🔭 astrophysics

10 Years of Archival High-Resolution NIR Spectra: The Raw and Reduced IGRINS Spectral Archive (RRISA)

This paper introduces the Raw and Reduced IGRINS Spectral Archive (RRISA), a public repository providing access to raw and reduced high-resolution near-infrared spectra from the IGRINS instrument across multiple telescopes, alongside the release of the improved IGRINS PLP v3 data reduction pipeline.

Original authors: Erica Sawczynec, Kyle F. Kaplan, Gregory N. Mace, Jae-Joon Lee, Daniel T. Jaffe, Chan Park, In-Soo Yuk, Moo-Young Chun, Soojong Pak, Narae Hwang, Ueejeong Jeong, Hwihyun Kim, Hyun-Jeong Kim, Kang-Min

Published 2026-06-11

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Erica Sawczynec, Kyle F. Kaplan, Gregory N. Mace, Jae-Joon Lee, Daniel T. Jaffe, Chan Park, In-Soo Yuk, Moo-Young Chun, Soojong Pak, Narae Hwang, Ueejeong Jeong, Hwihyun Kim, Hyun-Jeong Kim, Kang-Min Kim, Sanghyuk Kim, Huynh Anh N. Le, Hye-In Lee, Sungho Lee, Heeyoung Oh, Jae Sok Oh, Byeong-Gon Park, Woojin Park, Young-Sam Yu

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

Imagine the Immersion GRating INfrared Spectrometer (IGRINS) as a super-powered, high-definition camera for the invisible infrared universe. Just as a regular camera captures visible light to take photos, IGRINS captures the "heat signatures" and chemical fingerprints of stars, planets, and gas clouds in the dark, dusty regions of space that our eyes can't see. It's so precise it can split light into thousands of tiny colors, acting like a prism that reveals secrets hidden in the darkness.

For the last decade, this camera has been traveling between three different "studios" (telescopes) in the US, South Korea, and Chile, snapping millions of pictures. However, until now, these photos were locked away in private filing cabinets, accessible only to the specific scientists who took them.

This paper introduces RRISA (Raw and Reduced IGRINS Spectral Archive), which is essentially opening the doors to a massive, public digital library. Here is what the paper actually does, broken down into simple concepts:

1. The Library (RRISA)

Think of RRISA as a giant, organized digital warehouse. Before this, if you wanted to see a specific star's infrared photo, you had to know exactly who took it and ask them nicely. Now, RRISA puts everything on a public shelf.

The Raw Files: These are the "unprocessed negatives." They are the raw data straight from the camera, full of noise and imperfections, just like a photo straight out of a camera before editing.
The Reduced Files: These are the "edited masterpieces." Scientists have cleaned up the noise, fixed the colors, and sharpened the image so it's ready for anyone to use.
The Cross-Reference: Imagine finding a photo of a stranger in a crowd. RRISA doesn't just show the photo; it attaches a "Wanted" poster with their name, address, and history. It links every star IGRINS looked at with existing databases (like SIMBAD, Gaia, and 2MASS) to tell you exactly who the object is, how bright it is, and how fast it's moving.

2. The Cleaner (IGRINS PLP v3)

Taking photos in space is messy. The camera gets hit by tiny cosmic bullets (cosmic rays), the detector gets a static-like "flicker" (pattern noise), and the camera shifts slightly as the telescope moves (flexure).

The authors introduced Version 3 of their cleaning software (IGRINS PLP v3), which acts like a master photo editor with three new superpowers:

Cosmic Ray Removal: Imagine a photo where a bird flew in front of the lens for a split second, leaving a white streak. This software finds those streaks and "fills in" the missing pixels using information from other photos taken at the same time, making the streak disappear.
Pattern Noise Removal: Sometimes the camera sensor has a weird, repeating grid pattern, like a watermark on a piece of paper. The new software detects this pattern and subtracts it, leaving a clean image.
Flexure Correction: As the telescope tilts to look at different stars, the camera inside shifts slightly, like a shaky hand. This causes the "atmospheric fingerprints" (telluric lines) to blur. The new software acts like a steady hand, realigning every single photo so they stack perfectly, resulting in a much sharper, clearer view of the star's true light.

3. The Toolkit (Tutorials and Python Tools)

The paper also provides a "user manual" and a set of tools for anyone who wants to use this library.

They offer Python tools (named µler and gollum) that act like a Swiss Army knife for astronomers. µler helps you clean up and plot the data, while gollum lets you compare your star photos against computer-generated models to figure out the star's temperature and composition.
They provide tutorials that teach you how to find the best photos, how to download them without needing to log into complex systems, and how to fix common problems (like when the atmospheric correction isn't perfect).

The Bottom Line

This paper is an announcement that the IGRINS team has spent ten years taking high-quality infrared photos of the universe, and they have finally built a public library (RRISA) to share them. They have also upgraded their photo-editing software (PLP v3) to make the images cleaner and sharper than ever before. Now, instead of a few scientists holding the keys to this data, the entire astronomy community can walk in, pick up a star, and start studying it.

What the paper does NOT claim:

It does not claim to have discovered new planets or stars; it claims to have made the data for them accessible.
It does not claim to have solved all atmospheric problems; it notes that some "fuzzy" effects from the Earth's atmosphere still remain if the target and the reference star are at different heights in the sky.
It does not discuss medical or clinical uses; the focus is strictly on astronomical data reduction and archival access.

