Recalibration of Pre-SM4 STIS Echelle Throughputs

This paper presents a straightforward scaling method, derived from updated CALSPEC stellar models, to recalibrate pre-SM4 STIS echelle throughputs for eight modes, achieving typical flux accuracy improvements of 0.5–2.4% in the FUV and NUV.

The Gist

Here is an explanation of the paper, translated into everyday language with some creative analogies.

The Big Picture: Fixing the "Volume Knob" on an Old Telescope

Imagine the Hubble Space Telescope as a giant, incredibly precise camera that has been taking pictures of the universe for decades. One of its most important tools is the STIS (Space Telescope Imaging Spectrograph). Think of STIS not just as a camera, but as a prism that breaks starlight into a rainbow, allowing astronomers to measure exactly how much light (flux) is coming from a star at every specific color.

For this to work, the telescope needs a "volume knob" calibration. If the knob is set wrong, the telescope might think a star is brighter or dimmer than it actually is.

The Problem:
In 2009, astronauts fixed the telescope (a mission called SM4). After that, the team updated the "volume knob" settings based on the best science they had at the time. However, in 2024, scientists realized their "standard stars" (the reference stars used to calibrate the telescope) had been re-measured with better technology.

It turns out the old reference stars were actually 1% to 3% brighter than we thought. Because the reference was wrong, the telescope's "volume knob" was slightly off for all the data taken before the 2009 repair.

The Challenge: The "Moving Target" Issue

The paper focuses on fixing the data from before 2009 (Pre-SM4). This is tricky because of how the telescope worked back then.

• The Analogy: Imagine taking a photo of a painting, but every month you shift the camera slightly up, down, left, or right to avoid getting dust spots on the lens.
• The Reality: Before 2009, the STIS instrument did exactly this. Every month, it shifted its position slightly to spread the light evenly across the detector. This means the "dust spots" (or in this case, tiny imperfections in the detector) were different for every observation.

Because the camera was moving, doing a full, perfect recalibration for every single old photo is like trying to fix a moving target while blindfolded. It's incredibly difficult and requires a massive amount of new data that doesn't exist in the archives.

The Solution: A "Simple Scaling" Trick

Instead of trying to rebuild the entire calibration from scratch (which is the "hard way" used for post-2009 data), the team came up with a clever, simpler shortcut.

The Analogy:
Imagine you have a recipe for a cake that calls for 1 cup of sugar. You realize later that your measuring cup was actually 10% too small, so you were only putting in 0.9 cups of sugar.

• The Hard Way: Go back to the kitchen, bake a new cake, taste it, measure the sugar again, and rewrite the whole recipe from scratch.
• The "Simple Scaling" Way: Just take the old cake, sprinkle a little extra sugar on top to make up the difference, and serve it. It's not a perfect rewrite, but it tastes much better and takes 5 minutes instead of 5 hours.

How they did it:

1. They compared the Old Reference Model (CALSPECv04) with the New Reference Model (CALSPECv11).
2. They calculated the ratio: How much brighter is the new model compared to the old one? (It's about 1–3% brighter).
3. They applied this ratio as a multiplier to all the old data. If the old data said a star was 100 units bright, and the new model says the reference is 2% brighter, they simply adjusted the old data to account for that 2% difference.

They didn't change the complex "shape" of the data; they just turned the volume up slightly to match the new reality.

The Results: A Small Fix with Big Payoff

The team tested this "simple scaling" method on 8 different modes of the telescope (covering different colors of light, from ultraviolet to near-ultraviolet).

• The Outcome: By applying this simple math trick, they improved the accuracy of the old data by 0.5% to 2.4%.
• Why it matters: In astronomy, being off by 2% can change your understanding of how a star is evolving or how heavy a black hole is. Getting this right ensures that scientists looking at data from 20 years ago are still drawing the correct conclusions today.

The Catch (The "Moving Target" Warning)

The authors add a small warning. Because the telescope was shifting positions (the "monthly offsets") back in the day, the correction works best when the telescope was in its "center" position. When the telescope was shifted far to the side, the correction is still better than before, but it's not perfectly perfect.

However, since they don't have enough old data to study every single shift position, they decided this "good enough" fix is the best option available. It's better to have a slightly imperfect correction than no correction at all.

Summary

The paper is about updating the "volume knob" on Hubble's old photos. Instead of doing the impossible task of re-calibrating every single old photo from scratch, the team used a smart, simple math trick to adjust the brightness levels. This ensures that the scientific data from the early 2000s remains accurate and useful for future discoveries.

Technical Summary

Here is a detailed technical summary of Instrument Science Report (ISR) STIS 2024-02, titled "Recalibration of Pre-SM4 STIS Echelle Throughputs."

1. Problem Statement

The Space Telescope Imaging Spectrograph (STIS) on the Hubble Space Telescope requires periodic flux recalibration to maintain accuracy. This need arises from updates to the CALSPEC library of stellar atmospheric models used as standard stars (specifically the white dwarf G191-B2B).

