For decades, weather forecasts across Africa spoke in the language of certainty. But in an era of disrupted seasons and intensifying extremes, scientists and journalists are learning that explaining what might happen, rather than what will, could be the difference between preparedness and catastrophe
Across eastern and southern Africa, a quiet but consequential shift is underway in newsrooms and meteorological centers alike. Journalists are being asked to master a difficult craft, reporting uncertainty, at a moment when climate shocks are arriving faster, harder, and with less predictability than ever before.
The stakes extend far beyond professional practice. The ability to explain risk rather than promise outcomes may determine whether farmers protect their harvests, governments prepare for disasters, and communities endure what is increasingly becoming a volatile environmental future.
During a recent interview in climate-vulnerable northern Kenya, award-winning broadcaster Susan Mbodze leaned forward in her chair, reflecting on a career shaped by confident forecasts that often proved fragile.
“All my career, I have been given forecasts that speak like commandments. The rains will start on October 15. The season will be poor.
But the farmers I talk to don’t live in a world of ‘will’ and ‘won’t’. They live in a world of ‘might’ and ‘maybe’, “said Susan.
Her observation captures a growing communication crisis. For decades, much of Africa’s public weather information has relied on deterministic forecasts, single-outcome predictions that imply a fixed future. But in a climate-altered world defined by shifting seasons and compounding extremes, such certainty is no longer merely misleading; it can be dangerous.
A quiet revolution is now unfolding across the continent’s climate institutions and newsrooms. At its middle is probabilistic forecasting, an approach that does not predict exactly what will happen but instead quantifies what could happen and with what likelihood.
“This is not an academic shift,” Collison Lore, a climate adaptation expert and former regional coordinator for the World Meteorological Organization (WMO) told Impact AI News.
“For journalists, governments and farmers in the world’s most climate-vulnerable regions, understanding probability is now a matter of food security, economic survival and ethical reporting.”
The consequences of misplaced certainty are etched into recent climate history. In 2021, parts of the Horn of Africa were advised to prepare for a poor rainy season, prompting communities to brace for drought. Yet embedded within the seasonal outlook was a low-probability, high-impact risk of extreme rainfall, one that was barely communicated.
When the rains arrived, they did so violently, unleashing floods across drought-weary landscapes and destroying crops, homes, and infrastructure.
“We failed to communicate the full story,” admitted Guleid Artan, director of the IGAD Climate Prediction and Applications Centre (ICPAC), which issues regional seasonal forecasts for eastern Africa. “We focused on the most likely outcome and unintentionally silenced the less likely, but catastrophic, alternative.”
Similar misalignments have played out from Malawi to Mozambique. Farmers plant based on an expected onset date that never comes. Governments pre-position flood resources in years dominated by drought, or prepare for drought only to face inundation.
Scientists insist the problem is not a lack of data but a failure to translate uncertainty honestly.
Behind every modern seasonal outlook lies an ensemble, not one forecast but dozens or even hundreds of simulations generated by climate models, each beginning with slightly different atmospheric conditions.
“Think of it as consulting 100 expert meteorologists, each with a slightly different interpretation of the same initial data,” explained Dr Hussen Seid, a climate modeler at ICPAC.
“If 80 of them indicate below-normal rainfall, we don’t say drought is coming. We say there is an 80 percent probability of below-normal rains. The 20 percent who disagree are not wrong, they represent real alternative futures.”
This spread of outcomes is not a weakness. It is scientific honesty.
The World Meteorological Organization now promotes ensemble and probabilistic forecasting as global best practice, recognizing that the climate system is inherently chaotic, especially under human-driven warming.
For journalists, this evolution demands more than technical literacy; it requires a shift in professional identity. No longer mere messengers of prediction, reporters must become translators of risk.
Climate Frontline, a nonprofit that trains journalists in climate-vulnerable regions and works closely with forecasters, promotes a three-part framework for responsible reporting: lead with the most likely outcome, explain the range of possibilities, and highlight high-impact risks.
