WFC3/UVIS Geometric Distortion -- Time Evolution of Linear Terms w.r.t Gaia
This paper analyzes over 7,400 WFC3/UVIS exposures aligned with Gaia DR3 to characterize the time evolution of geometric distortion linear terms from 2009 to 2022, revealing minor temporal changes dominated by intrinsic scatter and filter-dependent offsets, while recommending specific realignment procedures for high-precision astrometry.
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 Camera That Slightly Shifts Over Time
Imagine the Hubble Space Telescope's WFC3/UVIS camera as a very high-end digital camera. To take a perfect picture of the universe, this camera needs a "map" (called a geometric distortion model) that tells it exactly how to translate the pixels on its sensor into real locations in the sky.
For years, scientists have used a specific map to correct the images. But this paper asks a simple question: Has this map stayed accurate over the last 13 years (2009–2022), or has the camera slowly changed its shape or alignment?
To find out, the authors acted like detectives. They took over 7,400 photos taken by Hubble and compared them against the Gaia DR3 catalog, which is essentially the most precise "GPS map" of the stars we have ever made. They checked if the stars in Hubble's photos lined up perfectly with the stars on the Gaia GPS map.
The Four Ways the Camera Can "Drift"
The authors looked at four specific ways the camera's internal geometry might have shifted over time. Think of these as four ways a photo on a table might get messed up:
- Shift (The Slide): Imagine sliding a photo slightly to the left or up.
- What they found: The camera does slide a bit over time. However, this "slide" is mostly caused by the telescope not pointing exactly where it thought it was, rather than the camera itself changing. It's like a photographer's hand shaking slightly, not the camera lens warping.
- Rotation (The Spin): Imagine the photo is slightly tilted, like a picture frame hanging crookedly.
- What they found: From 2009 to 2017, the tilt was very stable. But starting in mid-2017, the tilt became a bit wobbly and unpredictable. The authors suspect this is because the telescope itself started shaking (jittering) more during that time, making the "spin" measurement harder to pin down.
- Scale (The Zoom): Imagine the photo slowly zooming in or out, making stars look slightly closer together or further apart.
- What they found: There is a tiny, slow "zoom" drift over the years (less than 0.2 pixels over 13 years). However, there is a bigger issue: different color filters act like different lenses. A photo taken in ultraviolet light (F275W) has a slightly different "zoom" than one taken in infrared (F814W). This difference is up to 0.3 pixels. It's like having two different cameras where one is slightly zoomed in compared to the other, and the current map doesn't account for this difference.
- Skew (The Slant): Imagine a square photo that gets squashed into a parallelogram, where the corners are no longer perfect 90-degree angles.
- What they found: The camera's grid has become slightly less square over time, but the change is very small (less than 0.2 pixels) and is mostly hidden by natural "noise" or scatter in the data.
The Main Takeaways
- The changes are small, but real: The camera's internal geometry has changed slightly over 13 years, but the changes are tiny (less than a quarter of a pixel).
- The "Noise" is bigger: The natural uncertainty in how the telescope points and how the data is measured is often larger than the actual changes in the camera. It's hard to see the slow drift because the telescope's hand is shaking a bit more than the camera is changing.
- Color matters: The biggest consistent difference isn't time, but color. If you mix images taken with different filters without adjusting them, they won't line up perfectly because of that "zoom" difference mentioned above.
What Should Astronomers Do?
The paper concludes that while the official "map" (the IDCTAB file) used by the Hubble pipeline is good, it isn't perfect for high-precision work, especially when mixing different colors or looking at data from different years.
The Recommendation:
Astronomers should use a tool called tweakreg (part of the DrizzlePac software) to manually "re-align" their images before doing science.
- Think of this as taking the photos and manually sliding, rotating, and resizing them until they perfectly match the Gaia GPS map.
- The authors suggest using a "6-parameter fit" (which adjusts for shift, rotation, and scale separately for X and Y axes) to get the best possible alignment.
Summary
The Hubble camera is aging gracefully, but its internal geometry has shifted just enough over 13 years that we can't rely on the old maps alone. The telescope also shakes a bit, and different color filters act like slightly different zoom lenses. To get the sharpest, most accurate scientific results, astronomers need to manually "re-tune" their images using modern software to match the precise Gaia star map.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.