The High Resolution Camera (HRC) detectors are occasionally active while they are in a stowed position. Data taken during these active intervals can be used to study HRC backgrounds. Most "events" that HRC sees during the stowed position are cosmic rays which are missed by the anti-coincidence shield, and hence, we can also use the stowed data to study changes in the anti-coincidence shield efficiency.
In this study, we address following points:
HRC level 0 event data which satisfy all of the following criteria are extracted from archive.
The level 0 data are processed with the standard HRC function, hrc_process_events, but with no aspect solution applied.
Extracted data are categorized into:
If you are interested in a more detailed description of the selection criteria, please check: Detailed Description of Data Selections and Reduction.
When the HRC is in the stowed position it is usually in the next-in-line telemetry format; the amount of data telemetered to the Earth is very limited, and the data are usually telemetry-saturated. However, the HRC also reports three different count rates. The first rate is the total rate, which is the count rate of all MCP events (X-rays and high-energy particles). The second rate is the valid rate, which counts the fraction of the "total" events that pass the on-board event validity tests (non-vetoed rate). The last rate is the shield rate, events due to high-energy particles counted by anti-coincidence shield, Since when the HRC is in the stowed position, the MCP rates (total and valid) are dominated by high-energy particles, we expect the three rates should closely follow the Electron Proton Helium Instrument (EPHIN) Integral channel flux (see Comparison with EPHIN Integral Channel Flux below, or a memo 'Raw HRC Antico Rate v EPHIN Flux' by M. Juda).
Temporal variations of the
HRC-I valid, total, and shield rates, as well as
those of the EPHIN Integral channel flux and
the valid/total rate ratio are shown in figure 1a
Temporal variation in the event PHA peak for the HRC-I are shown in
Figure 1b.
The fitted lines in Figure 1b are computed using
a robust method on the data with an anti-coincidence shield PMT HV
output step level
Because of the modulation of the cosmic-ray flux due to the solar-activity cycle (see EPHIN Integral Channel PHAs), HRC anti-co shield and MCP valid and total count rates change non-linearly with time, and hence, we did not fit a line. The value of the EPHIN integral channel flux from the beginning of 2006 onward is artificially high due to the EPHIN running in a current-limited state.
Figure 1a: Time Evolution of MCP Valid Count Rate, Total Count Rate, and the ratio of Valid Count Rate/Total Count Rate
Figure 1b: Time Evolution of PHA Peak Median and PHA Peak Width
Table 1 shows links to data tables and plots of all cases.
Table 1: Data and Plots
| Detector | Data | Plot |
|---|---|---|
| HRC I RANGE_SWITCH_LEVEL=115 | Data | Plot |
| HRC S RANGE_SWITCH_LEVEL=125 Section 1 | Data | Plot |
| HRC S RANGE_SWITCH_LEVEL=125 Section 2 | Data | Plot |
| HRC S RANGE_SWITCH_LEVEL=125 Section 3 | Data | Plot |
| HRC S RANGE_SWITCH_LEVEL=125, high precision timing | Data | Plot |
The table (Table 2,) below shows intercepts and slopes estimated by robust method. Again, the error is computed by Bootstrapping (sample size = 100), and the data used are only with S2HVST = 8. The zero point for the intercept is set to the start of Year 2000 rather than Year 0.
The statistics of valid/total ratio of the three sections of HRC-S are identical, since the valid rate and total rate are same for all three (section 3 contains U = 0, V = 0). Note, the errors for the HRC-S slopes are slightly different, since they are computed by Bootstrap method.
All PHA peaks also evolve with time. Interestingly, PHA peak width of HRC-S high precision case increases with time, even though the peak position is shifting downward. Note, for a peak position, the median is a better indicator than the mean. This is because the number of PHA channel is limited, and the upper end of the distribution could be saturated.
Table 2: Robust Fitting Results against Time (zero point is set at Year 2000)
| HRC-I | HRC-S | HRC-S High Precision | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Sec 1 | Sec 2 | Sec 3 | ||||||||
| Int | Slope | Int | Slope | Int | Slope | Int | Slope | Int | Slope | |
| Valid/Total Mean | -23.2201 | 0.01173+/-0.00031 | -2.3371 | 0.00139+/-0.00007 | -2.3371 | 0.00139+/-0.00007 | -2.3371 | 0.00139+/-0.00007 | -2.2307 | 0.00160+/-0.00008 |
| Valid/Total Median | -23.1852 | 0.01171+/-0.00044 | -2.4749 | 0.00146+/-0.00007 | -2.4749 | 0.00146+/-0.00006 | -2.4749 | 0.00146+/-0.00008 | -1.8709 | 0.00142+/-0.00004 |
| PHA Peak Mean | 2263.099 | -1.067+/-0.055 | 5884.753 | -2.864+/-0.031 | 5157.454 | -2.496+/-0.037 | 5736.369 | -2.783+/-0.033 | 6041.942 | -2.929+/-0.123 |
| PHA Peak Median | 3135.530 | -1.511+/-0.058 | 8394.269 | -4.121+/-0.067 | 7508.641 | -3.674+/-0.047 | 9133.356 | -4.479+/-0.097 | 10594.835 | -5.201+/-0.121 |
| PHA Peak Width | 1387.5990 | -0.6510+/-0.0082 | 1091.5088 | -0.5052+/-0.0106 | 531.1949 | -0.2269+/-0.0141 | 674.4635 | -0.2980+/-0.0089 | -171.0422 | 0.1224+/-0.0244 |
From mid March 2003 until early November 2004, the HRC anti-coincidence shield was operated at a reduced HV level for most of the time when the HRC was not observing the sky. The HRC anti-coincidence PMT HV level (output read-back S2HVLV) is controlled via commanding its step level, S2HVST. For example, S2HVST = 8 corresponds to a HV read-back level of 83 ADU or 1700V (see HRC PMT HV Codes).
