OpenUniverse2024 Introduction¶
OpenUniverse2024 is a collection of large-scale simulated observations designed to resemble the expected data products of the Nancy Grace Roman Space Telescope and the Vera C. Rubin Observatory. Both observatories are simulated over the same roughly 70 square degree patch of sky, so the same objects appear in Roman’s high-resolution near-infrared imaging and in Rubin’s wide optical ugrizy imaging, and the two can be analyzed together. The dataset includes wide-area imaging, a time-domain survey that revisits a smaller footprint many times, and mock source catalogs. For Roman there are truth catalogs of the input object properties, noiseless truth images, calibrated single-epoch images, and coadds. For Rubin there are raw pixel data, calibrated exposures, coadds, and photometric catalogs.
All data are public, hosted on AWS in the S3 bucket s3://nasa-irsa-simulations/openuniverse2024/ (us-east-1 region), and require no credentials.
Catalogs are stored in Parquet, hdf5, and fits format.
Images are multi-extension FITS.
Documentation: OpenUniverse2024 Overview at IRSA
If you use the OpenUniverse2024 dataset in your work, please follow the “How to Cite” instructions on the Registry of Open Data on AWS: OpenUniverse2024 page.
Workshop Resources¶
The below materials were presented in a September 2026 workshop that provided a hands-on introduction to working with the OpenUniverse2024 dataset. Rather than surveying the simulations from a distance, we worked through Python Jupyter notebooks that access the data directly in the cloud and carry it through to a scientific result — locating a transient, building a light curve, fitting a spectral energy distribution. The emphasis throughout is on combining complementary survey data: what you can measure by using Roman and Rubin together that you cannot get from either one alone, and how to structure that work as a scalable cloud workflow rather than a local download.
Agenda¶
The workshop consisted of the following five segments. Slides and videos are provided where available.
- What Is OpenUniverse2024?
Recording of the talk “What Is OpenUniverse2024?” by Jessica Krick.
- Introduction to Fornax
Recording of the talk “Introduction to Fornax” by Troy Raen.
- Accessing OpenUniverse2024 Data
Notebook walk-through (no slides).
Recording of the “Quickstart” notebook walk-through by Jessica Krick.
- Visualizing OpenUniverse2024 Data
Notebook walk-through (no slides).
Recording of the “Firefly Visualization” notebook walk-through by Jaladh Singhal.
- Self-paced tutorial exploration
Participants worked through notebook tutorials at their own pace (no slides or video).
Fornax Links¶
Get help by posting in the Fornax forum’s OpenUniverse Workshop category (private to workshop participants and Fornax staff).
OpenUniverse2024 dataset questions
Data access questions
Tutorial questions
Fornax technical issues
Before You Start¶
The workshop runs entirely in the Fornax Science Console, NASA’s cloud-based science platform, so there is nothing to install on your own machine. The Fornax Quick Start Guide walks through the five steps of getting an account, logging in, starting a server session, opening a notebook, and shutting the server down again.
Registered workshop participants already have a Fornax account and can skip step 1 and start at step 2, “Log In”.
Everyone else should begin at step 1, “Get an Account”, and register at signup
.fornax .sciencecloud .nasa .gov. Requests made with a nasa.govaddress are approved immediately; other institutional addresses are reviewed by Fornax staff and can take up to two business days, so allow time before you plan to work through the material.
Tutorials¶
Workshop participants will explore the following tutorials, all part of the IRSA’s Simulated Data Tutorials. They assume some familiarity with Python and Jupyter notebooks, but not with Roman, Rubin, or cloud data access.
The tutorials are available on the Fornax Science Console in the directory ~/fornax-notebooks/irsa-tutorials/simulated-data/OpenUniverse2024/.
They should automatically connect to the irsa-tutorials environment kernel, which has all libraries and other dependencies pre-installed.
To run the tutorials outside of Fornax, first clone the irsa-tutorials repo. Then, follow the instructions in the first code cell of each notebook to install the dependencies.
Quickstart¶
Access OpenUniverse2024 Roman and Rubin images and catalogs.
Connect to the public S3 bucket and navigate the Roman image directory tree
Inspect the structure of a Roman FITS file and display a gallery of images
Load SNANA, galaxy, and galaxy-flux parquet catalogs using HEALPix region indexing
Query Roman and Rubin images overlapping a sky position via astroquery and SIA
Firefly Visualization¶
Use Firefly to explore the structure of the simulations and visualize their content.
Access cloud-hosted Roman and Rubin simulated images and find those covering a sky position with SIA
Launch an interactive Firefly instance inside JupyterLab
Display a Rubin image and overplot ds9 regions showing where the Roman data fall
Overplot Parquet truth catalogs and filter them to pick out galaxies of interest
Build three-color images and control stretch, pan, and zoom from Python
Roman Time Domain Survey Supernova¶
Explore the Roman Time Domain Survey observation sequence and animate a simulated SN Ia across epochs.
Read the Roman TDS Observation Sequence File to see how the survey is laid out in space and time
Filter the SNANA transient catalog for a SN Ia peaking during an observed window
Query SIA for every Roman image covering the supernova during its bright phase
Extract aligned cutouts from 31 epochs, correcting for the varying focal plane orientation
Animate the cutouts into a GIF showing the supernova brighten and fade
TDE Light Curve¶
Locate a simulated TDE, retrieve Roman images, and build a multi-epoch light curve.
Select a TDE from the SNANA parquet catalog and locate its host galaxy
Query Roman TDS images covering the host position using astroquery and SIA
Perform aperture photometry on individual Roman images
Build and visualize a multi-epoch infrared light curve
Display full images and cutouts centered on the TDE host
SED Fitting¶
Fit spectral energy distributions for supernova host galaxies using Prospector.
Access multiband photometry from OpenUniverse2024 (Roman + Rubin)
Convert simulated fluxes to physical units and plot broadband SEDs
Perform SED fitting using Prospector and FSPS
Compare host galaxy properties across supernova types
Visualize results as SED plots and histograms of fitted stellar masses
Questions¶
Get help by posting in the Fornax Helpdesk (public) or the OpenUniverse Workshop category (private to workshop participants and Fornax staff).
OpenUniverse2024 dataset questions
Data access questions
Tutorial questions
Fornax technical issues