Circumstellar Disk Models for use with WFIRST CGI data simulations ------------------------------------------------------------------ John Debes(1) and Ewan Douglas(2) 1 Space Telescope Science Institute 2 The University of Arizona (If you use these models, please cite Mennesson et al. 2018, Proc. SPIE, 10698) 20 January 2020 The files in this directory consist of 0.03 AU resolution disk models at 10pc that have then been resampled to 21 mas pixels and convolved with publicly available field dependent PSFs of the WFIRST CGI Hybrid Lyot Coronagraph (HLC) mask and Wide Field of View Shaped Pupil Coronagraph (SPCWIDE) mask designs. These files are appropriate for insertion into simulated observing scenarios for WFIRST CGI, and cover two possible types of disks: narrow disk rings similar to the circumstellar debris disk HR 4796A and disk annuli with radial dust density power-laws that approximate the density distribution of the Solar System Zodiacal cloud (Kennedy et al. 2015, ApJS, 216, 23). Both types of disks are then scaled to a surface brightness in V of 22.5 magnitudes per square arcsecond at the peak surface brightness of the given disk in a face-on (inclination=0 degrees) configuration. This surface brightness is chosen under the assumption that a face-on disk is equivalent to looking through our local Zodiacal cloud at an ecliptic latitude of 90 degrees and that the total optical depth of dust is 2x that of our vantage point within the disk (cf. Appendix B and C of Stark et al. 2014, ApJ, 795, 122). The value of 22.5 comes from Table 1 of TIR-CRDS-2015-01 by R. Diaz (http://www.stsci.edu/files/live/sites/www/files/home/hst/instrumentation/reference-data-for-calibration-and-tools/documentation/_documents/TIR-CRDS-2015-01.pdf), which gives updated values for the local zodiacal light as a function of ecliptic latitude and heliocentric longitude. The models have a range of inner radii to the disks. The vertical density distribution of the disk is approximated by a gaussian with a vertical full width at half-maximum of 0.2*R AU, where R is radial distance from the star. The dust in the disks is assumed to have a scattering phase function equivalent to the dust in Saturn's G ring (Hedman & Stark 2015, ApJ, 811, 67). This phase function captures the highly forward scattering nature of both water ice rich grains, such as in Saturn's rings and silicate rich dust such as in the Zodiacal cloud. The exact mass of dust in these models is not set, since the precise optical properties of exo-zodiacal dust disks is currently highly uncertain. However, the scaling is appropriate for the surface brightness of the Zodiacal Cloud at 1 AU. The literature often discusses "zodis" as a shorthand for discussing an amount of dust scaled from the Zodiacal Cloud and can depend either on being scaled from the total IR luminosity of dust relative to the star or on the scattered surface brightness of the dust. However, these definitions are not equivalent because the scaling between IR luminosity of dust and its scattering efficiency is model dependent. Since dust disks often come in a range of masses, physical extent, and constituent grains, we instead focus on those disks that have equivalent observed surface brightness to the Zodiacal Cloud since this is the primary observable feature of such a disk for WFIRST/CGI. Convolved images were constructed using the field dependent PSFs appropriate for each type of mask and are publicly available at IPAC at the following websites. Each site describes the pedigree of the PSFs and may be updated in the future as designs are finalized. HLC: https://wfirst.ipac.caltech.edu/sims/off_axis_PSF.html SPCWIDE: https://wfirst.ipac.caltech.edu/sims/off_axis_PSF_SPCWIDE.html The HLC convolutions were validated against Observing Scenario 6 (OS6; https://wfirst.ipac.caltech.edu/sims/Coronagraph_public_images.html#CGI_OS6). Point sources with flux ratios of 1e-8 and 3e-9 respectively were convolved at 3.5 and 4.5 lambda/D and their core photometry was compared to the noiseless OS6 images (both with and without MUFs) and verified to be consistent within a few percent of OS6 simulations when scaled by the reported integrated stellar flux (stellar flux value used: 1.15x10^8 ph/s for a V=5 G0 star). To scale any of the convolved models to counts on a detector, one must know the flux density of the target star in the given bandpass (F_nu_star(Jy)) and the conversion between the stellar flux density and total integrated counts/s of the star on the detector (F_star(cts/s)). The models can then be scaled by F_star/F_nu_star. File naming Convention ------------------------------------------------------------------ Full resolution images are named as "modeltype"_inc"inclination angle"_r"radius in arcseconds".fits and include the surface brightness in units of Jy/pixel. Resampled images are like the full resolution images but resampled to the Direct Imaging Camera (DICAM) plate scale of 21 mas and are named "modeltype"_inc"inclination angle"_r"radius in arcseconds"_DICAM.fits. Convolved and resampled files are named as "modeltype"_inc"inclination angle"_r"radius in arcseconds"_"masktype".fits. "modeltype" is given either by "ring" or "zodi", "inclination angle" ranges from a face-on disk and inclination of 0 degrees to a nearly edge-on disk with an inclination of 89 degrees with intermediate inclinations of 30, 45, and 60 degrees. "radius in arcseconds" for the ring models include disk radii of 1,2,3,4,5,6,8,10, and 12 AU. The "zodi" models are annuli with an outer radius of 14 AU and so "radius in arcseconds" refers to the inner radius of the disk and includes inner radii of 1,2,3,4,5, and 6 AU. Two files are included to simulate the field stops for both modes that correspond to the extent of the highest contrast regions of each mode. For the HLC, it was assumed that this extends from 3 lambda/D to 9 lambda/D (lambda=575 nm), and for the SPCWIDE it was assumed to extend from 6.5 lambda/D to 20 lambda/D (lambda=825 nm). Scaling Models ------------------------------------------------------------------ The scaling used for these models as a function of inclination angle implies that each model has the same total mass of dust for a given radius model. One can roughly infer how many zodis of surface brightness are present by multiplying by the ratio of the radius to 1 AU (or 0.1") squared. For example, the ring_inc0_r0.5.fits full resolution model would correspond to a surface brightness at the disk radius equivalent to 25x the surface brightness of Zodiacal dust at the same radius. Models can thus be linearly scaled up or down to the desired "zodi" level. Users interested in making use of self-consistent dust disk models are encouraged to use the radiative transfer models that include IR SEDs and images from the WFIRST Preparatory Science Project: The Circumstellar Environments of Exoplanet Host Stars (PI: C. Chen). While these models are more limited in geometry, they provide a useful additional foundation to explore how surface brightness relates to disk mass.