MINERVA-UDS Observations, catalogs, and data products Authors: Sam Cutler, Luke Robbins, Yoshi Asada, WG2 Version: internal version n2.3_m2.4_v1.1.1 RECOMMENDED USE ------------ USE THE SUPER CATALOG! It uses the best apertures for each object. Otherwise: Large, bright, isolated sources --> 0.70" diameter measurements (D070) Compact, faint, crowded sources --> 0.32" diameter measurements (D032) High-redshift, speculative sources --> 0.20" diameter measurements (D020) Other single aperture catalogs (0.48", 1.00", 1.40") also provided upon request. If a source has negative flux in a band, check the SUPER aperture (use_aper) and try using smaller single aperture catalog. Note that smaller objects are blended near larger sources and so we do NOT provide Kron-corrected total photometry (flag_kron = 1). ABOUT APERPY --------- Aperpy is an aperture photometry software which utilizes source detection and photometry from SEP v1.4.1 (Barbary et al. 2016) with aperture corrections broadly following Whitaker et al. 2011. It follows a modular structure written entirely in python allow for flexibility and coherent pipeline flow commanded from a single configuration file. More details below. Codebase: https://github.com/astrowhit/aperpy/tree/sep1.4/ DETECTION and IDs: --------- Long-Wavelength (LW) Detection (F277W+F356W+F444W) combined using a noise-equalized technique: LW_F277W-F356W-F444W_noise-equal The images used in the detection are the unmatched background-subtracted images. Detected sources are crossmatched to IDs of the previous version of the corresponding catalog within a 0.08" radius (e.g. IDs in the LW catalogs are matched to IDs from the n2.1_m2.0_v1.0 LW catalogs). New sources in the detection or source that have more than one match (i.e. were deblended) are assigned new IDs starting at the last crossmatched ID. The first digit of IDs is a field indicator: in this catalog, all sources begin with 1 to indicate the source is from the UDS field. The field indicators for all MINERVA fields are: 1 - UDS 2 - COSMOS 3 - EGS 4 - GOODS-N MOSAICS/BANDS: ------------ The included photometric bandpasses are: HST/ACS - F435W, F606W, F775W, F814W, F850LP HST/WFC3 - F098M, F105W, F125W, F140W, F160W JWST/NIRCam - F090W, F115W, F140M, F150W, F162M, F182M, F200W, F210M, F250M, F277W, F300M, F335M, F356W, F410M, F430M, F444W, F460M, F480M Mosaics units are Fnu, 10*nJy (AB zpt=28.9). Note: catalog fluxes are Fnu ALSO have zpt = 28.9 Note that all bands are measured at 0.04"/pixel scale. For SW bands produced at 0.02"/px, pixels are co-added to scale to 0.04"/pixel. Grizli mosaics (v8.0): https://s3.amazonaws.com/grizli-v2/MINERVA/mosaics/uds/40mas-v2.3/index.html SKY BACKGROUND: ------------ Sky background is now removed during the data reduction step. No additional background subtraction is done in the photometry pipeline. PSF MATCHING: ------------ For sources identified in each detection strategy, we perform photometry on both a set of a images PSF-matched to F444W. To prepare the PSF matched images, both the JWST and HST PSFs were constructed directly from stars in the FOV. Stamps are 4" FOV at 0.04"/pixel. The HST and JWST/NIRCam medium band PSFs are suboptimal in some of the bands with less coverage but sufficient for these purposes and the growth curves agreed to the 1% level or less (for most HST/JWST bands). NEW WITH v1.1 PIPELINE: PSFs are matched to F444W at the native pixel scale (i.e. with no oversampling) using the pypher code (https://pypher.readthedocs.io/). A regularization parameter of 3e-5 is used for HST/WFC3 filters, F430M, F460M, and F480M, while 1e-4 is used for all other filters. FLUXS & ERRORS: ------------- Fluxes are measured separately in 0.20", 0.32", 0.48", 0.70", 1.00", and 1.40" diameter apertures that are corrected to total fluxes based on the Kron radius plus a roughly 5-10% correction from Kron to total to account for missing light beyond 1" radius based