Experts emphasise the urgent need for Africa-specific climate models to improve forecasts and early warnings of extreme weather events.
The recent floods in southern Africa highlight the inadequacy of current models, which fail to capture regional nuances.
Locally developed models could better predict and attribute such events, aiding vulnerable communities in Mozambique, Eswatini, and beyond.
A climate change attribution analysis of the recent southern African floods across Mozambique, Eswatini, north-eastern South Africa and Zimbabwe that killed 200 people and affected thousands throws up concerns that require immediate attention.
The primary concern is the urgent need for weather and climate models developed in African countries for the region to enable better forecasts of extreme weather events and their attribution to warming and consequent climate change, according to experts of the World Weather Attribution (WWA) who conducted the attribution analysis.
WWA is a consortium of climate scientists from around the world which conducts research to understand and quantify the role played by human-induced climate change in the occurrence, frequency and intensity of extreme weather events, known as attribution studies.
In many of the affected regions, the communities are mostly dependent upon agriculture and live in extremely poor conditions, especially in Mozambique and Eswatini. “Recurrent flooding and other natural hazards have trapped rural communities in the lower Limpopo River basin in a cycle of poverty. As a result, flood impacts are disproportionately severe for low-income and marginalised communities,” according to the study.
“Historical mining practices, weak environmental regulation and limited awareness of long-term impacts have entrenched persistent and severe damage to aquatic ecosystems which, in combination with long term low investment and low maintenance of aging infrastructure increased exposure dramatically,” wrote the study authors. “Pre-existing food insecurity, driven by droughts and other hazards, will be sharply amplified by the floods affecting highly exposed agricultural communities in Mozambique and Eswatini,” they added.
The WWA scientists analysed observations of a 10-day heavy rainfall event in the December-February period, which is the primary rainfall season for the region, of the previous years. They found that such an event would occur once in every 50 years in a world warmer than pre-industrial period (1850-1900) by 1.3°C. Even though this was a rare event on a climate changed planet, it would have been even rarer without warming, according to the study.
The study also found that rainfall beginning in the end of December and intensifying in the first week of January has been made 40 per cent more intense by warming.
Another major factor in the likelihood and intensity of the rains was the weak La Nina conditions currently prevalent in the Equatorial Pacific Ocean. La Nina is the cooler-than-normal phase of the El Nino Southern Oscillation phenomenon and generally brings good rainfall to southern Africa. La Nina conditions intensified the rains by 22 per cent, according to the study.
These findings were concluded by WWA scientists using observations. They wanted to validate the observations with climate models and also quantify the exact role played by warming and consequent climate change in the intensification and likelihood of the event. This proved difficult for them as none of the models performed well enough for the validation or the quantification.
“The climate models we tested showed varying range of results for the 10-day rainfall event. They did not capture the regional changes,” said Izidine Pinto, senior climate researcher, KNMI Royal Netherlands Meteorological Institute and a WWA expert, in a press conference.
Further, “the climate models were not able to capture the impact of the La Niña even though the impact is quite clear from the observations and what we understand and expect from physics,” said Friederike Otto, professor of Climate Science at the Centre for Environmental Policy, Imperial College London, also part of the WWA. La Niña determines when and where it is going to rain and climate change determines how bad the rain is going to be, according to Pinto.
“The climate models that we can have and are freely available are all developed outside of Africa. The models have their strengths and weaknesses and cannot get the climate and its changes correctly for the entire planet,” said Otto. “So the models are usually better designed for the regions that they are originally meant for,” she added.
Tropical climates are also more complex and have many more contributing factors and interacting systems that are not usually well represented in models, according to Otto. “Better climate models developed regionally for Africa would tell us how weather systems are being affected by warming”, she concluded.
There are other localised nuances which increase the vulnerabilities of certain regions that climate models cannot capture well in general. “The flooding in Mozambique happened during a high-tide event along the coast, which prevented the flood water from the rivers from entering the sea,” said Bernardino Nantumbo, climate researcher, INAM National Institute of Meteorology, Maputo, Mozambique and a WWA expert.
“Improved weather models developed in the region would also provide better forecasts which can be useful for early warning systems for such flooding events. Funding from governments is needed for this,” said Pinto.