TLDR:
1) Lots of chemical info to exploit in the opticalโof interest for ๐ช host chemistry.
2) Beware of naively trusting M/K dwarf physical model spectraโthey aren't (currently) a good match to reality & this affects recovered stellar properties.
3) I've helpfully quantified some of this mismatch!
23.02.2024 16:02 โ ๐ 7 ๐ 0 ๐ฌ 0 ๐ 0
Observed low resolution blue spectra of benchmark M/K dwarfs ordered from warmest (top, smoothest) to coolest (bottom, most wiggly) as compared to a data-driven/machine learning (ML) model. The spectra match remarkably well!
Observed medium resolution red spectra of benchmark M/K dwarfs ordered from warmest (top, smoothest) to coolest (bottom, most wiggly) as compared to a data-driven/machine learning (ML) model. The spectra match remarkably well!
Observed blue+red spectra of benchmark M/K dwarfs ordered from warmest (top, smoothest) to coolest (bottom, most wiggly) as compared to theoretical spectra from physical models. The match is poor, pointing to missing physics or molecular data in the physical models.
<skeets into room at full speed>
Hi astro blueskyโit's paper day!
I dove into optical M/K dwarf spectraโrife with molecules as they areโ& found success with a data-driven ๐ model (i.e. ML) vs physical models.
1) ๐ฆ๐ recovery ๐
2) ๐ฅ๐ recovery ๐
3) Physical model vs ML ๐
arxiv.org/abs/2402.14639 ๐ญ
23.02.2024 15:56 โ ๐ 11 ๐ 1 ๐ฌ 1 ๐ 0