on the F444W curve of growth. Kron radii are measured in a stack of all bands used in the detection image, except using the PSF-matched images (e.g. the F277W, F356W and F444W PSF-matched images are stacked for the LW detection). LW detections use the '_f277w+f356w+f444w' suffix for the Kron band measurements, while the chi-mean detections use '_KRON'. A wavelength dependent correction for galactic extinction at the position of UDS has been applied (of order 2% and less). Errors are calculated with empty apertures where the conversion from the aperture error to total flux error is based on F444W filter and noise properties. All fluxes have a unit of 10 * nJy with a corresponding AB magnitude zeropoint of 28.9. Photometry for significantly blended objects (flag_kron = 1) is computed without Kron photometry (color apertures are scaled to total via the PSF curve of growth). Blends are identified by SEP, but those which are the brightest sources within their blends as determined by isophotal (segment) fluxes are not considered blends so long as the area of their Kron ellipse is within 50% of their isophotal area. Objects whose Kron ellipse areas are less than their isophotal areas are also not considered blends. The SUPER catalog combines the five catalogs with adaptive aperture selection based on their isophotal areas (see Labbe et al. 2003). Apertures for objects which are considered blends are selected based on a conservative 20% reduction of their isophotal area to ensure robust colors. STAR-GALAXY SEPARATION: ----------- Stars are flagged as 'flag_star' == 1. They are identified as satisfying at least one of the following of criteria: 1. Having a F200W flux ratio in 0.7" / 0.32" apertures between 1.0 and 1.2 & brighter than 25.0 AB in 0.7" apertures 2. Having a F160W flux ratio in 1.4" / 0.7" apertures between 1.0 and 1.2 & brighter than 23.8 AB in 0.7" apertures For completeness, we also adopt stars detected directly from the F160W mosaics, as those with significant proper motions are shifted from their JWST-detected centroids. Gaia-identified stars are also flagged. Lastly, we flag any source in any band that was included in our more conservative selection that contribute to our PSF stacks. ARTIFACTS: ------------ Artifacts are flagged as 'flag_artifact' == 1. False-positive sources around bright stars and their diffraction spikes are flagged using hand-made regions. Any source that overlaps with this region is flagged as an artifact. This is particularly useful for the brightest stars that are saturated, meaning they cannot be identified from photometry and often cause multiple false detections near their saturated cores. A copy of the region file is provided. Although most artifacts are now flagged, there are still some artifacts detected at larger radii. Visual inspection is strongly advised. We are working on more robust masking for bright stars in future versions. In addition, there appear to be a number of bad pixels in the LW channels (F277W through F444W) which can appear as bright SW-dropouts. To mitigate contamination we flag bad pixels as having a F444W flux ratio in 0.7" / 0.32" apertures below 1.1 brighter than 26.0 AB in 0.32" apertures. There are also a number of detections near mosaic edges where detector effects cause data loss, seen as regions of zeros in the detection weight map. We flag any object within 3" radius of such a region whose area is between 10 and 1000 pixels, calibrated visually. USE_PHOT: ------------ We provide a 'use_phot' flag to denote objects that are stars, artifacts, are marginally detected, or have no usable photometry. Marginally detected sources, based a signal-to-noise < 3 in the detection image, are also denoted by the 'flag_lowsnr' flag. A handful of sources have no usable photometry. This is particularly prevalent in the larger aperture catalogs where the aperture either contains a bad pixel or the edge of the mosaic. For