SpectroTranslator: a deep-neural network algorithm to homogenize spectroscopic parameters
arxiv(2024)
摘要
The emergence of large spectroscopic surveys requires homogenising on the
same scale the quantities they measure in order to increase their scientific
legacy. We developed the SpectroTranslator, a data-driven deep neural network
algorithm that can convert spectroscopic parameters from the base of one survey
to another. The algorithm also includes a method to estimate the importance
that the various parameters play in the conversion from base A to B. As a
showcase, we apply the algorithm to transform effective temperature, surface
gravity, metallicity, [Mg/Fe] and los velocity from the base of GALAH into the
APOGEE base. We demonstrate the efficiency of the SpectroTranslator algorithm
to translate the spectroscopic parameters from one base to another using
parameters directly by the survey teams, and are able to achieve a similar
performance than previous works that have performed a similar type of
conversion but using the full spectrum rather than the spectroscopic
parameters, allowing to reduce the computational time, and to use the output of
pipelines optimized for each survey. By combining the transformed GALAH
catalogue with the APOGEE catalogue, we study the distribution of [Fe/H] and
[Mg/Fe] across the Galaxy, and we find that the median distribution of both
quantities present a vertical asymmetry at large radii. We attribute it to the
recent perturbations generated by the passage of a dwarf galaxy across the disc
or by the infall of the Large Magellanic Cloud. Although several aspects still
need to be refined, in particular how to deal in an optimal manner with regions
of the parameter space meagrely populated by stars in the training sample, the
SpectroTranslator already shows its capability and promises to play a crucial
role in standardizing various spectroscopic surveys onto a unified basis.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要