As the study of the universe evolves and the data sets get larger and more complex, a new breakthrough means researchers can analyze huge data sets with just a laptop and a few hours.
Dr. Marco Bonici, a postdoctoral fellow at the Waterloo Centre for Astrophysics at the University of Waterloo, led an international team that designed Effort.jl, which stands for EFfective Field theORy surrogate. The device combines state-of-the-art numerical methods and clever preprocessing strategies to achieve exceptional computational performance with the precision needed in the field of cosmology. The team designed this novel and efficient emulator for the Effective Field Theory of Large-Scale Structure (EFTofLSS) to analyze data sets faster than ever before.
Bonici came up with the idea to create the emulator after spending hundreds of hours throughout his career running multiple computational models every time a change was made to the parameters. Often, one small change could lead to days of additional computational analysis, making it a slow process that required a lot of time and patience, and sometimes came at a high cost.
Portion of data from the Dark Energy Spectroscopic Instrument (DESI) shows objects in space up to billions of light-years away from Earth, which is in the centre of this image. Bright galaxies appear in yellow, the orange area represents luminous red galaxies, emission-line galaxies are blue, and quasars are in green. (Credit: DESI Collaboration/DOE/KPNO/NOIRLab/NSF/AURA/C. Lamman)
"Using Effort.jl, we can run through complex data sets on models like EFTofLSS, which have previously needed a lot of time and computer power," Bonici said. "With projects like DESI and Euclid expanding our knowledge of the universe and creating even larger astronomical datasets to explore, Effort.jl allows researchers to analyze data faster, inexpensively and multiple times while making small changes based on nuances in the data."
Emulators are trained shortcuts that mimic the behaviour of the full and expensive simulations, but run much faster, empowering scientists to test multiple cosmic scenarios without waiting hours for each one. Emulators also make it possible to use advanced techniques like gradient-based sampling to explore complex models efficiently.
"We were able to validate the predictions coming out of Effort.jl by aligning them with those coming out of EFTofLSS," Bonici said. "The margin of error was small and showed us that the calculations coming out of Effort.jl are strong. Effort.jl can also handle observational quirks like distortions in data and can be customized very easily to the needs of the researcher."
As smart as this tool is, it won't replace the physics knowledge of the cosmologists who input and interpret the data. While the tool can make predictions, the knowledge from the researchers and the parameters they set make it truly powerful.
Looking ahead, Effort.jl is poised to analyze next-generation cosmological datasets and to support joint analyses with complementary tools. Possible future applications include weather and climate forecasting.
The paper, Effort: a fast and differentiable emulator for the Effective Field Theory of the Large Scale Structure of the Universe, appears in the Journal of Cosmology and Astroparticle Physics.
Credit for banner image: NASA, ESA, the Hubble Heritage Team (STScI/AURA), A. Nota (ESA/STScI), and the Westerlund 2 Science Team
