Event Rate Survey and Background Study
of the NCT‟09 Flight Data
Jeng-Lun (Alan) Chiu
Institute of Astronomy, NTHU
2011/06/27 HEAG Group Meeting @ NTHU
Outline
< Introduction >
• Introduction & Current Status
• Overview of NCT‟09 Flight Data
< NCT’09 Detailed Data examination and System Check >
• Unexpected Increase of Event Rates & Hz-feature
• Variation of Event Rates
< Data Analysis >
• Spectrum Study of Flight Background
J.-L. Chiu (Institute of Astronomy, NTHU)
NCT – Instrument in 2009 Flight
J.-L. Chiu (Institute of Astronomy, NTHU)
Germanium Detector Array
Instrument Cradle
Instrument Gondola
Table 2. Summary of the NCT stratospheric balloon flights.
Launch Date Launch Location Primary Target Duration * GeDs References
June 01, 2005 Ft. Sumner, NM, USA Background (o) 5.5 / (t) 8.5 hrs 2 [8,9] **
May 17, 2009 Ft. Sumner, NM, USA Crab Nebula (o) 22 / (t) 38.5 hrs 10 [7,10,11]
April 29, 2010 Alice Springs, Australia Galactic Center Crashed / (e) 48-96 hrs 10 [7,12,13]
Note: * (o) = operational, (t) = total, (e) = expected durations ; ** NCT Taiwan team had not yet attended the collaboration before 2006.
Energy Range 0.2-10 MeV
Spectral Resolution 0.2-1% FWHM
Field of View (FoV) ~40º×70º FWHM
Angular Resolution ~5º FWHM (662 keV)
Effective Area * 5-18 cm2
Narrow Line Sensitivity (3σ) (1-2) ×10-5 γ cm-2 s-1
Continuum Sensitivity (3σ) (1/E) ×10-4 cm-2 s-1 MeV-1
100% Polarization Sensitivity 37 mCrab (0.2-0.5 MeV)
NCT – Characteristics & Flights
• NCT’05: Despite a failure of a servo system in the azimuthal pointing system, the
6-hour prototype flight allowed detailed characterization of the background at float
altitudes [J. D. Bowen, 2007], where observed background data were compared
with simulations at passing altitudes.
• NCT’09: The Crab Nebula was detected at a significance of 4σ in this flight [M. S.
Bandstra, 2011]. Figure (in next silde) shows the Crab image from 25 ks of data,
using 2-site events from 200–1500 keV. There are 49,000 photons in the back-
projection, where 700 are expected from the Crab. Five iterations of an image
reconstruction algorithm (i.e. MLEM) were applied to the image. It‟s the first
reported detection of an astrophysical source by a Compact Compton
Telescope (CCT). Besides, a very promising potential was also shown to detect
the Crab polarization [J.-S. Liang, 2010].
• Calibrations: Intensive ground calibrations from NCT’09 and NCT’10 (before
flight attempt) have been carried out and analyzed [J.-L. Chiu, 2010 ; E. C. Bellm,
2011] to give a better understanding of the detector response functions such as
the effective area and the polarization performance.
NCT – Scientific Results
J.-L. Chiu (Institute of Astronomy, NTHU)
NCT’09 Imaging – Crab Image (ep.02)
Methods to obtain Crab image (by Andreas Zoglauer): (1) Use only 2-site events (for some unknown reason 3+ are not working as good) (2) Use the Klein-Nishina-times-Photo-Absorption event reconstruction approach (the one we usually use for HEMI). (3) Use only events with Compton scatter angles maximum half as large as the distance down to the horizon (I do the
same for COMPTEL data)
The first reported detection of an astrophysical source by a CCT
J.-L. Chiu (Institute of Astronomy, NTHU)
NCT’09 Imaging – Significance of Crab Detection
(M. S. Bandstra, ApJ, in press (accepted June 1, 2011)) :
1. The Crab was detected at a significance of at least 4σ by NCT during the
2009 balloon flight.
2. Using image processing, the Crab appears in a Compton image made
from the data.
3. Examination of the spectrum and ARM histograms from the source are
consistent with the simulations of the Crab.
4. This result is the first significant detection of a celestial source by a CCT
and is an important step in establishing the viability of the compact Compton
telescope design for future space-based wide-survey instruments.
