Abstract
Context . Astrometric calibration of mosaic CCDs can be performed either through a full field of view (FoV) joint solution (JS), which solves all CCDs simultaneously, or through a per CCD solution (PS), which calibrates each CCD independently. At present, the PS is the most commonly adopted approach, although it is not necessarily optimal. On the one hand, the projection point, plate scale, and distortion derived from the PS do not fully represent the true optical properties of the telescope, and their impact on astrometric precision remains unclear. On the other hand, accurate full-FoV joint pixel coordinates (JPCs), a unified coordinate system for all CCDs derived from the JS, are essential for subsequent data processing, such as full-FoV stellar flat-field fitting and sky background modeling. Aims . This work aims to develop a full-FoV joint astrometric calibration method for mosaic CCDs and assess whether it offers superior astrometric precision over the PS. Additionally, we seek to test the stability of JPC to validate the CCD mosaic geometry measured in the laboratory. Methods . We propose a joint astrometric calibration method for mosaic CCDs that employs an alternating minimization algorithm to iteratively solve for both the mosaic geometry between each CCD and the plate model. An accurate plate model is critical for the JS. We used simulations based on the typical distortion-corrected optical design to find the proper plate model. Results . Simulations show that even with only 0.1% distortion, conventional third- and fourth-order x, y polynomial models are insufficient for fitting the plate model. The JS plate model must include radial terms of the form r = x 2 + y 2 . We tested the proposed JS method using data from the Wide Field Survey Telescope (WFST). The results demonstrate that the JS achieves slightly better astrometric precision than the PS. After averaging the positional residuals over all the included exposures as a function of pixel position, an unmodeled distortion of ∼15 mas emerged, likely due to the discontinuity in the distortion pattern across the four-corner CCDs. The relative translation and rotation precisions between CCDs reach 0.056 pixels and 1.37, respectively. Moreover, CCD gap sizes derived from the JPCs are consistent with measurements from the laboratory. Conclusions . We developed a joint astrometric calibration method for mosaic CCDs. Our results indicate that radial terms are essential in the plate model for wide FoV telescopes with distortion-corrected design. The averaged 2D positional residual map reveals small-scale unmodeled discontinuity in the distortion pattern across CCDs, highlighting the complexity of high-precision astrometric calibration of mosaic CCDs. The mosaic geometry derived from the joint calibration aligns well with measurements from the laboratory, confirming the robustness and precision of our method. This work provides a new perspective on the astrometric calibration of mosaic CCDs.
Funding Information
  • National Natural Science Foundation of China (12321003)

This publication has 29 references indexed in Scilit: