🔭 astrophysics

New Time-Dependent WFC3/IR Inverse Sensitivities

This paper presents new time-dependent inverse sensitivities for the WFC3/IR channel, derived from CALSPEC standard photometry and sensitivity slopes, which correct for 1–2% signal loss over 15 years to achieve sub-0.5% photometric precision and will be implemented via an updated \texttt{calwf3} package and MAST reprocessing in late 2024.

Original authors: Annalisa Calamida, Mariarosa Marinelli, Varun Bajaj, Aidan Pidgeon, Jennifer Mack

Published 2026-02-16

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Annalisa Calamida, Mariarosa Marinelli, Varun Bajaj, Aidan Pidgeon, Jennifer Mack

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 Hubble Space Telescope's WFC3/IR camera as a very sensitive, high-end digital camera that has been taking pictures of the universe for over 15 years. Like any camera, its "eye" (the detector) has been slowly aging. It's getting a little bit dimmer and less sensitive to light over time, much like how an old pair of sunglasses might get slightly scratched or how a battery slowly loses its charge.

For a long time, scientists treated this camera as if it were brand new every time they looked at a photo. They assumed that if a star looked a certain brightness in 2010, it should look exactly the same in 2020, assuming the star itself didn't change. But because the camera's "eye" was actually getting dimmer, the photos were becoming slightly darker than they should have been, leading to small errors in measuring how bright stars and galaxies actually are.

The Problem: The Fading Lens
The authors of this report discovered that the camera's sensitivity has dropped by about 1% to 2% since it was installed in 2009. It's not a uniform drop, either. It's like the camera is wearing different colored glasses for different colors of light: it's losing sensitivity faster for blue light than for red light.

If you didn't account for this, you might think a star is fading away when it's actually just the camera getting tired.

The Solution: A New "Time-Travel" Calculator
To fix this, the team created a new set of rules, called time-dependent inverse sensitivities. Think of these as a magical calculator that knows exactly how old the camera was on the day a specific photo was taken.

Here is how they did it:

The Reference Stars: They picked five "standard candles" (stars with known, unchanging brightness) that Hubble has been photographing regularly since 2009.
The Comparison: They compared what the camera actually saw over the years against what the stars should have looked like based on perfect computer models.
The Correction: They realized the camera was under-reporting the light. So, they built a new mathematical "lens" that adjusts the data. If a photo was taken 10 years ago, the calculator adds a tiny bit of brightness back. If it was taken yesterday, it adds a different, smaller amount.

The Result: Sharper, More Accurate Photos
Before this update, if you measured the brightness of a star over 15 years, your measurements would wobble by about 0.6%. It was like trying to weigh yourself on a scale that drifted slightly every day.

With this new update, the "wobble" is reduced to less than 0.5%. The scale is now perfectly calibrated for every single day of the camera's life.

What This Means for You (The User)

For the Archive: The Hubble data archive (MAST) is re-processing all the old photos from late 2024 onwards. When you download an image, the "header" (the metadata tag on the file) will now automatically say, "This photo was taken on [Date], so we adjusted the brightness by [Amount] to account for the camera's age."
For New Observations: If you are planning to use the software stsynphot to simulate what a star should look like, you can now tell the software, "Simulate this for the year 2024," and it will automatically apply the correct "aging" correction to the numbers.

The Bottom Line
This report is essentially a "maintenance manual" update for Hubble's infrared camera. By acknowledging that the camera gets older and adjusting the math accordingly, scientists can now measure the universe with even greater precision, ensuring that when they say a star is getting brighter or dimmer, it's the star changing, not the camera.

1. Problem Statement

The Wide Field Camera 3 (WFC3) on the Hubble Space Telescope (HST) has an Infrared (IR) channel that has been in operation since 2009. Over time, the detector's sensitivity has degraded due to aging and radiation damage.

Previous Limitation: The inverse sensitivities (PHOTFLAM) used to convert count rates to physical flux units were last updated in 2020 (calwf3 v3.5.2). These values were static, calculated for a single reference epoch (MJD 55008, June 26, 2009).
The Issue: The sensitivity of the WFC3/IR detector has declined by approximately 1–2% over 15 years, with the loss being wavelength-dependent (greatest in bluer filters). While users could manually apply correction slopes derived in Marinelli et al. (2024), the official calibration pipeline lacked a built-in, time-dependent mechanism. This resulted in a photometric precision limit of roughly 0.6% and introduced systematic errors for observations taken far from the reference epoch.

