Jeffrey Donatelli, a staff scientist in the Applied Mathematics and Computational Research Division (AMCR), is a 2024 Department of Energy Early Career Research Program (ECRP) awardee. With this award, he will be developing new computational methods for analyzing experimental data.

Now in its fourteenth year, the ECRP award supports exceptional researchers during critical stages of their formative work by funding their research for five years.

With advancements in technology, experiments are producing data at increasingly fast rates, involving more complex physics and highly sensitive measurements. As a result, significant challenges are emerging in extracting accurate and detailed scientific information from this data. Donatelli’s ECRP project, ‘Multi-Tiered Algorithms for Solving Extreme-Scale Inverse Problems Emerging from New Experiments,’ aims to develop new classes of data-analysis algorithms that leverage the mathematical structures of the physics and properties of emerging experiments to overcome these challenges.

“This project addresses a range of important problems I’ve encountered while analyzing data from various Department of Energy experiments,” said Donatelli. “This work will provide scientists with the capability to extract information about physical phenomena, biological specimens, and materials in unprecedented detail from data that are too large and complex for existing data-analysis techniques to handle.”

Growing up, Donatelli was always interested in science and originally considered pursuing a career in physics and astronomy. However, while pursuing his undergraduate degree at the University of Maryland, College Park, he developed a profound interest in the puzzles that arise when solving math problems. This interest ultimately led him to major in mathematics.

He continued his math education at UC Berkeley, where he became fascinated with the impact that mathematics can have on scientific applications. Under the mentorship of UC Berkeley Mathematics Professor and Berkeley Lab Mathematics Department Head James Sethian, Donatelli became a graduate student researcher at Berkeley Lab and met researchers from various scientific domains who needed new mathematical techniques to analyze different kinds of experimental data. Finding this work extremely interesting and rewarding, he decided to focus his Ph.D. thesis work on developing new mathematical algorithms for solving these experimental data analysis problems. Donatelli’s graduate research was supported by a DOE Computational Science Graduate Student Fellowship (CSGF).

After receiving his Ph.D. in applied mathematics, Donatelli continued his research as a postdoc at Berkeley Lab, where he helped form the Center for Advanced Mathematics for Energy Research Applications (CAMERA), which brings together teams of mathematicians, domain scientists, and experimentalists to tackle critical mathematical problems in DOE experimental science. CAMERA provided him with further exposure to the emerging mathematical challenges in data analysis, which served as key scientific drivers for his ECRP project. Donatelli now serves as the deputy director and math lead of CAMERA and leads AMCR’s Math for Experimental Data Analysis group, where he steers several important efforts in developing and deploying new mathematical solutions for analyzing data throughout DOE experimental facilities.

“One of my favorite things about being at Berkeley Lab, and the reason I’ve stayed here so long, is the opportunity to interact with people from a wide range of scientific fields. I enjoy learning about what’s important to them, understanding the challenges they face, and then finding new and interesting mathematical solutions to address these problems,” said Donatelli.

He adds that what inspired him to apply for the ECRP and propose this project is the alignment between the Department of Energy’s mission and the emerging mathematical topics he is eager to explore.

“DOE is investing in experimental facility upgrades and measurement technologies that can capture fundamentally new science, but the data being produced is becoming increasingly challenging to analyze. From a mathematical perspective, addressing these challenges involves exciting new research areas spanning machine learning, optimization, linear algebra, harmonic analysis, and statistics. This award provides an opportunity to further explore these research areas and develop new foundational mathematical ideas and solutions that will enable a host of exciting scientific breakthroughs,” said Donatelli.