Federica Chiti

As Hot Jupiters orbit around their host stars, the stellar flux changes according to their position relative to the star as illustrated in Figure 1a: when the planet transits in front of the star, a fraction of its light is blocked, hence the dip in the light curve. Once the transit ends, the light curve becomes flat because we see the star plus the planet’s nightside. Eventually, the planet passes behind the star during its secondary eclipse thus the light curve shows another dip before flattening again when the planet’s dayside becomes visible. Spectrophotometry measures such flux changes as a function of wavelength. By subtracting the stellar spectrum from the combined spectrum of both the star and the exoplanet before an eclipse, we obtain the planet’s emission spectrum (Figure 1b).
We aim to test the spectrophotometric technique as a new method to detect planets and, combined with an analysis of emission spectra, to put constraints on the composition of the planetary atmosphere.
We use flux-calibrated spectra from XSHOOTER, a medium-resolution spectrograph of the Kueyen Telescope, the second compact building from the left that is portrayed in Figure 2, with the central region of the Milky Way in shades of purple and pink in the background.
We write a Python code that classifies the spectra of our targets, WASP-18b and WASP-19b, as out-of- or in- eclipse spectra. The spectra with the best signal-to-noise ratio are normalised relative to the continuum, co-added and averaged. In Figure 3, we plot the normalised flux against the wavelengths in nanometres to obtain the spectrum of the WASP-18 system (solid black line), whose profile resembles the one of a black body. The atmospheric absorption bands (vertical light blue spans) and the transition region from the visible to the infrared arm (vertical grey band) are excluded from our analysis.
We integrate each spectrum over bands of 50 nm to obtain the photometric points. This is illustrated in Figure 4 where the flux in erg cm-2 s-1 angstrom-1 nm-1 on the y-axis is plotted against the time in Barycentric Julian Date (BJD). The photometric points computed from spectra taken on the same date are aligned vertically; photometric points obtained from integrating over the same wavelength band but on different dates are aligned horizontally. The out-of-eclipse points on the right reveal dips of 10-15% due to calibration issues that do not allow to detect any planetary signature (~3%).
Consequently, we integrate over narrower bands around the sodium (Na) doublet, thus introducing a correction factor of 50 in the calibration. The results for WASP-18b are shown in the left part of Figure 5: the level of the points during the eclipse is lower compared to the out-of-eclipse phase. On the right part, we plot the ratio between the normalised flux out-of-eclipse and the one during eclipse against the wavelength as a black solid line. The location of the two Na lines (vertical green dashed lines) correspond to the spikes of the solid line highlighting the presence of Na. In both planets the Na signature is >6 σ.
The same procedure is applied to the Helium (He) triplet, however Figure 6 shows that the He signature is <3 σ in both planets.
Despite the calibration level of the spectra does not allow to conduct spectrophotometry, we introduce a correction factor of 50 by focusing on narrower bands and using them relative to the local continuum, thus achieving a level that allows to use ground-observations to detect planetary signatures usually identified through space-observations. Indeed, Na signatures >6 σ are detected in both WASP-18b and WASP-19b while no He seems to be present.