🔭 astrophysics

Simulating continuum-based redshift measurement in the \textit{Roman's} High Latitude Spectroscopy Survey

This study simulates the \textit{Roman} Space Telescope's High Latitude Spectroscopy Survey to demonstrate that its G150 grism can achieve a redshift recovery completeness of at least 50% for red, quiescent galaxies up to $z=2.5$ with F158 magnitudes brighter than 20.2, while analyzing how survey parameters influence measurement accuracy and completeness.

Original authors: Zhiyuan Guo, Bhavin Joshi, Chris. W. Walter, M. A. Troxel

Published 2026-02-25

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Zhiyuan Guo, Bhavin Joshi, Chris. W. Walter, M. A. Troxel

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: A Cosmic "Blur" Camera

Imagine the Nancy Grace Roman Space Telescope as a giant, super-powerful camera floating in space. Its job is to take pictures of the universe to understand how galaxies move and grow.

Most of the time, this camera takes clear, sharp photos (imaging). But for this specific study, the scientists are testing a special "lens attachment" called a grism. Think of the grism like a prism in a pair of glasses. When light from a galaxy hits the grism, it doesn't just make a dot; it smears the light out into a long, colorful rainbow (a spectrum).

The Problem:
In space, galaxies are crowded. If you look at a busy city street through a prism, the light from one person's red shirt might smear across the street and mix with the light from a blue car next to them. In astronomy, this is called contamination. The "rainbow" of one galaxy gets messy because it overlaps with its neighbors.

The Goal:
The scientists want to know: Can we still figure out how fast these galaxies are moving away from us (their "redshift") just by looking at the smooth, continuous color of their light (the continuum), even when the picture is a bit messy?

They are specifically interested in red, quiet galaxies. These are the "old folks" of the universe—galaxies that stopped making new stars long ago. They are crucial for mapping the structure of the universe, but they are hard to study because they don't have bright, flashing lights (like emission lines) to help us identify them.

The Experiment: Simulating the Mess

Since we can't wait 5 years for the telescope to actually launch and take these pictures, the team built a virtual reality simulation of the universe.

The Setup: They took a "noiseless" (perfectly clear) map of the universe and ran it through a computer program called Grizli. Grizli is like a sophisticated photo editor that knows exactly how the Roman telescope's prism works.
The Simulation: Grizli took the clear images and smeared them out, added background noise (like static on an old TV), and mixed the light from neighboring galaxies together, just like the real telescope will.
The Test: They then tried to use the software to "un-smear" the light and calculate the distance (redshift) of the old, red galaxies.

The Three Rules for a "Good" Result

To decide if the simulation worked, the scientists set three strict rules, like a bouncer at a club:

The Signal Must Be Loud: The galaxy needs to be bright enough to be heard over the cosmic "static" (Signal-to-Noise ratio $\ge$ 5).
The Answer Must Be Clear: When the computer guesses the distance, it shouldn't be confused. It needs to have one clear, dominant answer, not two or three equally likely guesses.
The Accuracy Must Be High: The guess needs to be within 1% of the true distance.

The Findings: How Deep Can We See?

The team ran the simulation with different amounts of "exposure time" (how long the camera shutter stays open).

The Baseline: With the standard plan (about 46 minutes of total observation time per spot), the telescope can successfully measure the distance of red galaxies that are brighter than magnitude 20.2.
- Analogy: Imagine trying to read a street sign. At night, with a standard flashlight, you can read the sign clearly if it's within 50 feet. If it's 60 feet away, the letters get too blurry. The "20.2 magnitude" is that 50-foot limit.
The Trade-off: If they spend more time looking (longer exposure), they can read signs that are further away (fainter galaxies). If they spend less time, they have to stick to the bright ones.
The "Missing Exposures" Problem: The survey plan isn't perfect; some parts of the sky will be photographed 8 times, while others are only photographed 5 or 6 times. The study found that for every time you miss a photo, the "reading distance" gets shorter by about 0.1 magnitude. It's like taking fewer photos of a moving car; you get a blurrier picture and have a harder time guessing its speed.

Why Does This Matter?

The scientists found that even with the messy, overlapping light, they can successfully measure the distance of about 50% of the red galaxies that are bright enough (up to magnitude 20.2).

The "So What?"

The Numbers: Over the huge area the telescope will scan (2,000 square degrees), this method could identify roughly 120,000 red, quiet galaxies.
The Utility: While 120,000 might not seem like a lot compared to the billions of stars, these are the specific galaxies needed to map the "skeleton" of the universe. They act as tracers to help scientists understand Dark Energy and how the universe is expanding.
The Bonus: Even if these galaxies aren't perfect for every single study, they can be combined with other data sets to make the whole picture much clearer.

Summary in a Nutshell

The paper proves that the Roman Space Telescope's special prism camera is a powerful tool. Even though the images will be messy and crowded, the software (Grizli) is smart enough to clean up the noise and accurately measure the distance of the universe's "old, red" galaxies. This means we will get a massive, high-quality map of the universe's structure, helping us solve the mystery of how the cosmos is evolving.

1. Problem Statement

The Nancy Grace Roman Space Telescope (Roman) is designed to conduct the High-Latitude Spectroscopic Survey (HLSS) using a slitless grism (G150) to measure redshifts. While the primary scientific target of the HLSS is emission-line galaxies (ELGs) for Baryon Acoustic Oscillation (BAO) and Redshift Space Distortion (RSD) studies, red, quiescent galaxies (which lack strong emission lines) represent a fundamental component of the cosmic mass assembly history.

