Palle, E., Nortmann, L., Casayas-Barris, N., Lampon, M., López Puertas, M., Caballero, J. A., Sánz Forcada, J., Lara, L. M., Nagel, E., Yan, F., Alonso Floriano, F. J., Amado, P. J., Chen, G., Cifuentes, C., Cortes Contreras, M., Czesla, S., Molaverdikhani, K., Montes, D., Passegger, V. M., Quirrenbach, A., Reiners, A., Ribas, I., Sánchez López, A., Schweitzer, A., Stangret, M., Osorio, M. R. Z., Zechweitzer, M. (2020). A He I upper atmosphere around the warm Neptune GJ 3470 b. Astronomy and Astrophysics, 638 DOI: 10.1051/0004-6361/202037719
High resolution transit spectroscopy has proven to be a reliable technique for the characterization of the chemical composition of exoplanet atmospheres. Taking advantage of the broad spectral coverage of the CARMENES spectrograph, we initiated a survey aimed at characterizing a broad range of planetary systems. Here, we report our observations of three transits of GJ 3470 b with CARMENES in search of He (2(3)S) absorption. On one of the nights, the He I region was heavily contaminated by OH(-)telluric emission and, thus, it was not useful for our purposes. The remaining two nights had a very different signal-to-noise ratio (S/N) due to weather. They both indicate the presence of He (2(3)S) absorption in the transmission spectrum of GJ 3470 b, although a statistically valid detection can only be claimed for the night with higher S/N. For that night, we retrieved a 1.5 +/- 0.3% absorption depth, translating into aR(p)(lambda)/R-p= 1.15 +/- 0.14 at this wavelength. Spectro-photometric light curves for this same night also indicate the presence of extra absorption during the planetary transit with a consistent absorption depth. The He (2(3)S) absorption is modeled in detail using a radiative transfer code, and the results of our modeling efforts are compared to the observations. We find that the mass-loss rate, (M) over dot, is confined to a range of 3 x 10(10)g s(-1)forT= 6000 K to 10 x 10(10)g s(-1)forT= 9000 K. We discuss the physical mechanisms and implications of the HeIdetection in GJ 3470 b and put it in context as compared to similar detections and non-detections in other Neptune-size planets. We also present improved stellar and planetary parameter determinations based on our visible and near-infrared observations.