As part of DR20, we are releasing 169 LVM tiles, with a total of ~300,000 spectra across a broad sample of targets, demonstrating the exquisite data provided by LVM-I and giving the community a sampling of science-ready data. All data are reduced with the current version (v1.2.0) of the LVM Data Reduction Pipeline (DRP), and spectral fitting is performed using the LVM Data Analysis Pipeline (DAP).
Quality flags are provided for each exposure (see LVM Data Access), enabling users to apply additional filtering appropriate to their science requirements. These quality flags provide a quantitative assessment of the accuracy of relative and absolute flux calibration, flat-fielding, wavelength calibration, sky subtraction, and the quality of the DAP fit of emission lines. They also encode information about observational conditions, as well as results from visual inspections of the released data products. We strongly recommend consulting these flags before performing any quantitative analysis of the delivered data products.
Due to ongoing developments and testing of the pipelines, there are also a number of general known issues and caveats in the released DR20 data. These are reflected in the quality flags derived from visual inspection and, in some cases, from quantitative metrics. We include a summary of the key known issues below.
Known Data Issues
- One of the most common issues is a jump in absolute flux between the different spectral branches, particularly prominent between the blue and red spectral channels. This issue affects the stellar continuum spectra and is taken into account correctly by the DAP. The user should be careful when applying their own analysis codes to the reduced spectra.
- Another common issue present in many exposures is the over- or under-subtracted brightest night-sky airglow emission lines at [OI] 5577, 6300, 6364 Å. In particular, this significantly affects all measurements of the [OI] nebular emission lines for Galactic sources, as the redshift is not sufficient to separate them from the airglow lines. These lines should not be used for the analysis without bespoke reduction and fitting procedures.
- Small differences in absolute flux between adjacent tiles are sometimes seen outside of bright regions. These deviations from the median flux distribution are likely residuals of the night-sky continuum subtraction, and often correlate with the observation conditions.
- Several exposures exhibit systematic differences in fluxes between fibers from different spectrographs, which result in a prominent 3-sector pattern within the hexagonal LVM field. Such differences likely result from imperfect flat-field correction for some exposures, and are typically present only in faint areas. One exception is the Flame nebula in Orion, hosting one of the brightest stars in the sky, leading to the saturation across many fibers in one of the spectrographs (the nebula itself is not affected).
- For a few tiles, corresponding to relatively low surface-brightness areas, we noticed a non-physical line ratio of [NII]6584 Å / [NII] 6548 Å , with a value significantly greater than 3. This issue seems to result from the suppression of [NII]6548 Å and does not affect other lines (including the brighter [NII] 6584 Å line typically used in the science analysis).
- We note that the telluric line correction is not applied in the current DRP version, and therefore, the telluric absorption bands can be visible in the red part of the stellar spectra of all released exposures. The telluric correction is currently being tested and will be implemented in future DRP versions.