The core challenge addressed in this paper is determining the capability of the Roman G150 grism to measure redshifts for these continuum-dominated galaxies. Unlike ELGs, red galaxies rely on spectral breaks (specifically the 4000 Å break) for redshift determination. The authors aim to assess:

The efficiency of recovering redshifts for quiescent galaxies in the range $0.5 \le z \le 2.5$ .
How survey parameters, specifically exposure time and the number of exposures, impact redshift accuracy and completeness.
Whether continuum-based redshifts can meet the rigorous requirements needed for cosmological analyses like galaxy clustering.

2. Methodology

The study employs a comprehensive end-to-end simulation pipeline using the Grizli (Grism Redshift and Line Analysis) software package.

Input Data:
- Direct Images: Utilized noiseless F158 band direct images from the Troxel et al. (2023) simulation, representing the Roman High-Latitude Imaging Survey (HLIS) combined with LSST data.
- Spectral Energy Distributions (SEDs):
  - Quiescent Galaxies: Assigned old, quiescent spectra generated via the python-fsps package for galaxies with $m_{F158} < 22.5$ and $0.5 \le z \le 2.5$ . These were assigned at a 60% rate to ensure statistical significance while maintaining diversity.
  - Contaminants: Used EAZY SEDs from the 3D-HST survey for other galaxies (to simulate emission lines and background contamination) and Pickles stellar library spectra for stars.
- Configuration: Simulated the Roman WFI (Wide-Field Instrument) across 18 sensors at four roll angles (0°, 5°, 170°, 175°) to mimic the reference survey design.
Simulation Process:
- Generated 2D dispersed grism images including zodiacal background, dark current, and readout noise.
- Modeled and subtracted spectral contamination from neighboring sources using Grizli's iterative polynomial fitting.
- Performed redshift fitting on the cleaned 2D spectra.
Redshift Estimation & Validation:
- Primary Tool: Grizli (using 2D spectra).
- Validation Tool: Bagpipes (using optimally extracted 1D spectra) to verify that 1D results match 2D results.
- Selection Criteria: A galaxy was considered a "successful" redshift recovery only if it met three conditions:
  1. Accuracy: $\sigma_z = |z_{fit} - z_{true}| / (1 + z_{true}) \le 0.01$ .
  2. Signal-to-Noise: $S/N \ge 5$ .
  3. Uniqueness: A single dominant peak in the redshift likelihood function $P(z)$ .

3. Key Contributions

First Continuum-Focused Roman Simulation: This is one of the first studies to specifically quantify the performance of Roman's grism for continuum-based redshifts of quiescent galaxies, rather than just emission lines.
Parameter Sensitivity Analysis: The paper systematically quantifies how varying exposure times and the number of available exposures (dithering patterns) degrade or improve redshift recovery rates.
Validation of 1D vs. 2D Fitting: The study confirms that redshifts derived from 1D spectra (via Bagpipes) are consistent with 2D spectral fitting (via Grizli), validating the use of standard extraction techniques for Roman data.
Survey Strategy Optimization: Provides concrete metrics on how many exposures are required to reach specific magnitude limits for reliable redshifts, offering guidance for future survey planning.

4. Key Results

Baseline Performance: For the reference HLSS design (total exposure time of ~2,776 seconds per source across 8 exposures), the survey achieves a 50% redshift recovery completeness for quiescent galaxies brighter than $m_{F158} \approx 20.2$ .
Exposure Time Scaling: Increasing the total exposure time improves the depth of the survey:
- 1.5x exposure ( $\sim$ 4,164s): 50% completeness at $m_{F158} \approx 20.4$ .
- 2.0x exposure ( $\sim$ 5,552s): 50% completeness at $m_{F158} \approx 20.6$ .
- 3.0x exposure ( $\sim$ 8,328s): 50% completeness at $m_{F158} \approx 20.9$ .
Exposure Count Sensitivity: Reducing the number of available exposures (simulating areas with less than 8 dithers) results in a degradation of approximately 0.1 magnitude in the 50% completeness limit for every lost exposure.
Redshift Dependence: Redshift accuracy improves at higher redshifts ( $z \ge 1.4$ ). This is attributed to the 4000 Å break shifting into the optimal wavelength coverage of the Roman grism ( $1.0 - 1.93 \mu m$ ), providing a stronger spectral feature for template fitting.
S/N Threshold: A Signal-to-Noise ratio of $S/N \ge 5$ is identified as the critical threshold for achieving 90% completeness in redshift recovery.

5. Significance and Implications

Cosmological Yield: The authors estimate that the Roman HLSS will yield approximately 120,000 red, quiescent galaxies with accurate spectroscopic redshifts over the full 2000 deg² survey area (specifically in the $1 < z < 2$ range, yielding ~60 galaxies/deg²).
Scientific Utility: While this sample size may be insufficient for standalone BAO or galaxy clustering analyses (which typically require larger samples), it is highly valuable for:
- Calibration: Serving as a "gold standard" sample to calibrate photometric redshifts (e.g., for the RedMaGiC algorithm).
- Galaxy Evolution: Expanding the census of high-redshift quiescent galaxies to study their formation and assembly history.
- Hybrid Analyses: Combining with ELG samples to enhance the statistical power of large-scale structure studies.
Strategic Guidance: The results suggest that survey planners can trade off between survey area and depth. If the scientific goal prioritizes deep redshift measurements of fainter quiescent galaxies, increasing exposure time or prioritizing areas with more dithers is necessary.

In conclusion, the paper demonstrates that the Roman Space Telescope will be a powerful tool for studying the continuum properties of the red galaxy population, provided that survey strategies are optimized to ensure sufficient signal-to-noise and multiple exposure coverage.