For three decades, Yusufu Che Wadi farmed the same two acres of land near Somola Village in Machinga district, eastern Malawi. In a typical season, the land produced around 1,000 kilograms of food — enough, he says, to feed his eight-member family for the year.
Then came Cyclone Freddy.
When torrential rains swept through southern Malawi in March 2023, water tore through farms, homes and infrastructure. The World Meteorological Organization would later confirm Freddy as the longest-lasting tropical cyclone on record, spending 36 days at tropical-storm strength or above during its extraordinary journey across the Indian Ocean and southern Africa.
For Che Wadi, the record was intensely personal.
When the floodwaters finally receded, his field was buried beneath sand and rocks. Crops that had been approaching maturity were gone. So were animals his household had depended upon.
“The destruction was just unbearable, and to make matters worse, my seven goats and twenty local variety chickens were either swept away or found dead,” he tells me pointing where his goats kraal once stood.
Che Wadi’s experience was repeated across communities affected by the cyclone. WMO reported that extreme rainfall during Freddy’s final landfall inundated extensive agricultural areas in Malawi and Mozambique and inflicted severe damage on crops.
For Malawi, those losses strike at the foundation of everyday life. Agriculture remains central to the country’s economy and livelihoods, with the World Bank estimating that the sector employs more than three-quarters of the population. Much of that farming is small-scale and dependent on rainfall, leaving households particularly exposed when the timing or intensity of the rainy season changes.
In Machinga, however, the destruction has also accelerated an experiment in adaptation. Farmers are combining some very old ideas — shaping land to slow water, diversifying what they grow and watching the weather closely — with increasingly sophisticated networks of weather stations, digital forecasts, mobile phones and early-warning systems.
The technology ranges from automated observation equipment to something much simpler: a plastic rain gauge in a farmer’s field.
After the Flood, Farmers Began Reshaping the Land
In the aftermath of Freddy, Che Wadi and other farmers in the area began working with agricultural and climate specialists on ways to reduce their vulnerability to the next extreme rainfall event.
One intervention is visible directly in the landscape.
Farmers have been digging swales — shallow channels following the contours of sloping land — to interrupt fast-moving runoff. Instead of allowing rain to rush downhill, carrying valuable topsoil with it, the channels slow the water and give more of it time to infiltrate the soil.
That makes the same intervention useful at opposite ends of Malawi’s rainfall problem.
During heavy rain, the swales can reduce erosion and slow runoff. During a dry spell, moisture retained in the landscape may remain available to crops for longer.
For farmers around Somola Village, where cultivated land lies beneath surrounding hills, the memory of Freddy has made that protection tangible.
“Previously, our farming land which most of it is on the slopes, did not have water controlling and retention mechanisms which made it vulnerable to large scale soil erosion and as we saw with Cyclone Freddy, there was an avalanche too which cost us almost matured crops and livestock. Now everyone is working to control that,” he says.
Swales cannot make a farm flood-proof. Extreme rainfall can overwhelm even well-designed land-management measures. But the principle is increasingly important in climate adaptation: rather than trying to remove water from a landscape as quickly as possible, slow it, spread it and help the soil absorb what it can.
The Hardest Decision Can Be When to Plant
Not every climate shock arrives dramatically.
For smallholder farmers dependent on rainfall, an apparently promising start to the wet season can create its own risk.
Seeds planted after the first rains may germinate successfully, only for the soil to dry again if the rain disappears. When seed is expensive and household resources are limited, replanting is not a minor inconvenience. It can mean buying cheaper or uncertified replacements, planting less land or facing a poor harvest months later.
Knowing when enough rain has arrived can therefore matter almost as much as knowing how much rain will fall across the whole season.
Malawi has invested heavily in improving that information through the project known as M-CLIMES — Scaling Up the Use of Modernized Climate Information and Early Warning Systems.
The system extends far beyond a conventional national weather forecast. Automatic weather stations measure atmospheric conditions. Hydrological stations monitor water levels and flows. Other observation equipment gathers information that can be combined into forecasts and warnings designed for particular hazards and locations.
But collecting better data is only useful to farmers if the information reaches them in a form they can actually use.
Turning a Forecast Into a Farming Decision
That last step is where Participatory Integrated Climate Services for Agriculture, or PICSA, becomes important.
Rather than handing farmers a weather bulletin and leaving them to interpret it, PICSA brings climate information into agricultural decision-making. Farmers and extension workers can compare seasonal forecasts with historical rainfall information and consider different crops, varieties and planting dates.
UNDP describes the approach as putting farmers at the centre of climate services, helping them use seasonal forecasts alongside their own circumstances to make practical decisions.
In Machinga, those choices can be as specific as whether to plant a local maize variety that requires around 120 days to mature or a drought-tolerant hybrid that can mature in approximately 90 days.
Climate information does not make either option safe. It makes the trade-off visible.
The same observation network can also support warnings unrelated to crops. Kondwani Francis Kanjala of Malawi’s Department of Climate Change and Meteorological Services says lightning detection is among the system’s functions.
