Total: 1
Few gas giants have been observed in their accretion phase. ELT instruments will have a higher sensitivity and a smaller inner working angle than instruments up to now, allowing detailed characterisation. We study the observability of accreting gas giants with METIS, the first-generation ELT spectrograph with a resolution R = 1e5. We focus on the accretion-tracing hydrogen recombination lines accessible to METIS, mainly Brackett alpha and Pfund-series lines. Our approach is general but we take PDS 70 b as a fiducial case. To calculate high-resolution line profiles, we combine a semianalytical multidimensional description of the flow onto an accreting planet and its circumplanetary disc (CPD) with local NLTE shock-emission calculations. We assume the limiting scenario of no extinction, appropriate for gas giants in gaps, and negligible contribution from magnetospheric accretion. We use resolved photospheric models, and include detector sensitivities to compute required observing times. Both the planet surface and the CPD surface shocks contribute to the total line profile, which is non-Gaussian and much narrower than the free-fall velocity. When the accretion onto PDS 70 b is at its maximum observed strength, the Br a line peak is equal to the photospheric continuum, which is modulated mostly by water features. However, the rotation of the planet broadens the features, making the shock excess stand out. At Br a, already only the continuum of PDS 70 b should yield SNR = 12 in 4 h, and for the fiducial accretion rate, the line peak will have a per-bin SNR = 28 in 4 h. The peak excess should require only about 10 min to reach SNR = 3. Br a is a potent planet formation tracer accessible to ELT/METIS in little integration time. Resolved line profiles will place independent constraints especially on the mass and radius of accreting planets, and help identify the accretion mechanism(s) at work.