2. Methodology

The authors developed a new set of time-dependent inverse sensitivities and filter throughputs using the following approach:

Data Collection: Photometry was collected from five CALSPEC standard stars (GD153, GD71, G191B2B, GRW+70 5824, and the G-type star P330E) spanning the period 2009–2023.
Sensitivity Correction: Observed count rates were corrected for time-dependent sensitivity losses using the slopes measured in Marinelli et al. (2024), which are functions of the filter pivot wavelength.
Synthetic Photometry:
- Synthetic count rates were generated using stsynphot.
- Inputs included the latest CALSPEC Spectral Energy Distributions (SEDs), updated filter curves (2020), and specific correction files (aperture, in-flight quantum efficiency).
- The ratio of observed (corrected) to synthetic count rates was calculated for each filter.
Filter Curve Derivation: Multiplicative scalar factors were derived from the observed-to-synthetic ratios to update the filter throughput curves. New time-dependent filter curves were created to allow interpolation/extrapolation for any observation date.
Inverse Sensitivity Calculation: New inverse sensitivities ( $S' = C/F$ ) were calculated for the reference epoch (MJD 55008) and for specific observation dates using the updated filter curves and corrected count rates.
Pipeline Integration: A new IMPHTTAB (Image Photometry Lookup Table) reference file was created for the calwf3 v3.7.2 pipeline. This allows the pipeline to automatically populate the PHOTFLAM and PHOTFNU header keywords with values specific to the observation's epoch (EXPSTART).

3. Key Contributions

Time-Dependent Calibration: The report delivers the first fully time-dependent inverse sensitivities for WFC3/IR, moving away from static reference values.
Updated Reference Files:
- New filter throughput curves (version 008) for all 15 WFC3/IR filters.
- A new IMPHTTAB reference file (8bq13281i_imp.fits) compatible with calwf3 v3.7.2.
Extended Data Baseline: The calibration now incorporates 14 years of data (2009–2023), compared to the 10 years used in the 2020 calibration.
Updated Standard Models: Utilization of the most recent CALSPEC SEDs, including new CTE corrections for STIS data, updated BOSZ models, and a new extinction law.

4. Results

Precision Improvement: The new calibration achieves an internal photometric precision of better than 0.5% for all wide-, medium-, and narrow-band filters, an improvement over the previous 0.6% precision.
Validation with Standard Stars:
- Testing on the standard star GD153 showed that applying time-dependent corrections reduced the dispersion of observed-to-synthetic flux ratios.
- The mean observed-to-synthetic ratio for the new calibration is 1.003 ± 0.003, indicating high accuracy.
Validation with Sgr A Field:*
- Photometry of stars in the crowded Galactic Center region (Sgr A*) was analyzed over a ~14-year baseline.
- Before Correction: Using constant (2020) PHOTFLAM values resulted in a spurious magnitude trend (slope) of -0.08% to -0.11% per year, mimicking sensitivity loss.
- After Correction: Using the new time-dependent PHOTFLAM values flattened this trend to -0.02% to 0.02% per year, effectively removing the systematic drift.
Filter Curve Accuracy: The updated filter curves (008) corrected previous underestimations in throughput, bringing the observed-to-synthetic ratios closer to unity (RMS of 0.5%).

5. Significance and Recommendations

Impact on Data: The pixel values of WFC3/IR images remain unchanged. The update affects only the header keywords (PHOTFLAM, PHOTFNU) in the FLT (flat-fielded) images.
User Action Required:
- Reprocessing: Users with data downloaded from MAST prior to late 2024 must either re-download the data (after MAST reprocessing) or re-process their RAW data using calwf3 v3.7.2 or higher to obtain the correct time-dependent header values.
- Simulations: Users performing simulations with stsynphot must use the new time-dependent filter curves and include the mjd# keyword in the observation mode string (e.g., obsmode='wfc3,ir,f110w,mjd#57429') to account for sensitivity loss at specific dates.
Scientific Value: This update is critical for long-term monitoring programs and high-precision photometry (e.g., exoplanet transits, variable stars) where sub-percent accuracy is required. It ensures that flux measurements are consistent regardless of when the observation was taken relative to the detector's degradation.

Conclusion:
This report marks a significant maturation of the WFC3/IR calibration pipeline. By integrating time-dependent sensitivity corrections directly into the standard pipeline and reference files, the WFC3 team has eliminated a major source of systematic error, enabling photometric precision of <0.5% across the instrument's 15-year operational history.

1. Problem Statement

2. Methodology

3. Key Contributions

4. Results

5. Significance and Recommendations

More like this