3.4.8 CCD health, charge injections and release

Author(s): Michael Davidson

An IDU task creates an independent CCD health calibration library for each 10 revolution interval following the model described in Section 3.3.3. This library contains information on the dark signal observed in each column of the SM, AF and XP CCDs. The automated validation asserts that the library is complete (i.e. that it has a full complement of device calibrations) and healthy (i.e. the library start time is valid and each fitted coefficient is within an acceptable range). The key conclusions on the CCD health for Gaia EDR3 are:

• •

  Only a relatively small number of columns have significant dark signal.

• •

  The effect of serial CTI can cause columns adjacent to a strong hot column to also be flagged as hot. This is considered acceptable as it will ensure that the majority of samples affected by the real hot column are masked.

• •

  The behaviour of hot columns varies; some are very stable in strength while others strengthen or weaken over time.

• •

  The effect of the temperature increase during decontaminations is clear.

• •

  Some new hot columns have been observed to appear since launch.

An IDU task creates an independent charge injection calibration library for each 1039 revolution interval following the model described in Section 3.3.3. This library contains information on the injections observed in each column of the AF and XP CCDs. The automated validation asserts that the library is complete (i.e. that it has a full complement of device calibrations) and healthy (i.e. the library start time is valid and each fitted coefficient is within an acceptable range). The key conclusions on the charge injections for Gaia EDR3 are:

• •

  Charge injection levels are extremely stable in time indicating that the relatively coarse time resolution on which the profiles are recalibrated is appropriate.

• •

  High amplitude variation in the profiles as a function of AC position is observed on both the largest spatial scales and smallest (column–to–column).

• •

  The number of injection response anomalies detected is very low (around 100). These are associated with deep traps, and defects in the injection structure and anti-blooming drains.

An IDU task creates an independent charge release calibration library for each 10 revolution interval following the model described in Section 3.3.3. This library contains information on the charge release profile observed in each column of the AF and XP CCDs. The automated validation asserts that the library is complete (i.e. that it has a full complement of device calibrations) and healthy (i.e. the library start time is valid and each fitted coefficient is within an acceptable range). The key conclusions on the charge release profiles for Gaia EDR3 are:

• •

  The release levels usually increase steadily as expected due to accumulation of radiation damage.

• •

  Large step-increases are observed during major solar events.

• •

  The RP devices have the largest levels of release due to their larger population of manufacturing traps.