4.1.1.1 CCD Subassembly Calibration Requirements

The CCD subassembly calibration measurements were selected via the following two-step process. First, scientific investigations which are most sensitive to ACIS calibration accuracy were identified. The corresponding AXAF observations were simulated using a parametric CCD response model, together with models of other components of AXAF. The parameters of the CCD response model were varied to determine the sensitivity of the scientific results to errors in the assumed CCD model parameters. The result of this first step was thus a set of constraints on the accuracy with which CCD model parameters must be determined.

The second step in the measurement selection process determined the accuracy with which various proposed laboratory measurements can constrain CCD response model parameters. This step required analysis and simulation of feasible laboratory measurements. In principle, the product of this step was a suite of laboratory measurements that calibrates ACIS CCDs with accuracy adequate to meet scientific objectives.

The parametric CCD response model clearly plays a central role in the selection of calibration measurements. The CCD response model used to derive the CCD subassembly calibration measurements specified here is described in ACIS memos PS-73 and PS-79 (Rasmussen).

Practical considerations complicate the selection of calibration measurements. Specific, readily identifiable scientific measurement requirements do not provide a complete specification of calibration accuracy requirements. More fundamentally, one could not, even in principle, anticipate all measurements that will be made with AXAF. For these reasons, it is useful to have a clear, if somewhat arbitrary, set of calibration objectives with which to guide calibration planning. The objectives adopted to plan the CCD subassembly calibration measurements described here are therefore stated, though not justified, as part of each measurement description. These objectives are consistent with the calibration goals identified in MSFC-2229; the justification of the objectives is beyond the scope of this document.

The complexities of laboratory measurements and the oversimplifications inherent in feasible simulations of those measurements must also be accounted for in calibration planning. To allow for these uncertainties, this plan called for more measurements than the absolute minimum number that simulations suggest was necessary. For example, simulations show that the required accuracy in quantum efficiency can be achieved with measurements at as few as four energies (see ACIS memo PS-72 (Isobe) ); this plan called for measurements at seven energies (see section  4.1.3 below.) The additional measurements allow for the evaluation of systematic errors in both the models and the measurements.