Old calibration

NOTE The following old calibration section contains maps made with an incorrect relative weighting. They do not represent the final APEX-SZ cluster maps.

Filter Steps
szvar.tsfilter{end+1} = {'flag_demod_rails',7000};
szvar.tsfilter{end+1} = {'flag_glitches',6.0,5};
szvar.tsfilter{end+1} = {'flag_finalcuts',1,1,0.5,1};
szvar.tsfilter{end+1} = {'flag_noisychannels',3e-12,30e-12};
szvar.tsfilter{end+1} = {'flag_finalcuts',0.5,0.5,1,0.9};
szvar.tsfilter{end+1} = {'medianflatfield',1};
szvar.tsfilter{end+1} = {'flag_lowflatfield',3.};
szvar.tsfilter{end+1} = {'medianflatfield',1};
szvar.tsfilter{end+1} = {'polyrm',3,1};
szvar.tsfilter{end+1} = {'flag_rms',5,2};
szvar.tsfilter{end+1} = {'removecommonmode',1};
szvar.tsfilter{end+1} = {'flag_rms',5,2};
szvar.tsfilter{end+1} = {'removemediannoise',0.9,3};
szvar.tsfilter{end+1} = {'flag_rms',5,2};

Filter Steps
szvar.tsfilter{end+1} = {'flag_demod_rails',7000};
szvar.tsfilter{end+1} = {'flag_glitches',6.0,5};
szvar.tsfilter{end+1} = {'flag_finalcuts',1,1,0.5,1};
szvar.tsfilter{end+1} = {'flag_noisychannels',3e-12,30e-12};
szvar.tsfilter{end+1} = {'flag_finalcuts',0.5,0.5,1,0.9};
szvar.tsfilter{end+1} = {'medianflatfield',1};
szvar.tsfilter{end+1} = {'flag_lowflatfield',3.};
szvar.tsfilter{end+1} = {'medianflatfield',1};
szvar.tsfilter{end+1} = {'circle_template_removal',-1,1};
szvar.tsfilter{end+1} = {'flag_rms',5,2};
szvar.tsfilter{end+1} = {'removecommonmode',1};
szvar.tsfilter{end+1} = {'flag_rms',5,2};
szvar.tsfilter{end+1} = {'removemediannoise',0.9,3};
szvar.tsfilter{end+1} = {'flag_rms',5,2};

New Calibration Test

We've made a calibrated map of the HII region rcw38. From this calibrated source, we derive calibration data on days in which Mars is not observable.
To test this method, we've used days in which we have full beammaps of a planet and RCW38. We then compare the RCW38 derived calibrations with that derived from the planet scan.

*APRIL2007*

*MAY2008*

*APRIL2009*

*Note: RCW38 is calibrated off the WMAP/RUDY Mars model. This temperature is ~4% less than the JCMT FLUXES Mars Temperature. Since no RUDY model is available for Saturn, we use the JCMT FLUXES program and expect a ~4% offset in calibration.

There was one day in which the RCW38 calibration produced strange calibration factors for a small section of the array.

Red is the rcw38 calibration data and Blue is the Saturn calibration data. The low calibration factors produced extremely low rms K rj maps which dominated in the sample variance weighting coadd. The cause of the strange calibration factors is currently unknown, so for the following maps, the Saturn calibration data is used.

New Calibration (source masked)

szvar.tsfilter{end+1} = {'flag_source',2};
szvar.tsfilter{end+1} = {'flag_opticallydead',0.08};
szvar.tsfilter{end+1} = {'flag_demod_rails',7000};
szvar.tsfilter{end+1} = {'flag_glitches',6.0,10};
szvar.tsfilter{end+1} = {'flag_finalcuts',1,1,0.5,1};
szvar.tsfilter{end+1} = {'flag_noisychannels',3e-12,30e-12};
szvar.tsfilter{end+1} = {'flag_finalcuts',0.5,0.5,1,0.9};
szvar.tsfilter{end+1} = {'medianflatfield',1};
szvar.tsfilter{end+1} = {'flag_lowflatfield',3.};
szvar.tsfilter{end+1} = {'medianflatfield',1};
szvar.tsfilter{end+1} = {'circle_template_removal',-1,1};
szvar.tsfilter{end+1} = {'flag_rms',5,2};
szvar.tsfilter{end+1} = {'polyrm',4,1};
szvar.tsfilter{end+1} = {'removespatialtemplate',1,0,0,1};
szvar.tsfilter{end+1} = {'flag_rms',5,2};
szvar.tsfilter{end+1} = {'removespatialtemplate',1,0,0,1};
szvar.tsfilter{end+1} = {'flag_rms',5,2};
szvar.tsfilter{end+1} = {'polyrm',2,1};
szvar.tsfilter{end+1} = {'flag_rms',5,2};
szvar.tsfilter{end+1} = {'removespatialtemplate',1,0,0,1};

* 20090917fsignal_to_noise_17f.png:

Source Flagged Regions

To see if the source-flagging was introducing any cluster like signals in the map, I made a number of different maps, each with a different source-flagged region. It appears the source-flagging is introducing the cluster signal. The source-flagging has been applied before the glitch-flagging. This could be keeping glitches in the source region. Alternatively, the polysecant fit could be blowing up at the edges of the source flagged region. We've seen this at the beginning/end of timestreams. Both possibilities are being investigated.

-- JamesKennedy - 06 Oct 2009

---

This topic: APEX_SZ > WebHome > XMMJ1230 > OldCalibrationAndSourceMaskingStuff Topic revision: r1 - 2009-10-06 - JamesKennedy

---