We provide different photometric measurements, described in the MER Photometry Cookbook. Given the different techniques and algorithms, it is expected they can be discrepant to a few percent. However, in some cases the discrepancies are larger than expected, particularly for bright stars in NIR bands (TEMPLFIT/SERSIC are consistently fainter by
The various morphological measurements provided in the MER final catalogs are described in the MER Morphology Cookbook
CAS parameters are calibrated relatively for Euclid data. The absolute calibration with the literature (e.g CANDELS) might differ: Completeness and Asymmetry parameters show a good correlation with the literature for bright objects (
The current masking of persistence affected pixels by OU-NIR is not sufficient. The co-added NIR images produced in the OU-MER PF show elongated streaks that are produced by persistence from bright stars in NISP spectroscopy. Some parts of those streaks are (NIR) detected and cataloged in the MER Processing Function and end up as false positives in the final MER catalogs.
As part of our data validation we compare the fluxes in our final catalog with the expected fluxes for GAIA stars. The offsets we are finding are usually in the range of a few percent. We are constantly testing and also discussing with other OU’s to understand and improve these offsets. In particular:
We find a consistent offset of
We determine an offset of
The incorrect treatment of the PSF stamps leads also to an offset of
These offsets are roughly in agreement with flux corrections that are applied in OU-PHZ to get consistent photometric redshift estimates.
The rms maps that MER currently gets delivered from EXT stage 2 do not contain contributions from the object photon shot noise (See issue #28268). As a consequence this noise component is also missing in the error estimates of all EXT bands in the MER final catalogs and the signal-to-noise ratio especially of bright sources is too high.
In the no-detection Sersic fitting we need to provide an object size to the fitting code which is then translated to the size of the fitting area. This size is computed from the boundary positions of the segmentation area, which is stored in final cutout catalog (EUC_MER_FINAL-CUTOUTS-CAT_TILE*.fits), as:
sersic_size = max(dist((cutout[’DBL_CORNER_0_RA’],cutout[’DBL_CORNER_0_RA’]), cutout[’DBL_CORNER_1_RA’],cutout[’DBL_CORNER_1_RA’]), dist((cutout[’DBL_CORNER_1_RA’],cutout[’DBL_CORNER_1_RA’]), cutout[’DBL_CORNER_2_RA’],cutout[’DBL_CORNER_2_RA’]))
The computation strongly depends on the size of the fitting area, and it turned out that bright saturated stars, which yield a very large fitting area, need an extremely large computation time. Due to the saturation we do not expect reasonable results from the Sersic fitting, and we set the object size for source larger than 111” to the fix value of 5”, which of course completely degrades the results the fitting results for these source.
These sources have not been marked in the final catalogues.
Unfortunately it turns out that also the fitting area of some bright galaxies had been cut through this mechanism. We have identified 10 bright galaxies that suffer from this selection criteria, i.e., the results from their Sersic fitting are unreliable. The objects are:
| SOURCE_ID | name |
|---|---|
| -594244863489083553 | NGC1494 |
| -601220560490298837 | ESO201-014 |
| -621004880478969002 | NGC1527 |
| -622494400487269989 | ESO201-022 |
| -652863889482547324 | NGC1567 |
| -660878051475262694 | ESO202-015 |
| 2657204209664761230 | NGC6457 |
| 2658928450676035617 | NGC6463 |
| 2677813028655307424 | NGC6505 |
| 2760100549653060494 | UGC11230 |
We are working on discriminating such galaxies so that we can perform the fitting without truncating their size.
MER have been notified that some sources in the MER catalog have APhot errors that are several orders of magnitude larger than what they should be The problem is associated with regions with low coverage (one or two dithers). In these regions we have a significant fraction of masked pixels with 0 fluxes and high error values that are set to identify those pixels as masked. The APhot error calculation doesn’t take correctly into account those masked pixels and the result is that the high error values are included in the APhot error. (See issue #29081)
An analysis performed on the Fluxes of objects measured using aperture photometry of different aperture sizes shows some difference in the flux measured for points of same magnitude between the northern and southern hemispheres. It is possible that direct aperture measurements have different trends in different areas of the sky, because the sizes of the aperture are not constant but depend on the FWHM of the worst EXT band in the dataset - and the surveys of north and south hemispheres are different. We suggest to compute the total flux of the sources, using aperture fluxes only to obtain colors. (See issue #28907)
In Q1 data the columns FLUX_VIS_TO_[band]_TEMPLFIT are identical to FLUXERR_VIS_TO_[band]_TEMPLFIT for Y, J, H bands. This is due to a bug in the cataloguing routine of the MER pipeline. Note that the correct values in the catalog are the one stored in columns FLUX_*, while the not correct ones are FLUXERR_*. (See issue #28811)
A significant number of cluster members show colour offsets of 0.6 mag in g-r, when compared to Legacy Survey DR10 photometry. Those are certainly not faint galaxies as visible from the plot below. And their g-r colours in LS DR10 are fine, meaning consistent with the red sequence. The issue is evident in aperture photometry and even present, less strongly, in template fitting photometry measurements. The issue is related to the fact that the core pixels of the affected objects are erroneously flagged by the OU-EXT pipeline. The OU-EXT team investigated the issue and back-tracked the issue within their pipelines and fixed in OTF processing. See issue #29011