Coadded Spectroscopic Stellar Parameters and Abundances from the LAMOST Low Resolution Survey
aa r X i v : . [ a s t r o - ph . S R ] F e b Draft version February 10, 2021
Typeset using L A TEX
RNAAS style in AASTeX63
Coadded Spectroscopic Stellar Parameters and Abundances from the LAMOST Low ResolutionSurvey
Jacob H. Hamer Department of Physics and Astronomy, Johns Hopkins University, 3400 N. Charles Street, Baltimore, MD 21218, USA
ABSTRACTI combine duplicate spectroscopic stellar parameter estimates in the LAMOST Data Release 6 LowResolution Spectral Survey A, F, G, and K Type stellar parameter catalog. Combining repeat mea-surements results in a factor of two improvement in the precision of the spectroscopic stellar parameterestimates. Moreover, this trivializes the process of performing coordinate-based cross-matching withother catalogs. Similarly, I combine duplicate stellar abundance estimates for the Xiang et al. (2019)catalog which was produced using LAMOST Data Release 5 Low Resolution Spectral Survey data.These data have numerous applications in stellar, galactic, and exoplanet astronomy. The catalogs Iproduce are available as machine-readable tables at https://doi.org/10.7281/T1/QISGRU.
Keywords:
Astronomy data analysis — Stellar abundances — Fundamental parameters of stars —Stellar properties — Sky surveys — Astrophysics - Solar and Stellar AstrophysicsThe Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) Data Release 6 (DR6) contains a catalogof spectroscopic stellar parameters for 5,773,552 A, F, G, and K stars generated from the R ∼ tname as a unique identifier for each object in the catalog. However,matching the catalog to itself on-sky reveals matches between objects with different tname as well as tsource , withsimilar spectroscopic stellar parameters. Therefore, tname is not a sufficient identifier to find all parameter estimatesfor the same star. Additionally, I find that a number of objects from the PILOT survey for which a single tname corresponds to multiple objects with similar declinations but drastically different right ascensions. I therefore undertakethe following steps to combine parameter estimates for different observations of the same star. First, I group the catalogby tname and tsource . Then, I check that the positions of the grouped objects do not differ by more than 10 arcsecondsin order to exclude the sources from the PILOT survey that have identical tname but different on-sky positions. Next,I check that the radial velocity estimates are not discrepant, only combining measurements when the significance t = RV max − RV min q σ , max + σ , min (1)is less than 3. Such a discrepancy could indicate that the objects are being incorrectly grouped by our method, orthat the star is in an unresolved binary.For grouped objects which pass the two tests described, I combine their stellar parameter estimates as follows. Icalculate the arithmetic weighted mean of each parameter using the SNR in the SDSS g band as the weight:¯ x = P Ni SNR i x i P Ni SNR i , (2) Corresponding author: Jacob H. [email protected] where x i = { T eff , [Fe / H] , log g, RV } denotes a set of spectroscopic stellar parameters from one observation, and SNRis the signal-to-noise ratio in the SDSS g band. I also produce a right ascension and declination for the combinedobservation with the same equation. It follows from error propagation that the uncertainty on our new parameterestimate is σ ¯ x = qP Ni SNR i σ x,i P Ni SNR i . (3)I estimate the combined SDSS g band SNR of this “coadded” observation by adding the g band SNR of each observationin quadrature. I also report the number of observations which go into the new parameter estimate.At this point, I have combined observations which can be grouped based on their shared tname and tsource , buthave not yet grouped objects which differ in these identifiers but are co-located on the sky. To identify stars whichshould be combined based on their distance on-sky, I match the grouped catalog with itself on the sky using TOPCAT.I use a radius of 0.5 arcseconds, and I retain all matches. I use the GroupID outputted by TOPCAT as a key to identifywhich observations to group, and combine them using the same methods described above, including the check thatthe radial velocities do not have a discrepancy with a significance over 3.The catalog which results from these steps to combine multiple observations is 4,334,538 objects. 22% of these setsof spectroscopic stellar parameters are the result of coadding. 72% are the result of 2 coadded observations and 28%are the result of 3 or more. In addition, I produce a catalog which provides sets of obsid which have I determined tocorrespond to a single object at the listed right ascension and declination. This key allows for one to carry out theirown method of coadding repeat observations of an object.We apply a similar methodology to the catalog of stellar abundances produced by Xiang et al. (2019) using theLAMOST Data Release 5 LRS. I remove any observations which are flagged as poor quality based on the chi-squaredof the spectral fit. Additionally, for each observation I use the individual element quality flags to remove unreliableabundance estimates. As the [ α /Fe] estimate is the weighted mean of [Mg/Fe], [Si/Fe], [Ca/Fe], and [Ti/Fe], I removethis estimate if any of the elements are flagged as unreliable. I also omit unreliable estimates of T eff , log g , andmicroturbulent velocity V t . I combine the reliable parameter estimates to produce a weighted mean and associatederror as above. The resulting catalog contains 5,792,830 objects, 20% of which are the result of coadding multipleobservations. ACKNOWLEDGMENTSThis material is based upon work supported by the National Science Foundation under grant number 2009415.REFERENCES. I combine the reliable parameter estimates to produce a weighted mean and associatederror as above. The resulting catalog contains 5,792,830 objects, 20% of which are the result of coadding multipleobservations. ACKNOWLEDGMENTSThis material is based upon work supported by the National Science Foundation under grant number 2009415.REFERENCES