The HRC-I MCP "total" rate is not affected by the different anti-coincidence shield PMT HV settings; there is no mechanism for this to happen. However, the HRC-I MCP valid rate will change with step level as the response of the PMT is lowered with lower HV step level resulting in a lower veto rate and a larger valid/total ratio. The linear fit to the valid/total ratio is indicative of a change in the anti-co shield response - the increase in the ratio with time reflects a degradation in the efficiency of the anti-coincidence detector. This decreasing efficiency could be due to lowered performance of the PMT or due to degradation of the optical qualities of the plastic scintillater or joints between sections or to the PMT. Interestingly, there is also an affect on the event PHAs with the reduced PMT HV step; there is a tendency for a lower mean and narrower distribution (see figure 1b).
Figure 2 shows a plot of Valid/Total ratio against S2HVST. In order to generate this plot, the observed time trend seen in Figure 1 was removed first using a fitted line given in that figure. Because data are dominated by two S2HVST (5 and 8) points, the fitted line may not be accurate; still there is a clear trend. This linear relation might be used to determine a candidate update to the PMT HV step level to restore the Valid/Total ratio back to the start of the mission level in the future.
Figure 2: Mean Valid/Total Count Ratio against S2HVST
Since the anti-coincidence shield does not veto events from the HRC-S, there is no effect from different PMT HV setting.
Valid and total count rates seem to modulate with the EPHIN integral channel flux. This does make sense, since all are affected by cosmic rays. Figures 3a and 3b show scatter diagrams of the valid and total count rates against EPHIN integral channel flux. There seems to be two components in these diagrams. The first panel shows all data points with two regression lines fitted: one for the data set with EPHIN flux >0.3, and the other with <= 0.3. The second panel shows just between EPHIN flux between 0.1 and 0.3. The correlation statistics are significant for both data sets (p< 0.01%). Two component elements are more prominent in HRC-S (see Plots).
One thing to comment on with regard to the higher EPHIN Integral channel flux is that there are two mechanisms involved. One is real higher fluxes above a nominal quiescent level due to solar activity, and the other is "false" higher flux readings due to the anomalous EPHIN +27V operation. The anomalous behavior started in November 2003 and became a near constant presence after December 2005. The first of these effects can lead to a change in the MCP Total (or Valid) rate vs EPHIN flux due to a change in the particle spectrum, with less penetrating particles contributing to the elevated EPHIN flux but not to the MCP rate. In the second case the "false" Integral channel events are not expected to generate MCP events.
Figure 3a: Valid Count Rate against EPHIN Integral Channel Flux
Figure 3b: Total Count Rate against EPHIN Integral Channel Flux
Figure 3c shows Valid and Total Count Rates against Shield
Count Rates. Two solid lines fitted to the data are for S2HVST = 8 and 5 cases.
Although the lines are fitted with robust regression method, we dropped a few outliers.
The component along the dotted line is those with EPHIN flux < 0.25 which appeared as
a secondary component in Figure 3a and b.
Although there is a strong correlation between EPHIN Flux and the valid count rate, this component shows
only a weak correlation between the shield rate and the valid count rate.
Figure 3c: Valid and Total Count Rates against Shield Count Rates
Although the isotropic flux of high-energy particles might lead one to expect the background to be spatially uniform, shielding by spacecraft and detector structures can be expected to produce some variations in the spatial distribution. We created cumulative background maps to study the variations.
To create the maps, we first combined all the evt1 files for the individual time intervals (the resulting level event file). The data were binned by 256x256 pixels to create an image fits file. We apply a dead-time correction to the data to account for the effects of telemetry-saturation. The dead-time correction increases the count rates about 15 times of the telemetered count rates. The image is normalized to per sec per pixel. Figure 4a shows the HRC-I background map ( fits file ). The maps shown below had status bit filtering applied to their input event lists, which removes "bad" events from the data (for maps without the status bit filtering, please check Time Evolutions section). The red color indicates higher counts and the blue indicates lower counts. For this background map, we also corrected with an instrument map. For more details, please see instrument map correction page.
Figure 4b shows the distribution of count rate per sec per pixel area. Although there are many pixels without any counts, these are dropped from the plots.
Figure 4a: HRC I Cumulative Background Map
Figure 4b: HRC I Count Rate Distribution
Figure 5a shows the background maps of HRC S ( fits file section 1, fits file section 2, fits file section 3 ), and Figure 5b shows the distribution of the count rate per sec per pixel. The right panel has the origin of the coordinates. The dynamic range is between 1.5e-6 and 4.0e-6 (count/sec/pix).
Figure 5a: HRC S Cumulative Background Map
Figure 5b: HRC S Count Rate Distribution
Figure 6a: HRC S High Precision Timing Mode Cumulative Background Map
Figure 6b: HRC S High Precision Timing Mode Count Rate Distribution
The following links will bring up pages showing yearly background maps, similar to these shown above.
The background maps are created as follows:
As we see Figure 1, the count rates and peak widths increased with year (except HRC-S high precision case). HRC-I (status bit filtered) shows more counts at the lower left corner. This is because the instrument opening is located in that direction and more particles could reach the detector.
The stowed background maps created each year can be used to correct HRC I observations. We show a step by step procedure in: How to Create a Background Map for an Observation page.
If you have any questions about this page, please contact: tisobe@cfa.harvard.edu