GPC1 Static Masks

The static mask fraction for GPC1 is composed of three main sources.

Detector masks

The cell-by-cell masks remove problematic areas of the detectors. This includes the manually generated masks, flat and dark masks, regions with CTE issues, and cells that have been removed due to noise issues. This page shows the mask fraction from these masks, with the unvignetted region 0.097 masked and the full detector 0.147. Cells that are used for guide stars can also contribute to this fraction for a given exposure, as the pixels are flaggd as BLANK.

Inter-cell gaps

The inter-cell gaps are not included in the above numbers, but are counted directly by psastro when the cells are mosaicked into a chip. A quick calculation based on the sizes of chips (4846 x 4868) and cells (64 x 590 x 598) shows that the inter-cell gaps contribute 0.043 to the total mask fraction.

Inter-chip gaps

The interchip gaps are not directly counted by psastro, as psastro loops over the pixels contained in all chips. Because of this, the size of the interchip gap is only defined after the chips are assembled into an FPA. The FPA has a full astrometric solution, so the input chip pixels do not map one-to-one onto the FPA pixels. This complicates any contribution of interchip gaps to the fraction of masked chip pixels. The solution to this is to model the interchip gap in FPA pixels, and rescale that number to chip pixels.

Field of view

The interchip gap calculation requires a model of the detector extent in FPA coordinates. The location of the illuminated pixels can be output by psastro, and that set of pixels constrained by the detector field of view. The GPC1 FOV is specified as 3 degrees, which translates to a 20930 FPA pixel radius at a nominal plate scale of 0.258 arcsec/pixel. However, the detector is unvignetted to a 3.25 degree field of view, or a 22720 FPA pixel radius.

Calculation

Calculating separately for the two fields of view yield model detector sizes of MODEL_REF = 1 357 799 527 FPA pixels for the 3.0 degree FOV, and MODEL_MAX = 1 482 790 357 FPA pixels for the 3.25 degree FOV. Counting the number of pixels with values from the detector of FPA_REF = 1 258 523 504 pixels and FPA_MAX = 1 378 635 056 pixels. The difference gives the number of interchip FPA pixels: INTERCHIP_REF = 99 276 023, INTERCHIP_MAX = 104 155 301.

The number of chip pixels that contribute is CHIP_REF = 1 235 728 642 and CHIP_MAX = 1 353 439 062. The ratio of CHIP to FPA pixels gives a consistent scaling of 0.982 for both fields of view. Applying this to the interchip gap counts gives INTERCHIP_CHIP_REF = 97 477 897 and INTERCHIP_CHIP_MAX = 102 251 754. The contribution to the mask fraction is given by INTERCHIP_CHIP / (CHIP + INTERCHIP_CHIP), or about 0.07.

Median static mask

Based on the values listed above, the expected mask fraction is 0.097 + 0.043 + 0.07 = 0.21. This is somewhat of an overestimate of the loss of sky coverage for the specified 3.0 degree FOV, as the added pixels included in the MAX area outside this range are in effect a "bonus." The larger unvignetted FOV contributes {{{122 484 277}} pixels to the REF area. Assuming the same mask fraction for these pixels (0.21), this added area provides 7% of the reference FOV in usable pixels.

A 4x4 binned representation of the mask fraction is supplied below. The pixel values are the median number of pixels masked (ranging from 0 to 16) for a set of 12 ThreePi? exposures (chosen to ensure that the guide star cells would be randomly distributed).