J.-L. Chiu (Institute of Astronomy, NTHU)
• NCT gondola was supposed to be launched on April 29, 2010 from Alice Springs Airport for a „> 28-hour‟ flight, which could allow us at least a whole period of Galactic-center observation.
• However, it crashed during launch (ps. fortunately without hurting any person).
• The NCT instrument was recovered on April 29 and then shipping back to Berkeley on July 8, 2010.
• The cryostat was initially checked to be in a fine condition due to the protection of shields and cradle, while the rest of devices were damaged to different levels by visual inspection.
• Functionality tests were carried out to verify the status of the GeDs (ps. some need reprocessing), the BGO shields (ps. need further test for 4 of the 36 channels), and the readout electronics (ps. working fine, need minor replacements).
• A two volume set of investigation report was published by NASA on October 22, 2010. Several recommendations were made to the Balloon Program Office about the standard operating process, where some parts are arguable. More details can be checked in the report online ( http://www.nasa.gov/centers/goddard/business/foia/balloon_mishap.html ).
NCT - Current Status of the Project
J.-L. Chiu (Institute of Astronomy, NTHU)
• The instrument is planned to be reconstructed with a new configuration (with only 3 layers), which will enhance the efficiency of polarization measurement and hard X-ray imaging. The next balloon launch is planned for Spring 2013.
Energy & Position Measurements
• Energy measurement: Unipolar shaper, 6μs time-to-peak
• Timing measurement: bipolar shaper, 170ns
• X and Y positions: determined by orthogonal strips
Strip pitch: 2 mm (0.25 mm gap)
• Z position: calculated by time difference between X & Y strips
Preliminary (depth) resolution: 0.6 mm
- +
Depth
eh
Edge-On View mirror
X
YZ
J.-L. Chiu (Institute of Astronomy, NTHU)
< Estimate of Loss and Rates >
• EvtID loss: Sum of the discontinuity on event ID assigned in each CC.
• Frame loss: Sum of the discontinuity on frame ID assigned by flight
computer for all frames but GCUHK.
• Frame/Transmission/Event rates are obtained by dividing relevant counters into increment of GCU/GPS time recorded in the flash disk. If
there’s any GCUHK frame-loss, rates will be over-estimated.
Event types:
-Total E0AE: events with SYNC word received. (may lose the rest) -Valid E0AE: events with SYNC word and group header. -Valid events: Number of events with either timing or ADC data.
-E&T event: Events with both ADC and time data.-LLD only event: Events with ADC data only.-FT only event: Events with time data only.
-No-data event: Events with no data but only headers.
“Event” means “trigger” due to “single interaction”
< Data Structure & Event types in GSE >
(Definitions)
Clock Increment History (UHF Clock)
07456
27772
29336
46184
65141
02552-0
256
4: 0
9-0
5-1
7 0
8-2
4-3
1
CC
4S
ys-e
rr (
3 C
C r
ese
ts)
18432-1
8448: 09-0
5-1
7 1
4-3
7-4
7
CC
4S
ys-e
rr (
4)
21400-2
142
4: 0
9-0
5-1
7 1
5-4
8-0
4
CC
4S
ys-e
rr (
5)
24904-2
492
4: 0
9-0
5-1
7 1
7-1
0-1
3
CC
4S
ys-e
rr (
5)
32520-3
253
2: 0
9-0
5-1
7 2
0-0
9-0
1
CC
0S
ys-e
rr (
3)
30732-3
074
4: 0
9-0
5-1
7 1
9-2
7-1
3
CC
2S
ys-e
rr (
3)
44832-4
485
6: 0
9-0
5-1
8 0
0-5
6-1
8
CC
4S
ys-e
rr (
5)
(~ 6
.6 h
r)
46191-4
868
7: <
09-0
5-1
8 0
7-1
4-1
5
Syste
m W
arm
-up
(3)
54
10
7: 0
9-0
5-1
8 0
9-2
0-2
8 (
1)
Re
bo
ot a
fte
r S
oft
-ve
to s
hie
ld m
ode
55
38
7: 0
9-0
5-1
8 0
9-5
0-2
2 (
1)
Re
bo
ot a
fte
r S
oft
-ve
to s
hie
ld m
od
e
61
83
7: 0
9-0
5-1
8 1
2-0
4-2
8 (
1)
Re
bo
ot fo
r p
ow
er
an
om
aly
20984-21440:
dGPS anomaly
3800:
1st GCU
Time
Reset
34052:
2nd
GCU
Time
Reset
58915:
3rd GCU
Time
Reset
46191:
2nd-day
system
turned
on (dT~5hr)
61837:
system
reboot for
power
anomaly (dT~380s)
1a 1d 2a 2e 2j 2p 2r 3a 3d 3g 3l 4c 4e 4g
18432-1
844
8: 0
9-0
5-1
7 1
4-3
7-4
7
CC
4S
ys-e
rr (
4)
32508-
32804:
Shield
off
32520-3
253
2: 0
9-0
5-1
7 2
0-0
9-0
1
CC
0S
ys-e
rr (
3)
4d
Clock Histories T(GCU-GPS) vs. ΔT(UHF Clock)
• Q1: Why were the event rates in NCT‟09 much
higher than estimate from NCT‟05 to exceed the
transmission maximum? No-data events
• Q2: What caused the Hz-feature during NCT‟09
flight? Event loss during DAQ through GCU
Frame & Event Rate Histories
during NCT’09 Flight
< System – Event Loss >
Alan Chiu, 2011/03/10
Q1: Much higher event rates than
estimate from NCT’05?
