J/A+A/687/A205ivo://CDS.VizieR/J/A+A/687/A205doi:10.26093/cds/vizier.36870205CDSMas-Buitrago P.Gonzalez-Marcos A.Solano E.Passegger V.M.,Cortes-Contreras M.Ordieres-Mere J.Bello-Garcia A.Caballero J.A.,Schweitzer A.Tabernero H.M.Montes D.Cifuentes C.2024-11-21T20:02:03Z2024-07-15T08:37:52ZCDS support team

CDS, Observatoire de Strasbourg, 11 rue de l'Universite, F-67000 Strasbourg, France

cds-question@unistra.frm-starsdwarf-starsspectroscopyDeep learning (DL) techniques are a promising approach among the set of methods used in the ever-challenging determination of stellar parameters in M dwarfs. In this context, transfer learning could play an important role in mitigating uncertainties in the results due to the synthetic gap (i.e. difference in feature distributions between observed and synthetic data). We propose a feature-based deep transfer learning (DTL) approach based on autoencoders to determine stellar parameters from high-resolution spectra. Using this methodology, we provide new estimations for the effective temperature, surface gravity, metallicity, and projected rotational velocity for 286 M dwarfs observed by the CARMENES survey. Using autoencoder architectures, we projected synthetic PHOENIX-ACES spectra and observed CARMENES spectra onto a new feature space of lower dimensionality in which the differences between the two domains are reduced. We used this low-dimensional new feature space as input for a convolutional neural network to obtain the stellar parameter determinations. We performed an extensive analysis of our estimated stellar parameters, ranging from 3050 to 4300K, 4.7 to 5.1dex, and -0.53 to 0.25dex for Teff, logg, and [Fe/H], respectively. Our results are broadly consistent with those of recent studies using CARMENES data, with a systematic deviation in our Teff scale towards hotter values for estimations above 3750K. Furthermore, our methodology mitigates the deviations in metallicity found in previous DL techniques due to the synthetic gap. We consolidated a DTL-based methodology to determine stellar parameters in M dwarfs from synthetic spectra, with no need for high-quality measurements involved in the knowledge transfer. These results suggest the great potential of DTL to mitigate the differences in feature distributions between the observations and the PHOENIX-ACES spectra.2024A&A...687A.205Mhttps://cdsarc.cds.unistra.fr/viz-bin/cat/J/A+A/687/A205CatalogResearchIsServedByTAP VizieR generic serviceIsServedByConesearch servicehttps://cds.unistra.fr/vizier-org/licences_vizier.htmlhttps://vizier.cds.unistra.fr/viz-bin/VizieR-2?-source=J/A+A/687/A205https://vizier.iucaa.in/viz-bin/VizieR-2?-source=J/A+A/687/A205http://vizieridia.saao.ac.za/viz-bin/VizieR-2?-source=J/A+A/687/A205https://vizier.cds.unistra.fr/viz-bin/votable?-source=J/A+A/687/A205https://vizier.iucaa.in/viz-bin/votable?-source=J/A+A/687/A205http://vizieridia.saao.ac.za/viz-bin/votable?-source=J/A+A/687/A205GETtext/xml+votablehttps://tapvizier.cds.unistra.fr/TAPVizieR/tapCone search capability for table J/A+A/687/A205/tablea1 (The catalogue)https://vizier.cds.unistra.fr/viz-bin/conesearch/J/A+A/687/A205/tablea1?https://vizier.iucaa.in/viz-bin/conesearch/J/A+A/687/A205/tablea1?http://vizieridia.saao.ac.za/viz-bin/conesearch/J/A+A/687/A205/tablea1?GETtext/xml+votable180.050000true1.295371745.78656690.0055555555555555566/22 131 497 546 624 725 778 811 1033 1672 1822 1875 1938 2515 2545 2681 2737 2747 3031 3266 3345 3374 3658 3682 3715 3791 4033 4366 4444 4518 4595 5009 5041 5207 5241 5343 5420 5488 5534 5590 5849 6384 6552 6672 6676 6894 6980 7003 7103 7181 7262 7264 7395 7428 7471 7611 7775 7837 7921 7937 7940 7980 8023 8032 8123 8163 8208 8240 8394 8562 8627 8733 8843 8915 9188 9529 9578 9597 10037 10149 10225 10278 10448 10464-10465 10694-10695 10901 10999 11177 11255 11403 11636 11671 11962 12140 12184 12191 12621 12732 12865 13085 13427 13451 13567 13688 13875 14277 14395 14515 14709 14781 14850 14985 14993-14994 15052 15078 15311 15346 15431 15548 15598 15602 15635 15763 15792 15960 16041 16754 16758 17020 17271 17302 17348 17358 17412 17870 18195 18496 18553 18666 18789 18942 19050 19064 19131 19306 19352 19466 19516 19879 20023 20100 20109 20311 20459 20795 20896 21151 21171 21427 21747-21748 21953 21969 22014 22222 22277 22309 22626 22645 22709 22807 22904 22942 23014 23094 23363 23453 23564 23718 23731 23735 23743 23867 23910 23997 24213 24227 24239 24995 25009 25363 25388 25512 25528 25849 25989 26327 26487 26538 26560 26563 26583 26731 26757 26772 26821 26972 27201 27289 27320 27392 27459 27730 27734 27787 27820 27912 28223 28341 28343 28548 28646 28955 29146 29170 29684 30345 30429 30455 30490 30505 30771 30833 30893 31014 31019 31191 31229 31642 31686 31767 31781 31803 31811 32033 32076 32093 32115 32335 32392 32458 32491 32753 36732 40250 40359 40386 40436 40505 40513 40610 44518 44541 44728 44921 44936-44937 44967 48485 48500 48906 48935 48958 49101https://cdsarc.cds.unistra.fr/viz-bin/moc/J/A+A/687/A205?format=asciidefaultJ/A+A/687/A205/tablea1The catalogueSimbadNameSimbad column added by the CDSmeta.idcharnullablerecnoRecord number assigned by the VizieR team. Should Not be used for identification.meta.recordintKarmnCARMENES identifier (JHHMMm+DDdA)meta.id;meta.maincharNameDiscovery or most common name (1)meta.idcharRAdeg? Right ascension (J2000.0) (2)degpos.eq.ranullableDEdeg? Declination (J2000.0) (2)degpos.eq.decnullableTeffEffective temperature from our methodologyKphys.temperature.effectiveinte_TeffDelta Teff for -1 sigmaKstat.error;phys.temperature.effectiveintE_TeffDelta Teff for +1 sigmaKstat.error;stat.maxintloggSurface gravity from our methodologylog(cm.s**-2)phys.gravityfloate_loggDelta logg for -1 sigmalog(cm.s**-2)stat.errorfloatE_loggDelta logg for +1 sigmalog(cm.s**-2)stat.error;stat.maxfloat[M/H]Metallicity from our methodologylog(Sun)phys.abund.Zfloate_[M/H]Delta [M/H] for -1 sigmalog(Sun)stat.error;phys.abundfloatE_[M/H]Delta [M/H] for +1 sigmalog(Sun)stat.error;stat.maxfloatvsiniProjected rotational velocity from our methodologylog(km.s**-1)phys.veloc.rotatfloate_vsiniDelta vsini for -1 sigmalog(km.s**-1)stat.error;phys.veloc.rotatfloatE_vsiniDelta vsini for +1 sigmalog(km.s**-1)stat.error;stat.maxfloat_RARight Ascension (J2000) from SIMBAD (not part of the original data)degpos.eq.ra;meta.main_DEDeclination (J2000) from SIMBAD (not part of the original data)degpos.eq.dec;meta.main