the 'SUPER' catalog, we include the most suitable aperture size, but may include a smaller size if the most suitable one is unusable for either of these reasons. MIRI PHOTOMETRY: ------------ Total fluxes in MIRI filters are given in this catalog. The total flux in a MIRI filter is measured based on the color between NIRCam/F444W. The color is measured in a reference fixed-aperture on the PSF-matched F444W image vs the MIRI image, and then the flux is scaled to total to match the NIRCam SUPER photometry from MINERVA WG2. The reference aperture sizes are: F770W: 0.7" diameter F1280W: 1.2" diameter F1500W: 1.2" diameter F1800W: 1.5" diameter USE_PHOT_MIRI: ------------ "USE" flag for MIRI photometry. The flag is TRUE when: 1. FLAG_PERSISTENCE is False for all MIRI filters 2. FLAG_EDGE is False for all MIRI filters 3. S/N > 2 in F444W PHOTOMETRIC REDSHIFTS WITH EAzY: ------------ All catalogs are processed with EAzY to derive best-fit redshifts and physical properties. There are multiple photo-z catalogs with different templates (Larson and SFHZ_BLUE_AGN) with varying strengths and weaknesses. Generally, the Larson templates (see Larson et al. 2022) are preferred due to their photo-z performance, but we note that this template only provides photo-z fits and no stellar population measurements (e.g. mass, SFR, etc.). The SFHZ_BLUE_AGN templates have comparable photo-z quality. We run all catalogs (templates+apertures) both allowing for spatially-dependent corrections (ZPiter) and without corrections (noZPiter). These photo-z's are computed based on SUPER catalog photometry; photo-z's from single-aperture photometry are not provided but are available on request. The SUPER photometry has a lower outlier fraction and scatter, but the smaller single apertures may be preferred for small, high-redshift searches. We supply photo-z catalogs as well as .h5 files which can be used to visualize the SED fits, along with the .param, .zeropoint, and .translate files for further reference. More robust photo-z's and stellar population fits will be available soon from Prospector. PHOTOMETRY CATALOG FORMAT: ------------ id :: object identification number x/y :: centroid position in image frame ra/dec :: centroid position in world frame ebv_mw :: MW attenuation correction in E(B-V) computed from Schlafly & Finkbeiner (2011) faper_f277w+f356w+f444w/faper_KRON: flux computed in circular aperture of size corresponding to filename, in 10*nJy (Fnu) corresponding to ABMag zpt=28.9 eaper_f277w+f356w+f444w/eaper_KRON: flux error computed from empty apertures (same size as above), scaled by local weight / median weight fauto_f277w+f356w+f444w/fauto_KRON: flux computed within the Kron radius w_f277w+f356w+f444w/w_KRON: weight at the position of the source relative to the maximum weight (i.e. large values --> deep regions) f_fXXXw :: total flux computed in circular aperture corrected using circularized kron radius, in 10*nJy (Fnu) corresponding to ABMag zpt=28.9 e_fXXXw :: total flux error computed from empty apertures (same size as above), corrected to total using circularized kron radius w_fXXXw :: weight at the position of the source relative to the maximum weight (i.e. large values --> deep regions) t_fXXXw :: exposure time of observations at the position of the source, in seconds tot_cor :: aperture to total correction based on circularized kron radius kron_radius :: Kron "factor" (unitless), semi-major Kron radius can be computed via kron_radius * a_image kron_radius_circ :: circularized Kron radius in arcsec use_circle :: 1 when no kron factor is applied (most are due to unreliable kron radii, but some are just small point-like things) flag_kron :: 1 when kron radius is not robust, in which case total photometry assumes point-like light distribution iso_area :: isophotal area based on the source segment in sq. arcsec use_aper :: (SUPER catalog only) indicates arcsec diameter adopted for color