“As communicators, we must separate likelihood from impact,” said Dr Joyce Kimutai, a Kenyan climate scientist and contributing author to the Intergovernmental Panel on Climate Change (IPCC). “A low-probability, high-impact event deserves attention, especially where people are already living on the edge.”
For many African communities, this probabilistic mindset is less foreign than it might appear.
“My grandparents never looked at one sign,” said Purity Mutua, a community elder and smallholder farmer from Kitui County, Kenya. “They watched birds, trees, insects and winds. Even if most signs pointed to early rains, they still prepared for delay.”
Today, science is formalizing that intuition. In northern Kenya, ICPAC is working with pastoralist communities to blend ensemble data with traditional ecological indicators, producing hybrid advisories that people trust and act upon.
When probability is communicated clearly, it transforms decision-making.
A farmer in Zambia facing an 80 percent chance of delayed rains but a 15 percent chance of normal onset can diversify planting strategies rather than gamble an entire harvest.
“This is the goal,” said Dr Olga Wilhelmi, a geographer at the US National Center for Atmospheric Research (NCAR). “Probabilistic information shifts people from passive recipients to active risk managers. It builds agency.”
Yet institutional barriers persist. Many national meteorological services, under-resourced and sometimes under political pressure, default to deterministic language that projects confidence rather than clarity.
Journalists, Lore argues, must become more demanding consumers of climate data.
“The next time you receive a seasonal forecast,” he urged, “ask: What does the ensemble show? What are the second- and third-most likely scenarios? What are the high-impact risks at the margins?”
As climate disruption accelerates, Africa can no longer afford the old language of certainty. The emerging language of probability, precise, transparent, and contextual, is not about doubt. In the continent’s most vulnerable landscapes, it is becoming the foundation of resilience.
Collison Lore has asserted that at Climate Frontline they are committed to training journalists across the Global South in climate science, data literacy and ethical reporting.
Deterministic forecasting provides a single, specific prediction of the future. It is the type of forecast most people are familiar with: “It will rain tomorrow” or “The temperature will reach 30°C.” While simple to understand, it carries a profound limitation because it ignores uncertainty. When reality diverges from prediction, farmers, planners, and emergency responders may be left exposed.
Probabilistic forecasting, by contrast, embraces uncertainty. Instead of offering one outcome, it presents a range of scenarios, each with an associated likelihood, a 70 percent chance of below-average rainfall, for example, alongside a 30 percent chance of normal or above-average conditions.
Built on ensembles of simulations that capture the atmosphere’s inherent variability, probabilistic forecasts do not attempt to identify a single correct future. They provide a spectrum of possibilities, allowing governments, aid agencies, and communities to assess risk and prepare accordingly.
In essence, deterministic forecasts tell you what will happen; probabilistic forecasts tell you what might happen and how likely each outcome is.
In a climate-challenged world where extremes are becoming the norm rather than the exception, that distinction may spell the difference between preparedness and catastrophe, between fragile hope and durable resilience.
Get the latest news and insights that are shaping the world. Subscribe to Impact Newswire to stay informed and be part of the global conversation.
Annet Mbabazi (Rwanda)
Most Impactful Work: Her work as a radio journalist, winning awards like the OFAB Africa Journalist of the Year.
Short Bio: Annet Mbabazi is a Rwandan radio journalist known for her impactful reporting on agricultural innovations. Her work, which earned her the 2024 OFAB Africa Journalist of the Year award, often focuses on stories that boost output and raise the living standards of local farmers.
Chinua Achebe
Biggest Selling Title: Things Fall Apart
Short Bio: A Nigerian novelist, poet, and critic, Achebe is best known for his novel Things Fall Apart, which is considered one of the most important works in African literature. His writing addresses the complexities of post-colonial identity.
Shadow Gal
A buddy in Cape Town saw some dude cat calling a coloured woman with that kst kst kst sound, so she turns around and says \”What? You got a puncture in your face or something??\”
Landfill Max
Sweden: You have two cows. You offer them to the government, who gives everybody milk.