“This technology also has an advanced lightning detection system to track thunderstorms which is leading killer of people in rural Malawi during rainy season thereby saving lives and costs which could have been incurred with every injury or fatality,” says Kondwani Francis Kanjala from the DCCMS
The principle is straightforward: measurements become forecasts, forecasts become warnings, and useful warnings give people time to act.
Sometimes Climate Technology Is Just a Rain Gauge
The phrase modern climate information system can suggest satellites, servers and rooms filled with screens.
For Marriam Binali, it can mean looking down at a rain gauge.
The 35-year-old mother of three is among farmers who have been given simple gauges to measure rainfall where they actually farm. Instead of relying only on whether the ground looks wet or whether rain has traditionally arrived by a particular date, farmers can record the amount that has fallen and compare it with local climate information.
For Binali, that change represents a break with indicators farmers once relied upon to anticipate the seasons.
“Gone are the days when we used to have a mango calendar,” Binali says while referring to old traditional way of relying on the falling of mango tree flowers as time for rainfall. “that no longer works, these days the rains come later, stops earlier and fall harder.”
The observation captures something profound about climate adaptation.
Knowledge passed between generations remains valuable, but traditional indicators developed under one climate regime may become less reliable as seasonal patterns shift. The answer is not necessarily to discard local knowledge in favor of distant technology. In projects such as PICSA, the two can meet: farmers’ understanding of their land is combined with measurements, historical records and forecasts.
In that sense, meteorology becomes something intensely practical — not only the work of scientists analyzing instruments, but a farmer reading millimetres of rain before deciding whether a bag of seed should go into the ground.
The Forecast Has to Reach the Last Mile
Malawi’s climate-information challenge is not solved when a forecast has been produced.
Someone still has to receive it.
For farmers without constant internet access or sophisticated devices, some of the most useful communication technologies are neither new nor glamorous.
One of them is radio.
Che Wadi says community stations have changed the usefulness of the forecasts available to him. National broadcasts can describe conditions across a large region; a local station can provide information for the place where someone is deciding whether to plant tomorrow morning.
“Now when I tune in to Radio Liranguka, the weather forecast I get is no longer for the whole Eastern Region, but specifically for Machinga Boma and advising us whether to plant or wait. To me, this is priceless because it saves me from the prospect losing seed, money and most importantly the threat of food insecurity,” he says.
That combination of sophisticated forecasting and comparatively simple delivery is central to effective early-warning systems. UNDP says Malawi’s climate-information programmes have used combinations of mobile communications, print and radio to get localized agricultural advisories to communities.
A forecast that never leaves a meteorological office cannot protect a crop.
A Basic Mobile Phone Can Be a Climate-Adaptation Tool
For Sakina Kalichero, the forecast arrives with a beep.
Kalichero farms two acres but does not have a rain gauge of her own. Instead, her mobile phone receives messages from DCCMS twice a week, on Tuesdays and Fridays.
The information arrives by SMS in a local language and includes rainfall forecasts for her area.
That seemingly modest service changed a decision during the 2024/25 growing season.
“We are able to get rainfall forecast via Short Message Service (SMS) on our phones. For instance, in the 2024/25 farming season, using the information I got on my phone, I delayed planting by a week while others did plant but their crops withered in dry spells and had to be replanted with an extra cost,” she says.
Her story illustrates why the most useful climate technology is not necessarily the device with the most impressive specification.
An automated weather station may collect the observation. Computers may process the data. Meteorologists may build the forecast. But for a farmer standing in a field, the final technology can be an SMS telling her to wait another week.
Early Warnings Cannot Remove the Risk
There is a danger in telling climate-adaptation stories as if better information solves the underlying problem.
It does not.
A rainfall forecast cannot rebuild a field buried in stones after a cyclone. A rain gauge cannot replace livestock swept away in a flood. An SMS cannot make expensive seed affordable, and a swale cannot contain every extreme downpour.
Farmers can make better decisions when they know more about the weather ahead, but they remain constrained by the resources available to act on that information.
This is particularly significant in Malawi, where agriculture remains highly exposed to climatic shocks. The country’s experience demonstrates both the value and the limits of adaptation: better information can reduce avoidable losses, but resilience also depends on healthy soils, accessible seeds, infrastructure, functioning extension services, water management and households having enough financial margin to recover when precautions are not enough.
Early-warning systems are therefore not an alternative to broader climate action. They are one layer of protection for people already living with its consequences.
Learning to Measure a Changing Climate
There is an appealing tendency to describe climate technology through satellites, artificial intelligence and enormous datasets.
Those systems matter. Malawi’s increasingly sophisticated observation network would not function without instruments collecting and integrating data across the country.
But in Machinga, its success can ultimately be measured through much smaller decisions.
A farmer digs a channel across a slope so the next storm loses some of its force.
Another bends down to read the rainfall collected in a gauge.
A radio announcer gives tomorrow’s forecast for Machinga rather than an entire region.
A phone beeps on Tuesday, and a farmer decides not to plant yet.
None of those actions can restore the predictable seasons that farmers remember. Nor can they guarantee a harvest when extreme weather arrives.
They can, however, turn information into choices.
For households whose food security may depend on whether expensive seed goes into wet soil on the right day, that can be the difference between simply watching the climate change and having some ability to respond to it.