
In July 2020, Lillian Kay Petersen, a 17-year-old high school senior from Los Alamos, New Mexico, won first place and $250,000 in the Regeneron Science Talent Search, the oldest major science competition for American high school students. Her project was a tool that uses daily satellite images to forecast crop harvests early in the growing season. The aim: give aid agencies an earlier warning of food shortages, so they can act before a drought turns into a hunger crisis.
Headlines at the time said she had invented a tool that could stop famines. That overstates it a little. What she built is an early-warning method, and early warning is one important piece of famine prevention. Here is what the project did, why it matters, and what the competition is.
What Petersen's harvest-prediction tool does
Healthy, growing crops reflect light differently from stressed or dying plants. Satellites that photograph the Earth every day can measure this difference across whole regions using standard vegetation-health indices, which compare how much red and near-infrared light the plants reflect.
Petersen's method analyses those daily satellite measurements through the growing season and links them to harvest results. According to the Society for Science, which runs the competition, she first tested the approach against known crop data from the United States, then applied it to countries in Africa, where it predicted harvests with high accuracy compared with the yields that were later reported.
The big advantage is timing. Official harvest figures often arrive after the season ends, which can be too late. A forecast made early in the season, while crops are still in the ground, gives governments and aid groups weeks or months to arrange food, money or seed before shortages become severe.
Why she chose this problem
Petersen's interest was personal. Her family had adopted three younger siblings who had experienced food insecurity before joining the family. She told NPR: “I have watched my younger siblings struggle with the lifelong effects of malnutrition.” Reading about the severe 2015 drought in Ethiopia, which left millions of people needing food aid, pushed her to look for ways to help aid organisations track crop conditions.
She started monitoring African droughts with satellite imagery during her sophomore year of high school. Less than two years later she had published her first paper on the subject in a peer-reviewed journal, an unusual achievement for a teenager. Around the time of the award, she also met researchers from the International Food Policy Research Institute to discuss how the COVID-19 pandemic was affecting crop production in Africa.

Petersen at the 2020 awards, which were held online. Because of the pandemic, the usual black-tie gala in Washington, D.C., was replaced by a virtual ceremony, the first in the competition's history.
About the Regeneron Science Talent Search
The Science Talent Search began in 1942 and is run by the Society for Science (then called Society for Science & the Public). It was sponsored first by Westinghouse and later by Intel; Regeneron, a biotechnology company, took over in 2017 with a ten-year, $100 million commitment that nearly doubled the annual prize pool to $3.1 million.
Each year around 1,800 to 2,000 high school seniors enter, 300 are named scholars, and 40 finalists are judged by scientists. In 2020 the ten top finalists shared more than $1.8 million. Alumni of the competition have gone on to win 13 Nobel Prizes, two Fields Medals and dozens of other major science honours, according to the Society.
The 2020 top three were:
- Lillian Kay Petersen, 17, Los Alamos, New Mexico: $250,000 for her harvest-prediction tool.
- Jagdeep Bhatia, 18, Green Brook, New Jersey: $175,000 for fast machine-learning algorithms that learn concepts with guidance from an instructor, work that could help robots learn more like people.
- Brendan Crotty, 18, Muskogee, Oklahoma: $150,000 for a hybrid gas burner design that reduces nitrogen dioxide emissions from industrial burners.
Other top-ten projects ranged from pure mathematics to cancer immunology, bacteria-powered soil sensors and less biased AI for spotting toxic online content. Maya Ajmera, president and CEO of the Society, said in the announcement: “Students like Lillian Petersen are the stewards of our future.”
Can satellite forecasts really prevent famine?
Satellite crop monitoring is already a core part of global food security work. The Famine Early Warning Systems Network (FEWS NET), set up by the US government after the 1984–85 Ethiopian famine, and the UN Food and Agriculture Organization both combine satellite data with ground reports, market prices and rainfall records to warn of coming crises.
Better and earlier harvest forecasts make those warnings stronger. But famines are rarely caused by bad harvests alone. Conflict, poverty, high food prices, blocked roads and slow political decisions often decide whether a drought becomes a disaster. The 2011 Somalia famine, for example, happened even though warnings had been issued many months in advance. Prediction helps only when it is followed by funding and action.
That is why the most accurate way to describe Petersen's project is as a promising, low-cost improvement to early warning, one tool among many, rather than a cure for hunger.
The wider hunger picture
The world grows enough food, in total calories, to feed everyone. Hunger persists mainly because of poverty, conflict, waste and unequal access. The UN's 2020 report on the state of food security estimated that close to 690 million people were undernourished in 2019, with the largest number living in Asia and the fastest growth in Africa. The report also warned that the pandemic could push many millions more into hunger. Older articles often quoted a figure of almost a billion hungry people; that was the estimate from around 2009 and was later revised down with better data.
Within that picture, timely information matters. When aid agencies know early which regions face poor harvests, they can buy food when prices are lower, plan deliveries and support farmers before families have to sell livestock or skip meals.
What young scientists can take from her story
- Start with a real problem. Petersen's project grew out of something she cared about personally, which kept her working on it for years.
- Use open data. Much satellite imagery is free to use, which means students can tackle global questions with a laptop and persistence.
- Test against reality. She checked her forecasts against reported harvests, the step that turns an idea into evidence.
- Publish and share. Peer-reviewed publication and talking with organisations like IFPRI connected her work to the people who could use it.
After the award, Petersen planned to begin studying at Harvard that autumn. Watch the announcement and interviews below.
For another young inventor's story, read about Ann Makosinski's body-heat flashlight.
Sources: Society for Science, 29 July 2020: Regeneron Science Talent Search 2020 winners; GPB/NPR, 30 July 2020: High school senior's tool to end food insecurity wins national competition; Society for Science; FAO: The State of Food Security and Nutrition in the World; FEWS NET.

COMMENTS