The Nancy Grace Roman Space Telescope, is a flagship NASA mission. Science investigations with its Wide Field Imager (WFI) include dark energy and cosmology investigations via a high galactic latitude imaging and spectroscopic and Type Ia supernovae (SNe Ia) surveys. The WFI’s field of view has 100 times the area of HST imaged on 18 H4RG HgCdTe detectors. These semi-conductor detectors have an inherently non-linear response to light. One such artifact is count-rate non-linearity (CRNL),where the detector’s response depends on the flux of the incoming light. CRNL physics is still not well understood; the current hypothesis ascribes the effect to charge traps in the semiconductor (Mosby et al. 2020). At present, we assume a per pixel, wavelength dependent model where CRNL is proportional to a power law of the incoming flux.
The nature of this effect makes faint sources appear fainter and bright sources seem brighter. If uncorrected, CRNL limits the accuracy and confidence in the SNe Ia and weak lensing measurements of dark energy. To enable on-orbit measurement and tracking of the CRNL, Roman’s WFI carries a Relative Calibration System (RCS) that allows selecting the intensity and wavelength of light illuminating the focal plane. In the direct illumination mode flat fields at six levels of intensity can be acquired. In lamp-on/lamp-off (LOLO) mode, the detectors are simultaneously illuminated by the RCS while observing an astronomical scene.
Here we present the results of a study of the effects on the dark energy figure of merit (FoM) if the Relative Calibration System capabilities are reduced or increased from the original baseline, thus affecting the characterization of CRNL. A higher FoM value implies greater constraints on the dark energy parameters of interest; we only consider the effect of CRNL. We determine the FoM given a set of filter combinations in direct illumination (DI) and lamp-on/lamp-off (LOLO), relative to the baseline. We find that DI in the six SN/Deep-field filters (RZY JHF) or LOLO in the six SN/deep-field filters or DI in three filters (RZJ) plus LOLO in two (WF) meet the FoM requirement.