How scientists predict famine before it hits

For several weeks back in 2018, Yadira Martínez González suddenly had to feed 15 additional mouths. Her husband's relatives, who had emigrated from Colombia to Venezuela decades ago, returned as part of an exodus of millions leaving a crumbling country.

An economic crisis, attributed to alleged corruption and poor financial management, which the country's president and high ranking officials have denied, had caused inflation to spike by over 860%, according to official data. Martínez González’s work selling her crafts alongside one of the few roads traversing the La Guajira desert wasn't enough to buy food for the nearly 40 people who now made up her family. "We didn't eat that much," she says. "Maybe twice, most likely once every day."

She wasn’t the only one with unexpected guests from Venezuela. Without enough food, animals began to disappear in Martínez González’s rural settlement, Palenstu in La Guajira, one of the more than 2,500 villages – known as rancherías – of the Wayúus, Colombia's most numerous indigenous group. As tensions rose on the dusty ground, satellites and local weather stations showed that rains would not arrive. Without help, crops would fail. A famine seemed imminent.

But it never arrived for the 45 families in Martínez González’s community, nor for 7,000 other people in four municipalities of La Guajira.

You might also want to read:

How Hong Kong protects people from dangerous landslides

How Taipei discovered an active volcano on its doorstep

The animals that detect disasters

In June 2018, a group from the Food and Agriculture Organization (FAO), the World Food Programme and Action Against Hunger visited the communities. They found that the Venezuelan exodus had expanded the size of Wayúu families in the border by an average of four people, while continued drought had nearly halved crop, milk and meat production. With more people competing for jobs, wages were decreasing, and families were owing money just to buy a little bit of seed.

Coupled with the forecast of below-average rainfall and high temperatures, the organisations knew they needed to act before the crisis reached a crest. They set up a rapid action programme, and by September 2018, they’d deployed a four-step strategy to help communities. Within eight months, they had rehabilitated 18 water wells, distributed seeds, guidance and tools to establish community farms, and sent animal-health brigades to vaccinate and treat about 12,500 cattle, sheep and goats.

It allowed us not only to have food but to teach children and young members to value our lands – Yadira Martínez González

The $400,000 (£310,000) programme transformed the communities. Over nine months, the 17 community farms set up by the programme increased the availability of fresh vegetables for families involved from two to four days a week, according to the FAO. In total, the 1,000 households in five municipalities harvested about 115kg (254lb) of food from five different crops – including some new ones, like aubergine, the FAO said. In comparison, families outside the program harvested 35kg (77lb) from far fewer crops.

More importantly, Martínez González says, they recovered their elders' collective mindset towards food. "It allowed us not only to have food but to teach children and young members to value our lands," she says.

The intervention, which tried to tackle the crisis well before it struck, is a novel humanitarian response to severe hunger crises. "The issue is not to provide emergency assistance to hand out food, but to create conditions to adapt to drought," says Alan Bojanic, FAO’s representative in Colombia.

Instead of focusing on food donations and cash assistance, these interventions simultaneously build up people's resilience to immediate and future shocks, says sociologist Erin Lentz, who investigates food security and food aid at the University of Texas, Austin.

Early interventions like the one in Martínez González’s settlement are only possible because 40 years ago, after a devastating famine killed about one million people in Ethiopia in the mid 1980s, a group of earth scientists, climatologists, agronomists and data analysts combined forces to ensure that a famine would never again take the world by surprise. They became famine forecasters – and they've been wildly successful in their goal.

[T]hese forecasters can predict the rise of a famine months before it hits. Consequently, they allow local governments to act and international funds to hit the ground before starvation does

"Early warning systems have, by and large, predicted most of the [hunger] crises that we've had recently," says Dan Maxwell, a professor at the Friedman School of Nutrition Science and Policy at Tufts University in the US.

Using real-time analyses of rain and wind patterns, soil moisture levels, river flows, international grain prices, weather variations and satellite data on changes in vegetal cover, combined with information on conflict intensity and humanitarian crisis, these forecasters can predict the rise of a famine months before it hits. Consequently, they allow local governments to act and international funds to hit the ground before starvation does, preventing the worst possible outcome in already food-insecure regions.

In 2016, the Famine Early Warning Systems Network (Fews Net), one of the largest hunger early warning systems currently in place, predicted an unprecedented extreme drought in the Horn of Africa that would later push about 27 million people into severe food insecurity. The alert brought immediate food supplies for two million people in Somalia. Compared to a drought in 2011, early intervention helped to reduce the number of severely hungry Kenyans by over a million, one report found.