< Calculation from NCT’05 by Mark >
1. Average rate per strip: D0: ~52 counts/sec ; D1: ~50 counts/sec
2. Average event rate per detector: ( flight_03.bin: from frame 140 to frame 40140. Elapsed time: 1231 seconds (+/- 3 seconds) )
- Number of 80-word data frames: 30791 frames,
25.0 frames/sec = 32 kbps.
- D0 events: 57466 events, 46.7 events/sec
- D1 events: 46856 events, 38.1 events/sec
< Calculation from NCT’09 by Eric >
3. 100-MB data files were stored in GCU-disk for about 1500s,
which is about 546 kbps.
~ 42.4 events/GeD/sec
Outline
Q1: Why exceeding the transmission maximum?
Q1
Transmission Rate HistoryA: Comparable transmission rates on valid & no-data events!
(All events)
TRValid-events = TRValid-SYNC -TRNo-data events
Almost all of the no-data events were originally LLD trigger only!
Q1
No-data events = Valid SYNC – Valid events
(384 kbps)
-Valid E0AE: events with SYNC word and group header. -Valid events: Number of events with either timing or ADC data. -No-data event: Events with no data but only headers.
Event Ratio History (Valid vs. No-data)
A: Higher no-data-event rate than valid event rate!
Revise DSP code!
No-data events = Valid SYNC – Valid events
Event Ratio History (relative to total-E0AE)
Q: What caused the difference?
~29.8%~30.2% ~30.7% ~29.9% ~29.8%
Consistent ratio of valid-events in “normal” epochs
~91.9%
A: Because junk events with system errors were compressed in data on
CC4, making no-data events as E&T events.
E&T events from CC4 ↑during this epoch
No-data events in CC4 ~0during this epoch
~29.4%
Q1
Event Rate History
Valid SYNC
Cause of No-data event in NCT’09
• Signature 1: Valid events were about 92% of total in soft-shield-veto mode, while they were about 30% in other normal epochs.
• Signature 2: Almost all of the no-data events were originally LLD trigger only!
1. In the current signal timing diagram, LLD signals will be a few μs later than Shield and Guard-ring Veto signals. (See diagram below)
2. Therefore, following unexpected trigger conditions happened in NCT’09 flight.
-Veto-rejected events can trigger CardCage DAQ again by LLD only trigger.
-Since the signal timing of those events were wrong, most of them were “no-data” events.
-There was very minor effect to CardCage DAQ. Livetime was always greater than 90%.
Shall we keep “LLD-only” trigger? If yes, a wait window (several μs) will be added in the trigger logic to avoid this condition.
Q1
(by CH Lin)
NCT’09 Result form Mark (2010/01/28) NCT’09 Result form Justin (2009/12/08)
Q: Periodically event loss somehow?
In flight data epoch 01 and 02
Ep.012540-2560: CCs reset
Event-loss peak offset
from Pfotzer
maximum ???
Mark’s result
Rates result
Q2: What caused the Hz-feature during NCT’09 flight?
