Active Action Items
Draft paper to Mark by August 1, 2012 or as soon as he returns, whichever comes later; even if it's not complete – ms_v21.pdf emailed to Mark Aug 1.
Catherine work out an analytic model (what does all this mean?!?)- ALL think about MWB comment about frame-to-frame variations in background effect the FWHM; CI overwhelms the variation;
- Bev will plot and understand HE rejects for XIS vs time and COR; do BI first
- will probably not help for paper; use cumulative charge above 100 ADU instead
- Bev check whether she used COR or COR2; will try to get COR2 via ehk files
- Catherine continue to think about temperature dependence of CTI (finish writing up)
- Catherine (with Bev input) write text explanations of data processing to all figures in paper
- Catherine rename bibtex file to leaky bucket specific
- 2012-02-02 v12 of paper uses leakb.bib
- Eric maintain action items page
- Bev look bright Earth data O line w/ time to look for non-CTI gain changes
- decided it's the wrong energy, not sure about light leak, use Perseus instead
- Eric look at other sources (Perseus) to see if there is a 6 keV line to use for same
- Eric send Bev Perseus event lists for PHA/PI reconstruction
- Bev reconstruct Perseus PHA/PI values and send back to Eric
- 20120209 - check out the plots in the Data Products page
- come up with some metric for constraining non-CTI gain changes
- Bev fit central pixel PHAs for Perseus data
- like Catherine predicted, it's looking awfully hopeless (14feb12)
- actually, it's not looking hopeless after all and Bev will keep working on it (1mar12)
- still might be hopeless, but Bev believes it's possible to calculate CTI, but might not be interesting (might not constrain non-CTI gain changes); Bev will keep working on it (10apr12); perhaps we don't need this, summed pulseheight is better (31may12)
- Bev fit summed pulseheights in perseus data for two Y regions (~ 1/4 and 3/4 up from framestore), intercept of fit line should be ~ non-CTI gain change (actually non-CTI-in-IA change)
- Bev will overplot FI/BI, CI on/off intercepts on single plot, in terms of fractional change
- Bev will get linear fit to trend for CI on; this will be upper limit to non-CTI-in-IA change
- Bev recompute Gaussian fits for Catherine, but using COR2
- Bev check the COR2 = 11,12,13 fits visually to see if they are acceptable chi-by-eye-wise
- Eric will figure out the COR2 value during the XIS raw frame observations
- 20120716 – COR2 = 12.0-14.1 during the 10 raw frames; frame 2 is the one in the paper, and that has COR2=12.8
- Eric will try to estimate the error in CTI/FWHM values in COR bins (Bev will send him spectra)
- IDL gets about the correct uncertainties if weighted by 1/cts
- Bev will do this for all of the COR2 bins
- Bev will also verify that COR2s are correct from dp10
- Catherine put errorbars on COR2 figures, or say that they're small
- all think about why FWHM at COR2 10-14 are significantly different
- Bev will count particle events on raw frames (or rather pixels above some threshold that she will choose)
- Bev showed cumulative charge vs. threshold; will determine value for 100 ADU threshold for all frames, and that will give estimate of how BG varies
- Bev will make plots of spectra around Mn Kalpha for XIS3, SCI off and on, 2006-10 to 2006-11
- Bev make example spectra for XIS BI, ACIS BI early in mission on one plot (or whatever she thinks is good). don't overlay fits. try on same plot or as multiple panels. use energy as x-axis, range 1-10 keV, linear X and Y axes.