Technical Summary: The Raw and Reduced IGRINS Spectral Archive (RRISA) and IGRINS PLP v3

Problem
The Immersion GRating INfrared Spectrometer (IGRINS) has been operational since 2014, conducting high-resolution ( $R \sim 45,000$ ) near-infrared spectroscopy (1.45–2.45 $\mu$ m) across three telescopes: the 2.7m Harlan J. Smith Telescope (HJST) at McDonald Observatory, the 4.3m Lowell Discovery Telescope (LDT), and the 8.1m Gemini South Telescope. Despite generating over 17,700 observations of science targets and telluric standards across more than 1,000 nights, the data lacked a centralized public repository. Consequently, only a small fraction of the dataset was accessible to the broader astronomy community. Furthermore, the data reduction pipeline required updates to address specific instrumental challenges, including cosmic ray (CR) contamination, detector readout pattern noise, and instrumental flexure, which can degrade telluric correction and spectral resolution.

Methodology
The authors present two primary technical developments: the creation of the Raw and Reduced IGRINS Spectral Archive (RRISA) and the release of version 3 of the IGRINS Pipeline Package (IGRINS PLP v3).

RRISA Construction: The archive aggregates raw and reduced data from August 2014 to May 2023. The pipeline for creating RRISA involved:
- Logging: Automated scraping of headers from H-band raw images and digital night logs to track every raw frame.
- Target Identification: A hybrid approach was used to identify targets. While automated searches against SIMBAD were performed, approximately 20% of recorded names failed to match due to non-standard nomenclature or missing catalog entries. These were manually corrected using observer logs and coordinates.
- Cross-Matching: Identified targets were cross-matched with SIMBAD, 2MASS, Gaia DR3, APOGEE2 DR17, and PASTEL catalogs using astroquery. The process prioritized SIMBAD coordinates to ensure accuracy, flagging matches where identifiers did not align across catalogs.
- Data Packaging: Raw data, reduced spectra, and calibration files are hosted on Box, with access links provided via GitHub. The archive includes over 3,500 unique target observations and >790 unique standard stars.
IGRINS PLP v3 Improvements: The pipeline introduces three specific algorithmic enhancements:
- Cosmic Ray (CR) Correction: Utilizing astroscrappy, CRs are identified in individual exposures. Affected pixels are "filled in" using the median value of overlapping exposures from the same nod position (A/B or ON/OFF). For single-exposure cases, a median-smoothed clean array is used.
- Pattern Noise Removal: To address detector readout patterns (repeating every 32 channels), the pipeline removes the constant term using overscan columns and subtracts the repeating pattern derived from inter-order pixels. This involves mirroring and folding stacked A-B frames into 64-column segments to isolate and remove the pattern.
- Flexure Correction: A new frame-by-frame correction uses OH sky-emission lines as a reference. By cross-correlating the OH lines in each exposure against a reference SKY frame, the pipeline measures pixel shifts in the dispersion direction. Sub-pixel shifts are corrected via linear interpolation, while integer shifts are handled by rolling the echellogram. This minimizes misalignment between science targets and telluric standards.

Key Contributions

Public Data Access: RRISA provides the first centralized, public access to both raw and reduced IGRINS data, encompassing nearly 1,093 nights of observations and ~4,000 hours of integration time.
Enhanced Data Products: The archive includes cross-matched identifiers and astrophysical parameters from major catalogs (Gaia, APOGEE2, etc.), significantly enriching the metadata available for IGRINS targets.
Pipeline Updates: IGRINS PLP v3 is released as open-source software, featuring improved CR removal, pattern noise subtraction, and a robust flexure correction routine that reduces telluric residuals.
User Support: The authors provide Python-based tools (muler and gollum) for data analysis and visualization, along with tutorials for parsing RRISA metadata, downloading data, and performing custom reductions.

Results

Archive Statistics: The RRISA dataset includes 141,953 raw frames resulting in 35,460 reduced H- and K-band spectra. The sample covers 3,432 unique targets and 792 unique standard stars.
Pipeline Performance:
- CR Removal: The "filling in" method produces cleaner spectra free of holes where CRs impacted pixels.
- Pattern Noise: Visual inspection of 2D echellograms (Figure 2) demonstrates a dramatic reduction in the detector readout pattern after correction.
- Flexure Correction: Application of the flexure correction significantly improves the alignment of telluric absorption lines and sky emission lines. Figure 3 illustrates that correcting flexure reduces residuals when dividing a target spectrum by a standard star, particularly in the K-band where flexure shifts can reach up to 1.5 pixels.

Significance
The paper establishes RRISA as a critical resource for the community, democratizing access to a decade of high-resolution NIR spectroscopy. By standardizing data reduction through IGRINS PLP v3, the authors ensure that the archived data is of higher quality, specifically regarding telluric correction and instrumental artifacts. The integration of multi-wavelength catalog data facilitates broader scientific applications, ranging from exoplanet atmospheres and stellar parameters to planetary nebulae and young stellar objects. The authors note that a future data release in May 2025 is planned to include data from Gemini South currently under its proprietary period.

1. The Library (RRISA)

2. The Cleaner (IGRINS PLP v3)

3. The Toolkit (Tutorials and Python Tools)

The Bottom Line

More like this