Outline
Q2
ps. This epoch is found as the only epoch with almost no event loss!
(Please check details in 02-Rate-Variation)
(01) Pump-ON, GCU, GCU-sync, shield-off
(06) Pump-off, GCU, GCU-sync, shield-off
(02) Pump-ON, CCGSE, GCU-sync, shield-off
(05) Pump-off, CCGSE, GCU-sync, shield-off
Ground Hz-feature Survey – Loss Rate
No event loss with CCGSEEvent loss always relevant to frame loss. No GCU data to confirm event-loss only…
Q2
Shorter data range due to lots of data loss
No event and frame loss!
No event and frame loss!
No event and frame loss!
No event and frame loss!
Result of Ground Hz-survey by Mark
☆
☆
☆
☆ : 30-minute data were also conducted.
☆
☆
☆
☆
☆ : data with plots shown in former pages.
Q: 1-sec feature appeared with GCU as DAQ?
Pump ON
Pump OFF
(05)
(06)
(02)
(01)
There are occasional peaks in the FFTsof each detector, but there aren't anyrepeated from detector-to-detector, sothey'reprobably just noise. ~Mark
Q2
Cause of Hz-feature in NCT’09
1. The overall event rate showed a periodic 1-second dip, likely due to how the flight computer was handling data packets when in the (unexpected) condition of constant UHF telemetry saturation.
2. The 1 Hz dip is related to the way the flight code handled non-event packets, but there is no easy explanation for the higher-order Hz features. The rate of counter frames, etc. were all 1, 4, or 10 seconds.
Conclusion: change the flight code so it drops all packets when the buffers are full, without prejudice for packet type.
Q2
by Eric Bellm
• No-data events, which were resulted from the failure in LLD-only re-triggers of original shield-veto-rejected events, occupied almost half of the band-width in data transmission to exceed the telemetry upper limit.
Revise DSP code.
• Valid-event ratios to all in “normal” epochs were consistent (~30%).
• Event and frame rates were also checked in ground Hz-feature survey. The result as well as Mark‟s and Justin‟s former studies indicate that “1-Hz feature may be due to periodic event loss caused by UHF telemetry saturation during DAQ process in flight computer”.
Revise flight code.
Summary
< Data Transmission Rates > (NCT’05: 32 kbps)
- All / Valid / No-data events: ~ (599±121) / (308±63) / (291±61) kbps
< Event Rates from all GeDs on the 2nd day (e.g. ep.4e) > (NCT’05: 42.4 /GeD)
- Valid SYNC + lost: ~ 1622±336 events/sec (100%) (no D0, h~39km)
- Estimate lost evt.: ~ 1± 6 events/sec (~ 0%)
- Valid events: ~ 482±101 events/sec (~30%) ( ~53.6 /GeD/sec)
(More in 02-Rate-Variation!)
• Q1: What is the cause of event-rate
variation? Variation of event loss
Altitude transition or electronics issue (?)
• Q2: What were the relevant data/event
rates in NCT’09? Summary (1)
Frame & Event Rate Histories
during NCT’09 Flight
< Rate Variation & NCT’09 Rates >
~ Alan Chiu, 2011/03/10
< Estimate of Loss and Rates >
• EvtID loss: Sum of the discontinuity on event ID assigned in each CC.
• Frame loss: Sum of the discontinuity on frame ID assigned by flight
computer for all frames but GCUHK.
• Frame/Transmission/Event rates are obtained by dividing relevant counters into increment of GCU/GPS time recorded in the flash disk. If
there’s any GCUHK frame-loss, rates will be over-estimated.
Event types:
-Total E0AE: events with SYNC word received. (may lose the rest) -Valid E0AE: events with SYNC word and group header. -Valid events: Number of events with either timing or ADC data.
-E&T event: Events with both ADC and time data.-LLD only event: Events with ADC data only.-FT only event: Events with time data only.
-No-data event: Events with no data but only headers.
“Event” means “trigger” due to “single interaction”
< Data Structure & Event types in GSE >
(Definitions)
Event Rate History(Valid Sync + Lost)
(2) With steady “valid-sync + lost” event rates
(3) Q2: What caused event loss varied to affect event rate?
Valid E0AE = Valid events + No-data eventsValid events = E&T + LLD-only + FT-only evt.
4e 4g2a
Why?
No record was found for it!