- Catherine look at Tawa et al 2008 paper to see time- (COR-) dependence of XIS background
- Eric see if there is change in XIS BG with time on longer baseline than Tawa, modulo COR2
all look at Fig 8 and think about interesting features to point out/annotate (rocket ships, framestore wackiness, etc.)--wait to see if this is appropriate in context of final text- Catherine continue to write
- still need a general reference for radiation damage in CCDs
- "gain" decrease could include framestore CTI, confess
- use percent or fraction to express CTI and gain changes (not ADU or eV) (2012-07-27)
- more warnings, not representative of user data, unprocessed/uncorrected
- update plots with error bars when/if Eric comes up with something better
- Catherine has incorporated comments from Bev and Eric through 5/22/12
- Bev and Eric read what Catherine has written so far and comment, suggest places that need references
Completed Action Items
- ALL have an outline for Mark
- outline is sent to Mark (2011-01-28)
- here is a link to an attachment: Outline PDF
- 20100210--comments from Mark here
- Eric combine outlines
- Eric convert action items page
- ALL decide on common CTI metric (which grades, which pixels in island, which rows)
- use center pixels
- XIS cal source regions-
- both use all telemetered grades (minus 255 for XIS, minus 5 Bev grades for ACIS, S3 only uses g02)
- Bev will try restricting grades, esp. for BI
- 2011-05-05 decision to not pursue CTI metric for XIS
- Bev make CTI plots for XIS upper rows (1 month bins)
- 20110210Bev has IDL script set up, pixel fits (center pixel pulseheight and trailing pixel) are working for all detectors/quads, CTI calculation is not working completely
- Catherine replicate same plots for ACIS, pending Bev's results
- 2011-05-05 decision to not pursue CTI metric for XIS
- Catherine replicate existing Bev plot (monthly plots of FWHM, summed pulseheight centroid)
- 20110210--CEG has framework set up, waiting on flaky disk
- 20110217--CEG posted plots to Data Products
- 20110407--CEG made plots for IACHEC, For paper, want just combined plots, peaks (eV)/5894 eV – same y-range, widths in FWHM (eV)
- Eric add space for Leaky Buckets presentations (posters, talks slides) to Publications page (20110620--CEG)
- 20110407--CEG a few slides for IACHEC, summarizing our goals and what we're doing IACHEC slides
- We should talk about the mechanics of writing the paper. Does Eric keep a master copy and we send text and revisions to him? Do we post a master copy here that we can check-out to edit ourselves?
- Eric will keep master version of tex file, update wiki on occasion
- Bev and Catherine can borrow tex tarball for big jobs
- Catherine's Suzaku poster wishlist:
- Peak & FWHM vs COR for early (CI off) and late (CI on) for XIS1 and XIS3 (Thank you Bev!)
- Catherine send to Bev copy of her line fitting (20110607, and the forgotten bits on 20110620)
- Catherine show that line centroid is good proxy for CTI (20110621--CEG posted to Data Products)
- then we can just use that and not worry about XIS CTI metric
- Catherine spot check line fits (20110621--CEG convinced that results are consistent between different bkg levels and src count rates)
- Catherine convince herself that ACIS QE drop goes away with CTI correction (20110622--mostly goes away. QE drop goes from ~12% to ~3% over 11 years)
- Catherine find the cut-and-paste text from previous ACIS papers, send to Eric (20110705--sent some text from SPIE2007 paper)
- Catherine make half-life plots
- 20110324--CEG posted plots to Data Products
- Bev find the cut-and-paste text from previous XIS papers, send to Eric
- Catherine make XIS half-life plots using Bev's data which use Catherine's fitting method (20111018)
- 20111028--CEG doesn't make much difference, XIS1 line flux is still weird
- Bev make new plots of line center and FWHM vs. COR using Catherine's fitting method
- from 2008 paper, will look for script that made them and verify that it's still flat; revisit with non-XIS2 devices
- 20110504--bev posted xis1 & xis3 plots for 201101 to Data Products
- Catherine will think about equivalent for ACIS, flat FWHM, but are FI wiggles due to background wiggles? compile bkg data in same time bins
- 20110816--actually confirmed this a while ago and forgot to update, the FI wiggles are not correlated with the background, ACIS FWHM is not strongly sensitive to sacrificial charge
- Eric begin latex document for A&A (downloaded A&A template)
- Eric look for pictures of XIS for schematic
- Eric look into A&A photo and color policies
- color figures are E250 each for 1 or 2, E180 each for 3 or more; should be in CMYK, not RGB; online (PDF?) color figures are free, but will be greyscaled in print version
- page charges are E100 per printed page
- no specific photo policy
- Eric look into figure format (titles, fonts, etc.) policy
- no specific policy about title, but looking at several papers, none of the figures have them
- no specific policy about fonts, we should just be consistent
- ALL think about whether to include trailing charge info for both instruments
- not very fruitful; XIS is limited to the top of the array, so hard to separate trailing charge from normal event splitting; also SCI will complicate this
- Catherine. Look at Bev's plots (http://space.mit.edu.ezproxyberklee.flo.org/XIS/monitor/ccdperf), get a sense how ACIS would look on the same kind of plot. Qualitatively, how much does trailing charge matter. (20111031--CEG, plots on data products page)
- Bev make gain/FWHM vs. COR plots for XIS3, Oct+Nov 2006 (20111028--done)
- see if it's as good as XIS2 previous plots
- send data to Catherine for paper figures (20111028--done, figures to be added to tar ball shortly)
- Catherine change fig captions for gain/FWHM vs. COR; data are for ~ 1 month (or 2) straddling CI turning on, and for XIS2 (or XIS3)
- Catherine send paper appropriate figures to Eric (20111031--CEG In progress. See Data Products for new versions and one new one) and recipe to bev.
