We present H‐ and K‐band spectroscopy of OB and Wolf–Rayet (WR) members of the Milky Way cluster 1806−20 (G10.0–0.3) to obtain a revised cluster distance, of relevance to the 2004 giant flare from the (soft gamma repeater) SGR 1806−20 magnetar. From GNIRS (Gemini Near‐Infrared Spectrograph) spectroscopy obtained with Gemini South, four candidate OB stars are confirmed as late O/early B supergiants, while we support previous mid‐WN and late WC classifications for two WR stars. Based upon an absolute Ks‐band magnitude calibration for B supergiants and WR stars, and near‐infrared (IR) photometry from NIRI (Near‐Infrared Imager) at Gemini North plus archival VLT/ISAAC (Very Large Telescope/Infrared Spectrometer And Array Camera) data sets, we obtain a cluster distance modulus of 14.7 ± 0.35 mag. The known stellar content of the 1806−20 cluster suggests an age of 3–5 Myr, from which theoretical isochrone fits infer a distance modulus of 14.7 ± 0.7 mag. Together, our results favour a distance modulus of 14.7 ± 0.4 mag (8.7+1.8−1.5 kpc) to the 1806−20 cluster, which is significantly lower than the nominal 15 kpc distance to the magnetar. For our preferred distance, the peak luminosity of the 2004 December giant flare is reduced by a factor of 3 to 7 × 1046 erg s−1, such that the contamination of BATSE (Burst And Transient Source Experiment) short gamma‐ray bursts (GRBs) from giant flares of extragalactic magnetars is reduced to a few per cent. We infer a magnetar progenitor mass of ∼48+20−8 M⊙, in close agreement with that obtained recently for the magnetar in Westerlund 1.

We present near‐infrared Very Large Telescope/Infrared Spectrograph and Array Camera and mid‐infrared (mid‐IR) Spitzer/Infrared Spectrograph spectroscopy of the young massive cluster in the W31 star‐forming region. H‐band spectroscopy provides refined classifications for four cluster member O stars with respect to Blum et al. In addition, photospheric features are detected in the massive young stellar object (massive YSO) #26. Spectroscopy permits estimates of stellar temperatures and masses, from which a cluster age of ∼0.6 Myr and distance of 3.3 kpc are obtained, in excellent agreement with Blum et al. IRS spectroscopy reveals mid‐IR fine structure line fluxes of [Ne ii–iii] and [S iii–iv] for four O stars and five massive YSOs. In common with previous studies, stellar temperatures of individual stars are severely underestimated from the observed ratios of fine‐structure lines, despite the use of contemporary stellar atmosphere and photoionization models. We construct empirical temperature calibrations based upon the W31 cluster stars of known spectral type, supplemented by two inner Milky Way ultracompact (UC) H ii regions whose ionizing star properties are established. Calibrations involving [Ne iii] 15.5 μm/[Ne ii] 12.8 μm, [S iv] 10.5 μm/[Ne ii] 12.8 μm or [Ar iii] 9.0 μm/[Ne ii] 12.8 μm have application in deducing the spectral types of early to mid O stars for other inner Milky Way compact and UC H ii regions. Finally, evolutionary phases and time‐scales for the massive stellar content in W31 are discussed, due to the presence of numerous young massive stars at different formation phases in a ‘coeval’ cluster.

Aims. A near-infrared study of the main ionizing star of the ultracompact HII region G23.96+0.15 (IRAS 18317-0757) is presented, along with a re-evaluation of the distance to this source, and a re-assessment of H-and K-band classification diagnostics for O dwarfs Methods. We have obtained near-IR VLT/ISAAC imaging and spectroscopy of G23.96+0.15, plus archival imaging from UKIRT/UFTI. A spectroscopic analysis was carried out using a non-LTE model atmosphere code. Results. A quantitative H-and K-band classification scheme for O dwarfs is provided, from which we establish an O7.5V spectral subtype for the central star of G23.96+0.15. We estimate an effective temperature of T eff ∼ 38 kK from a spectral analysis. Conclusions. A spectroscopic distance of 2.5 kpc is obtained for G23.96+0.15, substantially lower than the kinematic distance of 4.7 kpc, in common with recent studies of other Milky way H ii regions. Such discrepancies would be alleviated if sources are unresolved binaries or clusters.

We have obtained Spitzer IRAC and MIPS mid-IR images of a sample of 43 radio selected UCHII region sources to ascertain (a) whether the newly born O stars within are found with other stars in their birthplaces and (b) the nature of the surroundings. 37 of the sources appear to be in small clusters, and 33 are found in connection with other hot star formation activity. Thus, for the most part, O stars are not born in isolation. Here we give examples of the mid-IR images of the various types of UCHII regions.

We highlight how the downward revision in the distance to the star cluster associated with SGR 1806-20 by Bibby et al. (2008) reconciles the apparent low contamination of BATSE short GRBs by intense flares from extragalactic magnetars without recourse to modifying the frequency of one such flare per 30 years per Milky Way galaxy. We also discuss the variety in progenitor initial masses of magnetars based upon cluster ages, ranging from $50 M for SGR 1806-20 and AXP CXOU J164710.2-455216 in Westerlund 1 to $17 M for SGR 1900+14 according to Davies et al. (2009) and presumably also 1E 1841-045 if it originated from one of the massive RSG clusters #2 or #3.