecasts over southern Africa could potentially help water managers to model streamflow patterns in Lake Kariba along the border of Zambia and Zimbabwe. CPT forecasts are generated by an ensemble of global models and show high level of predictive skill based on multi-model validation.

“The Climate Predictability Tool has been key in helping southern Africa and other developing regions build their capacity to create scientifically sound climate forecasts,” says Landman.

Today, the forecasting success of the CPT has led South Africa’s National Research Foundation to help fund its development, demonstrating IRI’s role in climate capacity-building in developing regions.

An entire special issue of Earth Perspectives is dedicated to the work and mission of the International Research Institute for Climate and Society. Read more about this here.


Landman, W. A. (2014). How the International Research Institute for Climate and Society has contributed towards seasonal climate forecast modelling and operations in South Africa. Earth Perspectives1, 1-13.

Mason, S. J. (2011). Seasonal Forecasting Using the Climate Predictability Tool. Science and Technology Infusion Climate Bulletin, 36th NOAA Annual Climate Diagnostics and Prediction Workshop, 180-182.


Photo credit: Werner Bayer, under Creative Commons license.