(1) Q2: Cause of the event-rate variation?
Outline
Q1
(b). Very little (~0) event loss
here!!
~2380: Pfotzer maximum
61619-62109: System Reboot
46191-48819: System on
(a). Red epochwith significant variation
Relevant System Records
Altitude History (Enlarged)
46191: 2nd-day system turned
on (dT~5hr)
09-0
5-1
7 1
0-1
9-5
4 (C
ST
)
09-0
5-1
801
-27-1
5 (C
ST
)
09-0
5-18
-06-
32-1
5 (C
ST
)
09-0
5-1
813
-15-3
4 (C
ST
)
09-0
5-1
707
-32-5
1 (C
ST
)
09-0
5-1
7 1
8-1
7-3
8 (C
ST
)
09-0
5-1
7 1
8-5
4-2
3 (C
ST
)
09-0
5-1
715
-21-3
2 (C
ST
)
09-0
5-1
716
-34-4
6 (C
ST
)
09-0
5-1
7 1
9-2
2-57
(CST
)
Red epoch: Is event-loss coincidence with altitude transition? Or...?
32508-32804: Shield off
18424-18444: Rate raised
18432-18448: 09-05-17 14-37-47CC4 Sys-err (4)
61837: system
reboot for power
anomaly (dT~380s)
55387: 09-05-18 09-50-22 (1)Reboot after Soft-veto shield mode
Q1
4e2a
Q2: What were the relevant
data/event rates NCT’09?
< Calculation from NCT’05 by Mark >
1. Average rate per strip: D0: ~52 counts/sec ; D1: ~50 counts/sec
2. Average event rate per detector: ( flight_03.bin: from frame 140 to frame 40140. Elapsed time: 1231 seconds (+/- 3 seconds) )
- Number of 80-word data frames: 30791 frames,
25.0 frames/sec = 32 kbps.
- D0 events: 57466 events, 46.7 events/sec
- D1 events: 46856 events, 38.1 events/sec
< Calculation from NCT’09 by Eric >
3. 100-MB data files were stored in GCU-disk for about 1500s,
which is about 546 kbps.
~ 42.4 events/GeD/sec
Outline
Were rates in NCT’09 similar with those in NCT’05?
Q1
Transmission Rate History
(All events)
No-data events = Valid SYNC – Valid events
[ Ep.2a ]<Total> 570 ±72<Valid> 287 ±36
<NoDat>283 ±38(kbps)
[Ep.2e-p ]<Total> 610 ±90<Valid> 317 ±45
<NoDat>293 ±46(kbps)
[ Ep.3g ]<Total> 463 ±54<Valid> 271 ±44
<NoDat>191 ±36(kbps)
[ Ep.3l ]<Total> 438 ±108<Valid> 228 ±51
<NoDat>210 ±59(kbps)
[ Ep.4c ]<Total> 549 ±80<Valid> 281 ±37
<NoDat>268 ±44(kbps)
[ Ep.4e ]<Total> 599 ±121<Valid> 308 ±63
<NoDat>291 ±61(kbps)
TRValid-events = TRValid-SYNC -TRNo-data events
[ Ep.4g ]<Total> 445 ±48<Valid> 229 ±24
<NoDat>217 ±24(kbps)
53787-54103, 54347-55387:
Soft-veto Shield
32508-32804: Shield off
~2380: Pfotzer
maximum
61619-62109: System Reboot
46191-48819: System on
18424-18444: Rate raised
18432-18448: 09-05-17 14-37-47CC4 Sys-err (4)
Relevant System Records
Q2
Event Ratio History (Valid vs. No-data)
Consistent ratio of valid-events in “normal” epochs
Event Ratio History (relative to total-E0AE)
~29.8%~30.2% ~30.7% ~29.9% ~29.8%
~91.9%
Long-term CC4 system-error epoch
E&T events from CC4 ↑during this epoch
No-data events in CC4 ~0during this epoch
~29.4%
Event Rate History
Valid SYNC
No-data events = Valid SYNC – Valid events
Q2
Event Rate History(Valid Sync + Lost)
Steady “valid-sync + lost” event rates
Ep.04e can be a reliable standard for rate estimate in other epoch
Valid E0AE = Valid events + No-data eventsValid events = E&T + LLD-only + FT-only evt.
4e 4g
Very little event loss
here!!