aperture a_image :: semi-major axis in pixels b_image :: semi-minor axis in pixels theta_J2000 :: position angle in degrees clockwise from the first image axis (along y) flux_radius :: radius containing 50% of the flux in arcsec use_phot :: 1 if photometry is reliable (f444w SNR > 3, not a star, not likely a bad pixel) flag_nophot :: 1 if the object has no viable photometry in any band at this aperture size flag_lowsnr :: 1 if S/N < 3 in the f277W+f356w+f444w stack flag_star :: stellarity flag (see above) flag_artifact :: 1 if bad pixel, bCG-subtraction residual, or image artifact (see above) flag_acs_coverage :: 1 if the source has both HST/ACS F435W and F606W flux, useful for vetting photo-z's n_bands_mb :: number of medium band filters with coverage of a source z_spec :: archival spectroscopic redshifts; includes spec-z's from 3D-HST (archival), MOSDEF (Z_QUAL>4), DJA NIRSpec v4.2 (grade=3) MIRI photometry columns --------------------------- f_{FILTER} total flux in the MIRI filter; in 10 nJy; e_{FILTER} flux error on f_{FILTER}; in 10 nJy; Scale_factor_{REF_APER}_{FILTER} scale factor from the fixed-aperture to the total. The scaling is already applied Localbkg_{FILTER} local background level around the source in this filter. The correction is already applied; 10 nJy/pix2; FLAG_PERSISTENCE_{FILTER} identifies sources whose reference fixed-aperture contains bad persistence pixels in the filter; FLAG_EDGE_{FILTER} identifies sources that are close to the edge of the MIRI filter footprint (less than 1.5 arcsec); use_phot_miri identifies sources with good MIRI photometry. PHOTO-z CATALOG FORMAT: ------------ Photo-z information is in the *.zout.fits files, including the following columns: z_phot :: best-fit photometric redshift (median likelihood) z025/160/840/975 :: percentiles of P(z), e.g. z025 -> 2.5% z_phot_chi2 :: total chi-squared at z_phot nusefilt :: number of filters used in the fit restU :: rest-frame U-band flux (same as photometry: Fnu, units corresponding to ABmag = 28.9) restU_err :: rest-frame U-band flux uncertainty (same as photometry: Fnu, units corresponding to ABmag = 28.9) restV :: rest-frame V-band flux (same as photometry: Fnu, units corresponding to ABmag = 28.9) restV_err :: rest-frame V-band flux uncertainty (same as photometry: Fnu, units corresponding to ABmag = 28.9) restJ :: rest-frame J-band flux (same as photometry: Fnu, units corresponding to ABmag = 28.9) restJ_err :: rest-frame J-band flux uncertainty (same as photometry: Fnu, units corresponding to ABmag = 28.9) rest414 :: rest-frame synth. u-band flux (same as photometry: Fnu, units corresponding to ABmag = 28.9; see Antwi-Danso et al. 2023) rest414_err :: rest-frame synth. u-band flux uncertainty (same as photometry: Fnu, units corresponding to ABmag = 28.9; see Antwi-Danso et al. 2023) rest415 :: rest-frame synth. g-band flux (same as photometry: Fnu, units corresponding to ABmag = 28.9; see Antwi-Danso et al. 2023) rest415_err :: rest-frame synth. g-band flux uncertainty (same as photometry: Fnu, units corresponding to ABmag = 28.9; see Antwi-Danso et al. 2023) rest416 :: rest-frame synth. i-band flux (same as photometry: Fnu, units corresponding to ABmag = 28.9; see Antwi-Danso et al. 2023) rest416_err :: rest-frame synth. i-band flux uncertainty (same as photometry: Fnu, units corresponding to ABmag = 28.9; see Antwi-Danso et al. 2023) mass :: median stellar mass at z_phot (M_sun) mass_p :: stellar mass percentiles at z_phot (M_sun) sfr :: median star-formation rate at z_phot (M_sun / yr) sfr_p :: star-formation rate percentiles at z_phot (M_sun / yr) Av :: attenuation normalized to V (ABmag) Av_p :: attenuation percentiles normalized to V (ABmag) star_min_chi2 :: chi2 of best-fit stellar template (NB: fit w/ MW attenuation applied...) uvj_class :: quick UVJ classification based on Whitaker+12 (0 = SF, 1 = Q, -1 = outside reasonable color bounds) flag_eazy :: 1 if fit is reliable