The race to fix Africa's poor weather forecasting

African nations are particularly vulnerable to severe weather events. These not only cripple economies and people's livelihoods but also cost lives across the continent.

In West Africa, for instance, more than 70% of the population is affected at least once every two years by flood, drought or sandstorms.

But like other less-developed regions, many parts of Africa have gaps in the ability to warn people of both imminent natural disasters, such as flash flooding, or future hazards, such as drought.

This has dramatic consequences. Over the last two decades, the average number of deaths caused by a flooding event in Africa is four times higher than in Europe or North America, according to a 2024 article in Nature, a science journal.

The continent, where the vast majority of people are dependent on rain-fed agriculture, also suffers the most deaths from droughts. From 2006 to 2015, 99% of deaths from all droughts worldwide were in Africa, even though it only experienced half of globally reported droughts in that time.

Push to improve Africa's forecasting

With climate change predicted to increase the intensity and frequency of such extreme events, said Victor Ongoma, an assistant professor of climate change adaptation at Morocco's Mohammed VI Polytechnic University, "there's pressure on weather providers in Africa to provide more reliable, accurate and timely weather forecasts."

They've made great strides in terms of quality and timing, he told DW. "It's still below average, but there is still lots of effort going on and I'm hopeful things will get better with time."

Weather forecasting is complex. In the best case scenario, it needs millions of data points — measurements of things like temperature, rainfall, humidity, solar radiation and wind intensity and direction.

These observations are collected on land, lakes and oceans, in the atmosphere and by satellites, and then crunched by powerful computers using complex mathematical models to predict weather patterns.

Data observation still poor in Africa

But weather forecasting still has a way to go on the continent, which has the world's least developed land-based weather observation network. And, according to a 2019 report by the World Meteorological Organization (WMO), this aging network is deteriorating due to lack of maintenance.

Angola, for example, went from more than 150 working weather meteorological stations before independence from Portugal in 1975 to 20 in 2022.

In West Africa, more than 60% of weather data is still collected manually by non-professional and voluntary staff, leading to poor quality data that can't be used for real-time monitoring.

According to a WMO database, Africa only has 37 radar facilities, which are vital for tracking weather fluctuations and rainfall and for forewarning of floods and other hazards. In comparison, 345 radars cover Europe and Russia, which together have a smaller land mass than Africa.

Moreover, more than half of Africa's radar stations are unable to produce accurate enough data to predict weather patterns for the coming days or even hours, the 2024 Nature article noted. 

Weather models weak for Africa

On top of this, the global models used to predict the weather perform poorly in Africa.

"The main disadvantage is that most of these models were developed for the global North," said climate scientist Benjamin Lamptey, explaining that these models work well in the mid-latitudes (about 30 to 60 degrees north or south of the equator), where much of Europe and the USA fall.

But African weather patterns, such as monsoons, haven't been incorporated into the models, said Lamptey, a visiting professor of meteorology at the UK's University of Leeds, which is why "we're seeing shortcomings in the forecasting."

"Just developing things in the global North and transferring them to the global South does not work. We need to have ownership and things must be viewed from the African perspective," he said.

Many African nations lack computing facilities powerful enough to run complex weather models, while power cuts and weak internet connections limit the access to global data sets.

Increasing efforts and funding

There is some good news. For climate expert Jeffrey N.A. Aryee from Ghana's Kwame Nkrumah University of Science and Technology, weather forecasting in Ghana, and across many other sub-Saharan nations, is quickly evolving.

Ghanaians used to complain that forecasts were "way off" what the actual weather was, Aryee tells DW. But the Ghana Meteorological Agency has "undergone a revolution," he said, where it is now able to provide a realistic short-term forecasts because of better access to satellite and radar data (although Ghana's radar recently broke down) and prediction models.

The numbers of surface weather stations across Africa sending data to the global observing network known as GBON nearly doubled in the first half of 2023, going from 589 to 1,045.

And thanks to the WMO and other organizations, African nations are making inroads into digitizing the millions of data points hidden in their historical weather archives.

"This will help make more data available because for you to know where you're going [with weather predictions], you need to know where you are coming from," said Victor Ongoma.

Both Ongoma and Lamptey also praise the development of regional climate centers.

"For example, if you take countries along the Gulf of Guinea in West Africa, they have two rainy seasons, and the rainfall originates in the east and moves towards the west," said Lamptey.

"So, weather systems are unique to the regions," he said, meaning the five climate centers can tailor their expertise to boost regional forecasting.

Amid the flurry of efforts to close Africa's weather forecasting gaps, it's important not to forget the end users, warn experts, and how they access information and in what languages.

Fishermen going out on the ocean want different weather information to pastoralists looking for water for their herd. And farmers trying to decide when to plant or harvest need a reliable seasonal forecast to help with decision making.

"You need to know what people really want to know, so you don't just tell them, 'OK, it's going to rain,'" said Jeffrey N.A. Ayree. "A fisher is not just interested in whether it's raining or not. He's also interested in the nature of the sea."

Edited by: Keith Walker 

While you're here: Every weekday, we host AfricaLink, a podcast packed with news, politics, culture and more. You can listen to AfricaLink wherever you get your podcasts.