• The Discrepancy: The transition from older models (CALSPECv04–v07) to the latest CALSPECv11 models revealed continuum flux differences of 1–3% in the Far-Ultraviolet (FUV) and Near-Ultraviolet (NUV) ranges.
• The Challenge: While the STIS team successfully recalibrated post-Servicing Mission 4 (SM4) data (post-2009) by deriving new sensitivity curves and blaze shift coefficients, this full recalibration is not feasible for pre-SM4 data (pre-August 2002).
  • Pre-SM4 observations utilized a "monthly offsetting" strategy to distribute illumination across the detector. This results in complex spatial and spectral variations that make a full, order-by-order recalibration using current archive data extremely difficult and labor-intensive.
  • Existing pre-SM4 throughputs were derived using CALSPECv04, leading to systematic flux errors when compared to modern standards.

2. Methodology

The authors propose a simplified throughput scaling approach as an alternative to full recalibration for pre-SM4 data.

• Reference Models: The method utilizes the ratio between the new CALSPECv11 model and the old CALSPECv04 model for the standard star G191-B2B.
• Continuum Fitting (Spline Method):
  • To avoid imprinting narrow metal-line absorption features (which are prominent in CALSPECv11 blueward of 2000Å) onto the scaling factor, the authors developed an automated procedure to identify continuum points.
  • They smooth the model flux with an 11-pixel boxcar filter and identify points with fractional deviations $< 0.001\%$.
  • A fourth-order spline is fitted to these continuum points. Knots are spaced $>12.5$Å apart and avoid the Lyman-$\alpha$ feature (1180–1240Å).
  • This generates a smooth, wavelength-dependent ratio curve ($CALSPECv11 / CALSPECv04$).
• Scaling Implementation:
  • The derived ratio is applied as a multiplicative scaling factor to the existing PHOTTAB reference files (which contain the throughput efficiencies) for 8 high-priority echelle modes.
  • Crucially, the previously derived blaze-shift coefficients are not modified; only the throughput magnitude is adjusted.
  • Since the new model continuum is brighter, the effective throughput must be reduced to compensate, ensuring the final calibrated flux matches the new standard.

3. Key Contributions

• Development of a Robust Scaling Algorithm: A systematic method to extract smooth continuum ratios from complex stellar models containing dense absorption forests, preventing artifacts in the calibration.
• Pre-SM4 Throughput Updates: Application of this scaling to 8 specific echelle modes (E140H/1271, E140H/1416, E140M/1425, E230H/2263, E230H/2713, E230M/1978, E230M/2415, E230M/2707).
• Validation Against Post-SM4 Standards: The study validates the scaling method by comparing it against the rigorous, full-recalibration results obtained for post-SM4 data, demonstrating that the simple scaling yields comparable accuracy for pre-SM4 data.
• Assessment of Monthly Offsets: The report investigates how the "monthly offsetting" strategy of pre-SM4 observations impacts the efficacy of this scaling, acknowledging limitations while confirming general improvement.

4. Results

• Flux Accuracy Improvements: The scaling approach results in typical flux calibration improvements of 0.5% to 2.4% across the FUV and NUV wavelength ranges.
  • The median difference from the CALSPECv11 model improved from approximately -2.0% (unscaled) to -0.1% to -0.7% (scaled) for most modes.
  • The maximum difference observed was a 2.4% change at 1738Å.
• Comparison with Post-SM4 Recalibration:
  • For the E230M/1978 mode, the simple scaling method reduced the flux difference from CALSPECv11 from -2.06% to -0.10%.
  • This is comparable to the post-SM4 full recalibration result (-0.22%), indicating the scaling method is highly effective despite its simplicity.
• Impact of Monthly Offsets:
  • Data with larger monthly spatial/spectral offsets showed slightly higher scatter and deviation from the model compared to zero-offset data.
  • However, even for offset data, the scaling consistently reduced the flux error (e.g., improving a -2.73% error to -0.54% for a specific offset setting), confirming the method's robustness.

5. Significance

• Data Consistency: This update ensures that historical pre-SM4 STIS data is consistent with modern CALSPECv11 standards, allowing for accurate cross-epoch comparisons and meta-analyses of ultraviolet spectra.
• Efficiency: The report highlights that full recalibration is too labor-intensive for pre-SM4 data due to the lack of suitable calibration targets and the complexity of monthly offsets. The proposed scaling method offers a low-effort, high-accuracy solution that can be applied to lower-priority modes in the future.
• Operational Impact: The updated PHOTTAB files will be implemented in the CALSTIS pipeline, automatically correcting fluxes for new and re-reduced pre-SM4 observations without requiring user intervention.

In conclusion, this report provides a pragmatic and scientifically rigorous solution to a legacy calibration problem, significantly improving the absolute flux accuracy of a decade's worth of Hubble STIS ultraviolet observations.