🔭 astrophysics

Reference File Updates Following the Application of New Geometric Distortion and Walk Corrections I: Overview

This paper summarizes the re-derivation of 56 reference files for the Cosmic Origins Spectrograph (COS) following the implementation of improved geometric distortion, delta-geometric, and walk corrections, an effort aimed at enhancing the wavelength and flux calibration of FUV spectra.

Original authors: Nick Indriolo, William J. Fischer, Elaine Frazer, David French, Sten Hasselquist, Jaquline Hernandez, Christian I. Johnson, Darshan Kakkad, Lauren P. Miller, Marc Rafelski, Ravi Sankrit

Published 2026-02-11

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Nick Indriolo, William J. Fischer, Elaine Frazer, David French, Sten Hasselquist, Jaquline Hernandez, Christian I. Johnson, Darshan Kakkad, Lauren P. Miller, Marc Rafelski, Ravi Sankrit

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 Cosmic Map Update: Fixing the "Blurry Lens" of the Hubble Space Telescope

Imagine you are a professional photographer trying to take pictures of distant, tiny fireflies in a pitch-black forest. You have a very expensive camera, but there’s a problem: the lens has a slight, invisible warp to it. Because of this warp, when you look through the viewfinder, a firefly that is actually in the center looks like it’s slightly to the left. Even worse, if you try to measure how bright the firefly is, the warp makes it look dimmer or brighter than it actually is.

For years, astronomers using the Cosmic Origins Spectrograph (COS)—a high-tech "camera" on the Hubble Space Telescope—have been dealing with this exact problem. They knew the "lens" (the detector) had some geometric distortions and "walk" (where light shifts slightly as it hits the sensor).

This paper, COS 2026-01, is essentially the "Software Update" announcement for how we interpret those photos.

The Problem: The Shifting Ground

Think of the COS detector like a giant sheet of graph paper. To do science, astronomers need to know exactly which square (pixel) a photon (a particle of light) landed on.

In 2024, scientists realized they could make their "map" of this graph paper much more accurate. They created a new, better way to correct for the warping and shifting. However, there was a catch: once you change the map, every single instruction written on the old map becomes wrong.

If your old map said, "The treasure is at square A1," but your new, more accurate map shows that square A1 has actually shifted to square B2, you can't just use the old instructions anymore. You have to rewrite every single note, every measurement, and every guide in the entire manual.

The Solution: Rewriting the Manual

The researchers had to go back and "re-derive" 56 different reference files. You can think of these files as the "Instruction Manuals" for the telescope. These manuals tell the computer:

"Watch out for bad spots!" (Identifying "potholes" or dead pixels on the detector).
"Where is the light?" (Correctly identifying the wavelength/color of the light).
"How bright is it?" (Converting raw light counts into actual energy levels).

Because the telescope has aged over time (moving through different "Lifetime Positions"), they couldn't just do this once. They had to work backward through the telescope's history, from its most recent state back to its earliest days, ensuring that every era of its life had a perfectly calibrated manual.

The Result: A Sharper View of the Universe

So, what happens now that the "Software Update" is complete? It’s like switching from a blurry, old map to a high-definition GPS.

Better Precision: Before, if we tried to measure the speed of a star, our "ruler" was a bit shaky. Now, the accuracy has doubled. It’s like going from measuring a room with a stretchy rubber band to using a steel tape measure.
Smoother Data: Previously, the warping caused "fake" spikes or dips in the data—like seeing a bump in the road that wasn't actually there. Those "ghost" errors have been wiped away.
True Colors and Brightness: We can now trust the brightness of the light we see much more. This allows us to study the chemistry and temperature of distant galaxies with much higher confidence.

Summary

In short: The scientists didn't change the telescope itself; they changed the mathematical lens through which we view its data. By rewriting the "instruction manuals" to match a new, more accurate map, they have given astronomers a much clearer, more honest view of the deep universe.

