Detectability of exoplanet transits with Athena's WFI instrument: testing for white and correlated noise

One of the science goal of the Athena mission is to detect and characterise, in the X-ray domain, transits of hot Jupiter-like planets orbiting their parent stars. To date, the only candidate for this kind of studies is HD 189733b, a Jupiter-size planet in a 2d orbit, for which a transit depth of 6-8% has been observed accumulating several Chandra and XMM-Newton observations. We simulate in this work realistic light curves of exoplanet transits using the Athena end-to-end simulator, SIXTE, and derive the expected signal-to-noise ratios (SNR) for different instrument configurations and planetary system parameters. We first produce at light curves for the currently existing WFI instrument designs and for different source fluxes to extract the expected (white noise) standard deviation. Next, moderate levels of correlated noise and transits of different depths are added to the light curves. As expected, for pure white noise the SNR is proportional to the square root of the flux, to the light curve bin size and to the number of co-added transits, and by definition proportional to the transit depth. When correlated noise starts to be significant, rebinning the data will only slightly increase the SNR, depending on the noise characteristics. Considering only white noise, a transit observed in a source like HD 189733, that has a flux around 5x10-13 erg s-1 cm-2 and a transit depth of about 5% can be detected with a SNR>3 in a unique transit. With correlated noise, several transits might be necessary. We also simulate trapezoidal shaped transits and try to recover the ingress/egress times after addition of noise. The relative error on the fitted ingress times is below 10% for most of the light curves with SNR>1.