Hot subdwarfs from the SDSS/BOSS One month in Bamberg Peter Nemeth et al. March 8. 2014
SEGUE and BOSS targets in DR10 1159 spectra of 860 stars, 238 SEGUE + 921 BOSS Average magnitude: g = 17.43 ± 2 90% over ∣b∣ = 30°
Some challenges with SDSS spectra Continuum normalization near the Balmer jump, always a question Spectral resolution, in particular R(λ) is important for a precise analysis Further challenges: SNR flux calibration fiber loss R(λ, fiber) BOSS SDSS Spectral range 3650-10400 3800-9200 Resolution 1400-2600 1800-2000
Spectrostatistics We are not (yet) after the fine details, just look for a reasonable distribution of stars in the T-g and T-He diagrams TLUSTY/SYNSPEC NLTE models in sampling mode XTGRID fits, steepest descent method with statistical error determination Simple analysis assuming: H and He composition vmicro = ξ = 0 ; vmacro = 0 vr sin(i) = 0 ; E(B-V) = 0 (determined from the fit) Done on a single node of the MAIDS cluster 8 cores, x4 GHz ~20 days, 3 hrs/spectrum
GALEX sample (Nemeth et al. 2012, Vennes et al. 2011) 166 bright subdwarfs from the disk Consistent results with the SPY survey (Stroer et al. 2007) Clumpings and trends were observed Abundance extremes at 38000 K sdB: gravity too low Some He-sdO: gravity too high
SDSS sample 1158 spectra Dive and Giraffe diagrams
SDSS sample 860 stars sdB: gravity even lower He-sdO: for some gravity is still high Next steps: Evaluate results Go for CNO models with all the three MAIDS nodes
He-sdO gravity One possible explanation: Magnetic fields + rotation Requires better data and appropriate models
Radial velocity variations All combined spectra have been modeled and shifted to the lab wavelength RV variations are “estimated” for 6038 individual spectra of the 860 stars Monte Carlo method: resampling the original data and do a χ2 minimization Mean velocity from 13 trials Correlates with SNR Eccentric orbits?