NAM2019
  • NAM2019
    • Registration
    • Key Dates & Outline Schedule
    • Practical Information
    • Exhibitors
    • Grants & Bursaries
    • Contacts
  • Science
    • Science Programme
    • Parallel Sessions
    • Plenary Talks
    • Community Session
    • Special Lunches
    • Posters
    • Presenter Guidelines
  • Social
    • What's On
    • Welcome Reception
    • RAS Awards Dinner
  • Media
  • Outreach
    • Outreach and Education Day
    • Fringe Event
    • School Visit Day
  • Lancaster
    • Travel
    • Accommodation
    • Childcare
    • Campus Map
    • About Lancaster
    • Code of Conduct
  • NAM2019
    • Registration
    • Key Dates & Outline Schedule
    • Practical Information
    • Exhibitors
    • Grants & Bursaries
    • Contacts
  • Science
    • Science Programme
    • Parallel Sessions
    • Plenary Talks
    • Community Session
    • Special Lunches
    • Posters
    • Presenter Guidelines
  • Social
    • What's On
    • Welcome Reception
    • RAS Awards Dinner
  • Media
  • Outreach
    • Outreach and Education Day
    • Fringe Event
    • School Visit Day
  • Lancaster
    • Travel
    • Accommodation
    • Childcare
    • Campus Map
    • About Lancaster
    • Code of Conduct

Thursday

Schedule

id
date time
PM1
16:15
Abstract
Deep-learned baryons: Hybrid emulators for high-speed cosmological simulations
Thursday

Abstract details

id
Deep-learned baryons: Hybrid emulators for high-speed cosmological simulations
Date Submitted
2019-02-21 19:32:52
Ben
Moews
Institute for Astronomy, University of Edinburgh
Machine Learning in Astrophysics
Talk
Ben Moews and Romeel Dave
While cosmological dark matter-only simulations relying solely on gravitational effects are comparably fast to compute, baryonic properties in simulated galaxies require complex hydrodynamical simulations that are computationally costly to run. To solve this issue, we merge an extended version of the equilibrium model, an analytic formalism describing the evolution of the stellar, gas, and metal content of galaxies, into a deep learning framework and retrieve fully Bayesian posteriors for baryon cycling parameters. In doing so, we are able to recover more properties than the analytic formalism alone can provide, creating a high-speed hydrodynamical simulation emulator that populates galactic dark matter halos in N-body simulations with baryonic properties. Our results demonstrate that this novel hybrid system enables the fast completion of dark matter-only simulations by accurately mimicking full hydrodynamical suites of choice, offering an orders-of-magnitude acceleration of commonly deployed simulations in cosmology.

RAS Logo

Lancaster University Logo

STFC logo

Science Programme

  • Monday
  • Tuesday
  • Wednesday
  • Thursday
  • Posters

All attendees are expected to show respect and courtesy to other attendees and staff, and to adhere to the NAM Code of Conduct. To report harassment or violation of the code of conduct please click here.

© 2022 Royal Astronomical Society

Login