ACIS CTI Plots

Charge Transfer In-efficiencies (CTIs) of ACIS CCDs are graudually changing with time, and it is essential to know how the degradation progresses. Since CTI are strongly depending on focal temperature, to see the degradation, it is better to remove the temperature dependency of CTI observations. MIT/ACIS team suggested that there is a (close to) linear relation between a focal plane temperature and CTI, and the CTI temperature dependency can be easily corrected (MIT ACIS Memo #204). Related memos and detail analysis can be found at ACIS CTI @-120C page maintained by Catherine E. Grant at MIT.

Mean CTI Plots

The following three plots show time trends of monthly averaged CTI of imaging CCDs, spectral CCDS, and backside illuminated CCDs computed using Mn K_α line. CTIs are averaged over four quadrants of each CCD. The CTI values used here are corrected for temperature dependency and detrended (see How CTI are computed?), except those for the backside illuminated CCDs. The latters are temeprature dependency corrected, but not detrended. To see a full size plot, please click the plot.

Imaging CCDs	Spectral CCDs
Back Side CCDs

Individual CCD/Quad CTI Plots

The following table lists links to individual CCD/Quad CTI trend plots computed using Mn K_α line. Each page shows five different CTI trend plots. From the top to the bottom, they are:

• Un-corrected cases with an integration time > 2000 sec
• The focal temperature ≤ -119.7 C cases with an integration time > 2000 sec
• The focal temperature ≤ -119.7 C with an integration time > 7000 sec
• Newly computed temperature dependency corrected cases with an integration time > 2000 sec
• MIT/ACIS temperature dependency corrected cases with an integration time > 2000 sec

The third case is the cleanest estimation, but we lose many data points, especially after year 2012. The fourth/fifth cases keep all data points, but they may introduce errors from temperature corrections. Two trending lines are fitted on each plot. The dividing point are slightly different from that of the combined data. For the imaging and the spectral CCDs, the division point is year 2011 and year 2009 for the backsdie CCDs.

CTI data are far more scattered after year 2012 and extreme data points are dropped before fitting the trending lines. These scatters were introduced because less and less time, the focal tempearture stays at or around 119.7C and fractuates widely.

Detrended CTI Plots

The following table shows the links to individual CCD/Quad plots similar to above, but the data are detrended by dtracning drop amp effect. See How CTI are computed? section for details. Backside Illuminated CCDs 5 and 7 are not detrended and not included.

CTI based on Al K_α and Ti K_α

Although they are not clean as CTI computed from Mn K_α line, we also computed CTI based on Al K_α and Ti K_α.

Detrended CTI Plots

Fitting Results

The following links show CTI change rate per year of each quadrant and those of averaged cases are listed.

None-detrended CTI

• Fitting Results: Int Time > 2000 sec
• Fitting Results: Int Time > 2000 sec & Temp <= -119.7 C
• Fitting Results: Int Time > 7000 sec & Temp <= -119.7C
• Fitting Results: Int Time > 2000 sec, MIT/ACIS Temp Factor Correction
• Fitting Results: Int Time > 2000 sec, Updated Temp Factor Correction

Detrended CTI

• Fitting Results: Int Time > 2000 sec
• Fitting Results: Int Time > 2000 sec & Temp <= -119.7 C
• Fitting Results: Int Time > 7000 sec & Temp <= -119.7C
• Fitting Results: Int Time > 2000 sec, MIT/ACIS Temp Factor Correction
• Fitting Results: Int Time > 2000 sec, Updated Temp Factor Correction

ASCII Data (Not Adjusted)

Correction Factors for Temperature Adjusted Plot

• For Al K alpha
• For Mn K alpha
• For Ti K alpha

CTIs are computed in the following steps:

• ACIS evt1 data file for CTI observation is extracted from Database
• Find the time span of the observation and extract the focal temperature data for that duration.
• Compute an average focal tempearture of that observation.
• If the integration time is more than 600 sec, compute CTI using MTA CTI computing tool.
• The extracted data are listed in ASCII data. The columns are:
  1. start time
  2. quad 0 CTI
  3. quad 1 CTI
  4. quad 2 CTI
  5. quad 3 CTI
  6. obsid
  7. end time
  8. averaged focal plane temperature
  9. standard deviation of the temperatre
  10. minimum temperatre of the period
  11. maximum temperatre of the period
  12. start time in sec from 1998.1.1
  13. end time in sec from 1998.1.1
• If the averged focal plane temperature is colder than or equal to -119.5 C, the data are used to create a plot without any adjustment.
• This data are further refined by selecting only data with the integration time longer than 7,000 seconds.
• If the focal plan temperature is warmer than -119.5 C. A correction is done for CTI. The correction factor is computed as follows:
  • Fit a straight line with time as an independent variable to the data with the focal plane temperature ≤ -119.7 C and the integration time longer than 7000 sec.
  • Use the fitted values to remove a time dependent CTI evolution.
  • Fit a straight line onto this data, but this time use the focal plane temperature as an independent variable. Set the correction at -119.7 C to zero.
  • The resulted The correction factors are in CTI chage per degree.
  • Using these correction factors, adjust all CTIs with focal temperature warmer than -119.7 C. Anything data colder than -119.7 C are used without the correction.
  You may want to read A Step by Step Example of Correcting Temperature Dependence of CTI for a better understanding.
• Similar plots with MIT/ACIS team corrections are also plotted. The correction factor for front side CCD is 0.036, and the backside CCD is 0.045.
• To detrace drop amp effect, detrend correction factors are computed
  • A corresponding acis stat fits file is extracted from database.
  • Selecte out drop amp data for CCD7, and compute an average of the drop amp
  • The average is devided by 680 to normalize the correction factor. the value 680 was computed by MIT/ACIS team (08/18/2000).
  • The correction factor is added to the CTI to remove drop amp effect.
  All data, except for CCD5 and CCD7, are detrended and plotted above.

Does The CTI Correction Factor Changes With Time?

How to Select Data?

A Step by Step Example of Correcting Temperature Dependence of CTI

If you have any questions about this page, contact swolk@cfa.harvard.edu.

Last Update: Sep 24, 2021.