Echelle-MEPSICRON time-resolved spectroscopy of the dwarf nova SS Cygni

High-dispersion time-resolved spectroscopy of the dwarf nova SS Cygni was obtained during two consecutive days at the end of an outburst. The emission lines present a double-peaked profile with a peak-to-peak separation of $$526 \pm 14\,\text{ km s}^{-1}.$$ The lines have a complex structure which includes a narrow asymmetric component coming from a hot spot in the accretion disc or from the chromosphere of the red dwarf. The radial velocity analysis yields semi-amplitude values of $$K_\text{em}=96 \pm 5\,\text{ km s}^{-1}$$ for the wings of Hγ and Hβ, and of $$K_\text{ab}=152 \pm 2\,\text{ km s}^{-1}$$ for the cross-correlated secondary star. The rotational velocity of the red star, measured from the correlation profile is $$V_\text{rot}=99 \pm 8\,\text{ km s}^{-1},$$ consistent with the calculated value of $$85 \pm 8\,\text{ km s}^{-1}$$ for a semi-detached tidally locked binary with similar K semi-amplitudes. Detailed analysis of the absorption lines of the secondary indicate a K2V spectral type at the time of the observations. The recently discovered light modulations in the optical and ultraviolet suggest that the inclination angle might be higher than the expected canonical value of 38°, probably around 50°. If this is correct, the mass and spectral type of the secondary should be 0.40 M_⊙ and M1V, respectively. We suggest that a heating effect on the secondary by the white dwarf and accretion disc might be responsible for this discrepancy in spectral type. This idea is supported by our finding of a K2 spectral type when the system is approaching quiescence instead of a K5 spectral type found by other groups when the system was quiescent. It is also supported by theoretical calculations of the semi-amplitude of the radial velocity curves in SS Cyg B when heating effects are taken into account (Robinson, Zhang & Stover) although, as we show in our discussion, observational support for these K_ab variations is not conclusive.