Revisiting KELT-19Ab, WASP-156b and WASP-121b in the TESS Era.

We present a re-analysis of transit depths of KELT-19Ab, WASP-156b, and WASP-121b, including data from the Transiting Exoplanet Survey Satellite (TESS). The large $\sim21''$ TESS pixels and point spread function results in significant contamination of the stellar flux by nearby objects. We use Gaia data to fit for and remove this contribution, providing general-purpose software for this correction. We find all the three sources have a larger inclination, compared to earlier work. For WASP-121b, we find significantly smaller values (13.5 degrees) of the inclination when using the 30 minutes cadence data compared to the 2 minutes cadence data. We demonstrate using a simulation that 30-minute binning data in general, yields a smaller inclination than the true input value at 25$\sigma$. Also, we find that inclination and semi-major axis are biased small when applying a larger sampling time interval which is particularly important for deriving sub-percent transit differences between bands. If we constrain the inclination to previous work, we find the radius ratio of exoplanet to star ($R_{p}/R_{\ast}$) in the broad TESS band is 3.5$\sigma$ smaller than previous work for KELT-19Ab, and consistent to within $\sim$2$\sigma$ for WASP-156b and WASP-121b. The result for KELT-19Ab favors a haze-dominated atmosphere. We do not find statistically significant evidence for the $\sim$0.95\,$\mu$m water feature contaminating the transit depths in the TESS band for these stars but show that with photometric precision of 500ppm and with a sampling of about 200 observations across the entire transit, this feature could be detectable in a more narrow $z-$band. Furthermore, we find that with the inclusion of the TESS photometry, WASP-121b can be fitted just as well with a cloud-free atmosphere with water vapor as with an opaque featureless model.