- bev will give catherine pre-launch CTI values for xis.
- from http://space.mit.edu.ezproxyberklee.flo.org/XIS/about/lamarr_spie_5501_385.pdf Figure 4; BI (8c2) = 0.55, FI = 0.30-0.35
- Eric make schematic diagrams (20111121--EDM, see Data Products Item 9)
- Eric edit ACIS schematic to make cal source regions squares (for XIS too?)
- Eric add CI rows and coord axes and move S3 aimpoint
- Eric fix crappy lines and circles (20111220--EDM done)
- Bev confirm that ACIS and XIS are using the same high energy reject metric; time units and spatial area? what is being rejected on-board?
- XIS uses only cal source area, ~ 9-10% of full 1024x1024
- Bev rejects summed event PHA of >= 3750 on ground to match ACIS rejection; but some single pixels are rejected on board XIS, possibly below 3750
- seems no way to compare XIS rejection to ACIS
- also dependence on input energy spectrum
- Bev will look for XIS frame data to see how well the background metric matches that of ACIS
- Bev will grab raw frames for the paper to show background, and bias-correct; XIS BI & FI, CI on & off, ACIS BI & FI
- Bev add labels to colorbar (horizontal), add frames around each FOV, fix ACIS bias levels
- Bev will upload final frame EPS files and colorbar to wiki
- Eric choose best Bev frame images
- Eric help Catherine with latex layout of raw frames and colorbar
- Catherine change A4 formatting to Letter in dvips
- 2012-01-10 probably actually an A&A issue
- Eric tell Bev which raw frames to annotate
- Bev add "ACIS FI" etc. to upper right corner of the four raw frame PS files
- Bev search for reference on XIS Lesser process for thinning BI
- Section 4 of Mark's 2004 spie paper (available locally at http://space.mit.edu.ezproxyberklee.flo.org/XIS/about/bautz_5501_14_4june.pdf) discusses this improvement. He references M. Lesser and V. Iyer, “Enhancing the back illuminated performance of astronomical CCDs,” Proc. SPIE 3355, pp. 446--456, 1998. B. E. Burke, J. A. Gregory, A. H. Loomis, M. Lesser, M. W. Bautz, S. E. Kissel, D. D. Rathman, R. M. Osgood, M. J. Cooper, T. A. Lind, and G. R. Ricker, “CCD soft-X-ray detectors with improved high- and low-energy performance,” IEEE Trans. Nuclear Science in press, 2004.
- Catherine add comment in paper or Table 1 about different BI processes. 2012-02-01 added comment in v11. it's pretty vague but it's there, still needs an XIS reference
- Catherine add paragraph to end of Section 2.2? on how we're only analyzing parallel CTI, serial CTI (define them) is low except for ACIS BI, because of different BI process; therefore we're not talking about serial CTI cause it hasn't changed
- 2012-02-01 added to Sec. 2.1 in v11