"There's a lot of things to be depressed about in the world," says Chris Funk, director of the Climate Hazards Center (CHC) at UC Santa Barbara, which collects most of Fews Net meteorological data. "But I do think it's important to realise that this is [humans] being our best selves."

Funk's hero is a 20th-century statistician named Gilbert Walker. An English mathematician obsessed with finding and measuring patterns and cycles, Walker was recruited to lead the Indian Meteorological Department in 1903. It placed an enormous task on his shoulders: to forecast the arrival of monsoon rains – which, just 26 years earlier, had made crops fail, unleashing the worst famine ever recorded in that country.

Walker set out to understand how rains, temperatures and pressure relate to each other and shape global weather patterns. His and others’ work led to the understanding of El Niño and La Niña, periodic changes in the temperature of the Pacific Ocean that can lead to droughts or excessive rains in certain parts of the world. "We're kind of still in his footsteps. We're doing what he set out to try to do," Funk says.

In 1985, as a then-kid Funk watched the Live Aid concert for Ethiopia on his TV, the US Agency for International Development (USAID) was creating Fews Net in a bid to help governments and relief agencies plan their interventions in humanitarian crises.

The task was not insignificant. Seeing the complete picture of drought is like focusing simultaneously on something massive and far away and something tiny nearly grazing your nose. The big, far away aspects are the global variations that affect rainfall: changes in the temperature of winds and the sea surface or fluctuations in atmospheric pressure. Satellites can easily spot those.

But satellites have a hard time seeing things like local soil moisture, vegetal cover, river and stream flows and air temperatures. When these shifts are visible to satellites in space, it's often too late: food aid is already overdue.

Early warning systems are now good at warning about an impending shock, but decision-makers are reluctant to spend dollars until they see the crisis unfolding

Fews Net worked using satellite information and local weather stations' spotty data for years. Then, in 2002, it started funding a small climate research group at UC Santa Barbara called the Climate Hazard Group (now Center).

Co-led by Funk, the team’s first report in 2002 combined scarce historical information on rainfall data in Ethiopia and data on crop conditions in Southern Africa to document the 2002 El Niño’s impacts in the region. As a result, USAID sent some $280 million (£220 million) in food aid.

"More or less, we've been doing a lot of the same thing since," Funk says. "But now we've gotten much better at doing it."

To get better data, the team developed a two-way strategy. First, they managed to get data from a series of European satellites collecting weather information about Europe and Africa since the 1980s. These satellites measure the temperature of clouds and use this as a proxy for rainfall: if the temperature of a particular cloud clump drops below −38C (-36F), it is likely raining on the ground directly under it. With this data, the team mapped the history of rains in Africa.

Simultaneously, they convinced local meteorological agencies to share their weather station data. It's been a slow but rewarding process, says Diego Pedreros, a field scientist for the US Geological Survey and a long-time collaborator of Funk. Throughout the years, they've established partnerships with local researchers, field scientists and government agencies all over the US and in countries like Guatemala, Kenya, Botswana, Niger and Ethiopia. They now receive the information of 2,000 meteorological stations in 17 countries, updated every couple of days.

With the ability to now see both the monstrously big and the tiny elements of upcoming droughts, the Climate Hazard Center sends a report every five days to its partners at the US government, humanitarian agencies and international agencies like FAO. The report is a series of maps of hotspots in Africa, Asia and Central America, with real-time rainfall information and projections for the next 15, 30 and 60 days. It also compares current levels to previous years, contextualising the data – is it below or over the average? How likely is it that things will change? How bad is it compared to previous years?

Based on the maps, officers from US government agencies and food security analysts introduce other variables like grain prices, recent economic shocks in local economies and the exacerbation of armed conflicts to create a definitive famine forecast for more than 35 countries. Its work is integral in guiding how to spend USAID's nearly $4bn (£3.2bn) annual food aid budget.

What I find really frustrating is that this happens so frequently, and we're mostly responding in a reactive way – Chris Funk

Early warning about famines has give people a chance to react to and even recover from crises, says Lentz. “I do think this speed really makes a huge difference in keeping people from using worse and worse coping strategies, [ending up in situations] that then they can't get out.”

Still, the information produced by Fews Net doesn't always mean decision-makers will act on time. Early warning systems are now good at warning about an impending shock, but decision-makers are reluctant to spend dollars until they see the crisis unfolding, says Maxwell. "The problem is, by the time you see what's actually happening, it's far too late."

Fews Net and similar tools have warned about troubling signs in Southern Madagascar since May 2021, predicting the continuation, for the third year in a row, of the worst drought in 40 years. By September 2021, Funk was writing articles trying to bring attention. Now, nearly 1.5 million people are living on the edge of starvation, feeding on insects and cactus leaves.

