Using 23 Years of ACS/SBC Data to Understand Backgrounds: Significant Reductions in Expected Background Levels

TL;DR

This work analyzes 23 years of HST ACS/SBC imaging to empirically quantify background levels and their variability, highlighting that airglow drives large fluctuations in several SBC filters while dark current dominates only in a couple of filters. The authors develop a robust masking-based pipeline to measure backgrounds on 8,640 suitable exposures, revealing substantial discrepancies with ETC v33.2 predictions, with measured backgrounds significantly lower in airglow-dominated filters due to weaker OI lines than assumed. They perform a detailed comparison, explore a partial reconciliation via line-scaling, and recommend adopting empirical background percentiles in ETC v34.1 for more accurate exposure-time planning. The findings imply higher practical sensitivity for short-wavelength SBC observations and provide practical guidance for users on background expectations, bright-object limits, and background prediction approaches for SBC planning.

Abstract

We have used 23 years of Hubble Space Telescope ACS/SBC data to study what background levels are encountered in practice and how much they vary. The backgrounds vary considerably, with F115LP, F122M, F125LP, PR110L, and PR130L all showing over an order of magnitude of variation in background between observations, apparently due to changes in airglow. The F150LP and F165LP filters, which are dominated by dark rate, not airglow, exhibit a far smaller variation in backgrounds. For the filters where the background is generally dominated by airglow, the backgrounds measured from the data are significantly lower than what the ETC predicts (as of ETC v33.2). The ETC predictions for `average' airglow are greater than the median of our measured background values by factors of 2.51, 2.64, 105, and 3.64, for F115LP, F122M, F125LP, and F140LP, respectively. A preliminary analysis suggests this could be due to certain OI airglow lines usually being fainter than expected by the ETC. With reduced reduced background levels, the shorter-wavelength SBC filters can conduct background-limited observations much more rapidly than had previously been expected. As of ETC v34.1, a new option will be included for SBC calculations, allowing users to employ empirical background percentiles to estimate required exposure times.

Figures (7)

• Figure 1: Illustration of the steps of our masking procedure, using exposure JEC0C6AEQ, an F150LP observation of an area within nearby spiral galaxy NGC 7793. Full descriptions of each step in the process are given in Section \ref{['Section:Background_Measurement']}. Note that panel I, the final masked rate image, uses a different color scale than the other versions of the rate image -- panels A, B, and D -- which all share a color scale. Panel I uses a color scale tailored to the low count rates in the un-masked areas of the image, with dark blue used for pixels with a rate of 0 counts s$^{-1}$, and black reserved for masked regions. Panel I is also shown at twice the size of the other panels, to allow closer inspection. Panels F, G, and I show the boundaries between tiles, and the color of the masked pixels in panels F and G indicates which tile the masked pixels belong to; see Section \ref{['Subsection:Tile_Masking']} for details.
• Figure 2: Same as for Figure \ref{['Fig:Masking_Example_1']}, except for exposure JE2054ZIG, an F165LP observation of globular cluster NGC 104, which has been positioned in the low-dark region in the lower-left of the detector. Dark current is prominent in this exposure, especially in the high-dark region in the center and upper-right of the detector.
• Figure 3: Same as for Figure \ref{['Fig:Masking_Example_1']}, except for exposure JEDI12XPQ, an F165LP observation of star-forming dwarf galaxy SDSS J1044+0353.
• Figure 4: Same as for Figure \ref{['Fig:Masking_Example_1']}, except for exposure JEG201A4Q, an F115LP observation of open cluster NGC 6681 (a standard SBC calibration target).
• Figure 5: Distributions of backgrounds measured in each filter. Median value is plotted for each, and 16th--84th percentile ranges are shaded. Also shown for comparison are predicted ETC backgrounds for 'low', 'average', and 'high' airglow (zodiacal light and Earthshine being negligible). The standard assumed dark current rate (as of ETC v33.2) of 0.0082 ${\rm counts\,sec^{-1}\,arcsec^{-2}}$ is also plotted, although this can vary over the detector. Prism measurements included for completeness; for instance, note that the background distribution for PR130L is very similar to that of F125LP, which has a very similar transmission function.