Technical Summary: Instrument Science Report COS 2026-01(v1)

Problem: Coordinate System Shift in COS Calibration

The Cosmic Origins Spectrograph (COS) utilizes the CalCOS data processing pipeline, which relies on a hierarchy of reference files to calibrate photon events. In 2024, a major update was implemented to improve the far-ultraviolet (FUV) geometric distortion, delta-geometric, and walk corrections.

Because these corrections occur at the very beginning of the CalCOS workflow, they shift photon events into a corrected coordinate system ($(XCORR, YCORR)$). This shift effectively changed the "map" of the detector. Consequently, all "downstream" reference files—those used for wavelength assignment, spectral extraction, and flux calibration—became misaligned with the new coordinate frame. To maintain calibration accuracy, every downstream file had to be re-derived using data that had already undergone the new geometric corrections.

Methodology: Iterative Re-derivation Workflow

The researchers undertook a massive effort to re-derive 56 reference files applicable to Lifetime Positions (LP) 1 through 6. The methodology was structured by file category and detector position:

Whole-Detector Files: Files affecting the entire detector (e.g., BRFTAB for detector edges, BPIXTAB for bad pixels, GSAGTAB for gain-sagged pixels, and SPOTTAB for hot spots) were updated first. Notably, a new automated "Laplacian of Gaussian" method was implemented to detect and track hot spots more reliably than previous manual methods.
Lifetime Position (LP) Specific Files: The team worked backward from LP6 to LP1. For each LP, they used three types of data:
- White Dwarf (WD) calibrators: To define extraction regions and derive flux calibration.
- Emission/Absorption line stars: To derive dispersion (wavelength) solutions.
- Pt-Ne lamp observations: To tie dispersion solutions to a reference point.
Iterative Extraction Process: For LPs 3–6, the team employed a sophisticated iterative loop between BOXCAR and TWOZONE extraction methods. They would extract spectra, create intermediate flat-field (FLATFILE) and flux (FLUXTAB) files, and then re-process the data until the differences in the 1D profiles converged to $\leq 0.01\%$ .
Time-Dependent Sensitivity: The TDSTAB file (which corrects for sensitivity decline over time) was re-derived by analyzing the entire history of COS operations, consolidating previous files into a single version that includes dependence on the central wavelength (cenwave).

Key Contributions

Comprehensive Re-calibration: The successful re-derivation and delivery of 56 critical reference files to the HST Calibration Reference Data System.
Improved Detection Algorithms: Implementation of an automated, statistically driven method for hot spot detection in the SPOTTAB.
Standardization of Blue Modes: The retirement of specialized, mode-specific gain-sag files for the "blue modes" (G130M/1055 and 1096) in favor of the more robust TWOZONE extraction method.
Unified Time-Dependence: Consolidation of the time-dependent sensitivity corrections into a single, more accurate model.

Results and Validation

The new reference files were validated by comparing newly produced data products against those created with the old files. The results showed significant improvements in calibration precision:

Wavelength Accuracy: Improved from $\sim 0.5$ resolution elements (3 pixels) to $\sim 0.25$ resolution elements (1.5 pixels).
Systematic Velocity Errors: The maximum systematic offset between different detector locations was reduced from $\sim 20\text{ km s}^{-1}$ to $\sim 10\text{ km s}^{-1}$ for medium-resolution modes.
Flux Stability: Artificial flux spikes and deficits near detector edges were eliminated, and absolute flux calibration errors at the edges were brought well within the required $5\%$ threshold.

Significance

This work represents a fundamental "reset" of the COS calibration framework. By aligning all downstream calibration products with the improved geometric and walk corrections, the COS team has significantly enhanced the scientific utility of the instrument. The improvements in wavelength and flux precision allow for more accurate astrophysical measurements, particularly for studies involving high-resolution spectroscopy and faint ultraviolet sources where edge-of-detector effects were previously problematic.