Longitudinal Variation of H2O Ice Absorption on Miranda

Near-infrared reflectance spectra of the Uranian satellite Miranda.

Title: Longitudinal Variation of H2O Ice Absorption on Miranda
Authors: Riley DeColibus, Nancy Chanover, Richard Cartwright

Description of contents:
This .tar package is a collection of reduced near-IR (~0.9 to 2.5 um) disk-integrated reflectance spectra of the Uranian moon Miranda.
The package includes:

• 23 spectra acquired with the TripleSpec spectrograph on the ARC 3.5m telescope at Apache Point Observatory.
• 4 spectra acquired with the GNIRS spectrograph on the 8.1m Gemini North telescope on Maunakea.
• 10 spectra acquired with the SpeX spectrograph on the 3m IRTF telescope on Maunakea.
• 5 "grand average" spectra constructed by robustly averaging all TripleSpec data frames acquired covering certain longitude ranges on Miranda. The SpeX and GNIRS data are not included in these grand averages.
• 1 "grand average" TripleSpec spectrum of Uranus, averaged using observations on five nights in fall 2020.

The filenames are in the following format:

Mir_[mean sub-observer longitude]_[UT date]_[instrument].txt

The files themselves are plain-text, fixed-width ASCII tables.
The first several lines of the file header are commented with "#" characters.
An example header and explanation is discussed below in the "Additional comments" section.
One uncommented line with column titles precedes the data itself.

The 'wl' column is wavelength in micrometers (um).
The 'flux' column is an arbitrary flux ratio, normalized to unity at 1.72 um.
The 'err' column is the 1-sigma errors on the 'flux' column.

The process of data acquisition, reduction, and analysis for the TripleSpec and GNIRS spectra are described in the text of the main paper.
The SpeX spectra consist of 7 spectra published by previous authors (Gourgeot et al. 2014, Cartwright et al. 2018), and
3 spectra acquired in 2017 that were not previously published, but were reduced via the same procedures as in Cartwright et al. 2018.

These data have been corrected for telluric (atmospheric) absorption and divided by the spectrum of an early G dwarf standard star.
We did not attempt to correct differing spectral slopes between different standard stars and the Sun.
While we only included one 'grand average' Uranus spectrum as an example, we obtained a spectrum of Uranus on every night that we observed Miranda.
Users interested in the individual spectra of Uranus from each date should contact the corresponding author.

System requirements:
No special programs should be required to read these spectra.
They were converted to ASCII tables with the Astropy table reading/writing functionality and should be inherently compatible with it.
Note that some spectra do include some invalid values represented with 'nan', and these values should be ignored.

Additional comments:

The commented lines in the file headers include useful metadata regarding observational details for each spectrum.
The main text includes more details on how these quantities were determined.

Example header and explanation:
# PI = DeColibus             [PI of the observing proposal]
# INSTRUMENT = TripleSpec       [Instrument/instrument mode, if applicable]
# UT_DATE = UT201206            [Date of observation in UTYYMMDD format]
# BODY = Miranda                [Observed astronomical body]
# MEAN_LAT = 47.66              [Mean sub-observer latitude]
# MEAN_LONG = 274.6             [Mean sub-observer longitude (IAU east-positive)]
# START_LONG = 241.5            [Sub-obs longitude of first exposure]
# END_LONG = 304.7              [Sub-obs longitude of last exposure]
# LONG_RANGE = 63.2             [Degrees of longitude spanned by spectrum]
# MEAN_PHASE = 1.72             [Mean phase angle (alpha)]
# TOT_EXPTIME = 12960           [Total exposure time (seconds)]
# N_FRAMES = 72                 [Number of frames contributing to spectrum]
# STANDARD = HD 16275           [Star(s) used as telluric standards, typically early G dwarfs]
# WAVELENGTH UNITS = micrometers (um)
# FLUX UNITS = arbitrary (normalized to unity at 1.72 um)
# ERROR UNITS = 1-sigma errors on the above
           wl            flux           err     
     0.910701      -1.165e+00      9.00e+00     
     0.910824      -9.227e-01      7.27e+00     
     0.910946      -3.486e+00      8.03e+00     
          ...           ...           ...

The spectra are presented at native resolution and instrumental sampling.
The TripleSpec spectrograph has a detector gap between 1.85-1.88 um, and there is no data in this region.

As discussed in the main text, the TripleSpec spectra suffer from a "bump" in apparent flux between about 0.95-1.15 micron.
This 'bump' is not a real feature in Miranda's spectrum, but results from spurious spectral slopes in different spectral orders.
In regions of the spectrum where the wavelength coverage of these spectral orders overlap,
merging the spectral orders together results in inflection points that create the resulting 'bump'.

Due to its proximity to Uranus, the raw spectra of Miranda are strongly contaminated by scattered light.
This contamination (see figure 15 in main text) is strongest between 0.90-0.965 um, 1.02-1.12 um, 1.24-1.32 um, and 1.48-1.63 um.
A correction procedure was applied to the TripleSpec spectra to remove this Uranus contamination, described in Appendix A of the paper.

However, there may be low-level, residual spectral features resulting from inadequate correction, which may appear as either 'bumps' or 'dips'.
We warn the user to be cautious during analysis, as we cannot guarantee that either the spectral slopes
or any apparent spectral features at wavelengths shorter than 1.4um are reliable.
To a lesser extent, this may also be true of apparent spectral variations <1.4um in the GNIRS and SpeX spectra.

For the raw data files, files from intermediate processing steps, more Uranus spectra, or other questions,
please contact the corresponding author:
Riley DeColibus

(contact information omitted, but is included in readme file in download)