SANTA FE – At 17, recent Los Alamos High School graduate Lillian Kay Petersen already is making a difference in the world.
Her science project – which is really so much more than that – “Real-Time Prediction of Crop Yields from MODIS Relative Vegetation Health: A Continent-Wide Analysis of Africa,” has drawn the attention of the International Food Policy Research Institute, the United States Agency for International Development and the U.S. Department of Agriculture.
Lillian Kay Petersen (Courtesy of Lillian Kay Petersen)
And, oh, yes, it recently landed her the top prize in the 2020 Regeneron Science Talent Search, a kind of international science fair on steroids, earning her a $250,000 prize.
That’s not a bad nest egg for the Harvard-bound teenager who will study applied math and molecular biology.
“I’m not actually sure what I’m going to do with the money,” Petersen said. “I haven’t had a chance to think about it. … I was listening to them calling the awards. When I heard my name called after first place, it took me a couple of seconds to realize what happened. I was just blown away. It was so unbelievable.”
What may be even more unbelievable is the amount of work that went into this project.
A mentee at the Santa Fe-based Descartes Labs, which uses the Descartes Labs Satellite Platform in its research, Petersen had access to daily satellite imagery from across the world.
Knowing that a severe drought in 2015-16 had wrecked food production in Ethiopia, she was inspired to research that area because she has three younger, adopted siblings, all of whom suffered from food insecurity before their adoption.
“I have seen the life-long effects of malnutrition,” Petersen said. “And the physical and mental delays they had to overcome. When I heard about the major drought in Africa in 2015 and 2016, with 18 million people facing starvation, I thought there had to be a better way to respond to food crises.”
The model Petersen created is able to accurately predict crop yield three to four months before the harvest based on early satellite imagery.
She computed the monthly Normalized Difference Vegetation Index, a measure of crop health, finding correlations between crop yields and the index anomalies by judging the health of the fledgling plants.
Having early, accurate crop yield predictions, Petersen said, allows countries to better prepare for potentially catastrophic conditions ahead of time.
As a matter of fact, her model is taking on added importance in this COVID-19 era as crops are suffering.
“I met with the International Food Policy Research Institute last week,” Petersen said. “We discussed the effect of COVID-19 on crop health. They were interested in my model because they were getting some worrisome reports in Kenya. And Rwanda. My model has the potential to help organizations make decisions about policies to prevent food crises.”
And that makes her success all the more rewarding, she said.
“That is one of the motivating parts of why I do my research,” Petersen said. “I want to make a difference in my community and the world at large. This project opened my eyes to the power of data analysis and computing to impact real change in the world.”
While there are plenty of crop predictors available, what makes her model special is that it is adaptable to any part of the world and to any crop.
“Current early warning systems are highly tuned to specific crop types, and climate and conditions,” Petersen said. “They rely on very large amounts of crop-specific data and they even rely on observers to report on crop health. But, especially during COVID, a lot of these systems are breaking down. I took a step back (and) created a model that was much simpler and could be applied anywhere.”
It may be simpler, but it required digesting some 15 terabytes of data.
“It took me a year and a half to create the code,” she said.
Petersen was selected from nearly 2,000 candidates across the U.S. That group was trimmed to 300, then a final 40, which also included Valencia High School grad Makayla Gates, whose “Comparative Analysis of Lovastatin Introduction Through Botanical Dietary Supplementation in Apis mellifera for Treatment of Social Anxieties in Fragile-X Syndrome and Autism Spectrum Disorder Patients” earned her $25,000.
Those final 40 were put through a rigorous screening process in which judges were seeking the entrants’ knowledge beyond their own research, said Maya Ajmera, president and CEO of the Society for Science & the Public, which oversees the competition.
“The judges are not only looking for an established body of research, but that they also have an incredible knowledge of science and engineering,” she said. “They’re looking for how well do you know science? Are you a ruthless disciplinarian and well rounded?”
And that’s what truly separated Petersen from the competition, Ajmera said.
“Lillian just stood out for her tenacity with the judges. Her research that she has completed has been incredibly hard and she stuck with the research project over time. She’s trying to solve the biggest question that has been at the heels of humanity for a long, long time.”
Lillian Kay Petersen presents her winning project to judges in the 2020 Regeneron Science Talent Search. Petersen, 17, of Los Alamos, won a $250,000 prize.