Never in his wildest dreams did Pascal Caraccioli Salinas imagine that he would be working with the U.S. Geological Survey as an undergrad, let alone an internship exploring geothermal resources — his dream career.
"I thought I wasn't going to be able to do this until after grad school," Caraccioli Salinas said.
For the internship with USGS's Geology, Minerals, Energy and Geophysics Science Center, he’s using machine learning to predict the favorability of geothermal resources in Nevada's Great Basin.
The opportunity came about through the Climate and Community Resilience Internship Program, a partnership between PSU's Institute for Sustainable Solutions and the Louis Stokes Alliance for Minority Participation (LSAMP) that provides minoritized students in STEM fields like Salinas with 9-month paid internships in climate resilience and water resources fields. USGS matched him with an internship that combined his passions for geology and data science.
Caraccioli Salinas, who studied geology in his native Chile before immigrating to the U.S. four years ago, says he's long been interested in tapping into geothermal potential as a way to reduce dependence on fossil fuels and mitigate climate change. He sees machine learning as a faster, cheaper and more accurate alternative to exploratory drilling, but knew he needed a more in-depth knowledge of statistics, math and computer science to take his career in that direction.
"You need more of a mathematical and data-driven approach to find these resources," he said. "I had this theoretical approach when I studied geology, but I needed the math."
By chance, he landed at PSU in spring 2022 after his husband's job required their family to relocate to Oregon. He decided to pursue a degree in data science.
Caraccioli Salinas says there's no substitute for on-the-job training and he's been relishing the experience with USGS. Over the past few months, he's been learning the codes and getting familiar with the dataset, always thinking about the geology behind the codes and math.
"These algorithms and tools developed for machine learning didn't think about geology, but we can try to tell the machine, 'OK, this is a site where we know there's geothermal. What are the common factors here that might also be common at another site that we haven't explored,’" he said.
Over the next few months, he'll be using the data in machine learning to predict resource favorability and assess which datasets are the most important in making those decisions. He's looking forward to being able to submit a conference paper to a geothermal workshop by the internship's end. He credits his supervisor and mentor, Stanley Mordensky, and the rest of the USGS geothermal team — Cary Lindsey, Erick Burns and Jacob DeAngelo — with guiding and helping him as he learns the ropes.
"We don't have experiences like this in Chile," he said. "Studying geology, you think about USGS, the research they have and everything they do, but working there is something you only dream of. And then they opened up this internship through this program and now I'm a junior researcher. I never thought it would be possible, and it's only happening because of PSU. I'm extremely grateful."