"What I find really frustrating is that this happens so frequently, and we're mostly responding in a reactive way," Funk says.

To fill this gap between forecasts and response, some experts propose, requires looking at famines not as an immediate shock but as the fatal outcome of a larger, quieter emergency often looming over these communities: chronic hunger.

Response to famine and hunger has been traditionally treated as two separate, almost unrelated issues, says Lentz. Famine responses often focus on bringing momentary attention, while hunger-reduction programs tend to be related to development projects.

Yet, in the real world, hunger and famines are intimately related. When a crisis hits a chronically hungry population, "they already start closer to the edge of acute food insecurity", Lentz says.

Martínez González experienced this first-hand. When she suddenly found herself needing to feed 15 extra mouths, her community's ability to feed themselves had already been deteriorating for nearly a decade. She remembers her family routinely harvesting vegetables when she was a child. Women would cross to Venezuela to buy cheap products there. But starting in 2013, La Guajira went through one of its worst droughts on record. Water wells were damaged or dry. Herders saw most of their animals die. Beans, watermelons, pumpkins, yams, corn and melons grew sparsely, if at all.

When communities get over the shorter term crisis, they're often left more vulnerable to new shocks than before

Venezuela banned any exports of essential foods in 2014, and border patrols began stripping Wayúu women of the few pounds of rice, oil or flour they were bringing home for their families. Then when inflation made food a luxury in Venezuela, thousands of migrants began walking into Colombia and other Latin American nations. By November 2018, about 1.2 million people had settled in Colombia, many of them in La Guajira. The latest estimate shows that nearly 6 million Venezuelans have moved, desperately escaping poverty, violence and hunger. About 1.8 million have stayed in Colombia, more than 106,000 settled in La Guajira, according to Colombian government data.

Yet most humanitarian and state interventions in the peninsula are short-term, and focus on helping people get through the immediate hurdle, says anthropologist Claudia Puerta Silva, who's been working in the region for the last 20 years. When communities get over the shorter term crisis, though, they're often left more vulnerable to new shocks than before, she says. If their resilience is not increased, they can become trapped in a vicious cycle, she says.

CARBON COUNT

The emissions from travel it took to report this story were 0kg CO2. The digital emissions from this story are an estimated 1.2g to 3.6g CO2 per page view. Find out more about how we calculated this figure here.

This case of an ongoing crisis coupled with a short-term response is not an exception. A report by the Humanitarian Policy Group found that in 2016, most humanitarian food aid help was concentrated in the same 10 countries, and 86% of that money went to communities where crises had lasted more than three years. "Crises have endured for decades," the report reads, "but humanitarian aid has remained annualised and short term."

A better response, Lentz says, is early interventions that help communities build capacities for the upcoming emergency and the long term. That's precisely what FAO's intervention did in La Guajira, which experiences a continued situation of chronic hunger that can easily spill into famine if circumstances deteriorate.

"Most of the projects that come to La Guajira are emergency projects," says Bojanic, from FAO Colombia. "They’re six-month or one-year projects to address a critical situation. But we know that this is not the answer for such a complicated area."

With the drought forecast in hand, FAO and the community prioritised planting short-cycle, drought-tolerant seeds like maize, cassava and various beans. The project also supported their economic activities, creating an online shop for communities to sell their handicrafts, which came in handy to support them during the Covid-19 lockdowns two years later.

It gives Funk hope that people are using hunger forecasts beyond specific crises. Government agencies in Kenya and Malawi are now using observed conditions to design insurance policies for small farmers expecting variations of drought to help them get loans to increase production when needed. PlantVillage, a partnership of 20 non-profits and local governments, is helping farmers plan for the season ahead by informing them via text or WhatsApp message what kind of seeds might work best in the near future, and whether rains are coming. It has reached about 350,000 farmers in Kenya so far.

These kinds of early, planned interventions can also save millions of dollars. For example, USAID has estimated that about $2.6bn (£2bn) in humanitarian aid could be saved in Kenya, Zambia and Ethiopia over the next 15 years by combining early interventions with small cash transfers for families to have a safeguard in case of emergencies.

In a world where the number of severely food insecure people has doubled in just two years due to conflict, drought and natural disasters, and where climate change is already intensifying drought in many regions, there is more need than ever for early planning to prevent famine.

"Drought is looking for opportunities to create more havoc," Funk says. "If we can build up the resilience of people [with forecasting tools], when these events occur they'll be less catastrophic."