[ Ep.4e ]<VE0AE+L>
1622 ±336<Valid-Evt>
482 ±101(Hz)
(~30% here)
2a
[ Ep.2a ]<VE0AE+L>
1929 ±225<Valid-Evt>
478 ± 64(Hz)
(~25% here)
Why?
No record was found for it!
Q2
Event rate (excluding reset zone): ~ 1903 ±239 Hz @ Ep.02 ~ 1958 ±259 Hz @ Ep.03 ~ 1650 ±291 Hz @ Ep.04
Ep.02 Ep.03 Ep.04
Event Rate History (Valid Events) While shield off, E&Tevent rate ↑, LLD-only event rate ↓
[ Ep.2a ]<Valid> 478 ±64<E&T> 252 ±34<LLD>225 ±31
(Hz)
[Ep.2e-p ]<Valid> 505 ±79<E&T> 268 ±42 <LLD>237 ±38
(Hz)
[ Ep.4e ]<Valid> 482 ±101<E&T> 276 ±58<LLD>206 ±44
(Hz)
[ Ep.4g ]<Valid> 359 ±40<E&T> 204 ±23<LLD>154 ±18
(Hz)
[ Ep.4c ]<Valid> 446 ±73<E&T> 257 ±42<LLD>189 ±31
(Hz)
[ Ep.3g ]<Valid> 481 ±88<E&T> 251 ±98<LLD>155 ±26
(Hz)
[ Ep.3l ]<Valid> 367 ±103<E&T> 199 ±56<LLD>167 ±47
(Hz)
[ Soft-Veto Mode 02]<Valid> 1515 ±268<E&T> 1411 ±251<LLD> 95 ±17 (Hz)
Valid E0AE = Valid events + No-data eventsValid events = E&T + LLD-only + FT-only evt.
Q2
< Data Transmission Rates > (NCT’05: 32 kbps)
- All / Valid / No-data events: ~ (599±121) / (308±63) / (291±61) kbps
< Event Rates from all GeDs on the 2nd day (e.g. ep.4e) > (cf: NCT’05: 42.4 /GeD)
- Valid SYNC + lost: ~ 1622±336 events/sec (100%) (no D0, h~39km)
- Estimate lost evt.: ~ 1± 6 events/sec (~ 0%)
- Valid events: ~ 482±101 events/sec (~30%) ( ~53.6 /GeD)
- E&T events: ~ 276± 58 events/sec (~17%)
- LLD-only events: ~ 206± 44 events/sec (~13%)
< Event Rates from all GeDs on the 1st day (e.g. Ep.2a) >
- Valid SYNC + lost: ~ 1929±225 events/sec (~119%) (+D0, h~35km)
( original counts can be estimated following the above ratio in ep.4e)
- Valid events: (~ 478± 64) events/sec 579 ( ~57.9 /GeD)
< LV0&LV1 Trigger Rates >
- D1, D8-D9 (outer): ~ (40±6)(ep.02), (45±21)(ep.03), (39±6)(ep.04) triggers/sec
- D2-D7 (inner): ~ (30±5)(ep.02), (32±19)(ep.03), (28±5)(ep.04) triggers/sec
< Shield Rates >
- (15472±214) / (16017±288) / (15327±248) Hz @ ep.02 / 03 / 04
Summary (1) – NCT’09 Rates
Because of consistent valid-event rate (~30%) in ep.2a
Q2
Outline
• Features in data/event rates during NCT‟09 flight were surveyed in
details. (more in 03-supplement)
• Event loss rates varied during flight in different epochs.
• Rates concerning data-events in NCT‟09 were checked and then briefly
summarized in previous page.
• Valid-event ratios to all in “normal” epochs were consistent (~30%).
• “Valid-sync + lost” event rates were steady at float.
• Very little event-loss may enable ep.04e to be a reliable standard for rate
estimate in other epoch, if considering steady total-event rates.
< Discussion >
• Was variation during the red epoch due to altitude transition? Or system
problem?
• What caused very little event loss on ep.4e?
Summary (2) – Results
Spectrum Study of Flight Background
• Moltivation:
- Improved background model for future balloon experiments
- Cosmic Background
• Method for the Flight Background Study
• NCT‟05 Result
• NCT‟09 – Corrections & Current Status
- Preliminary results – Flight Spectrum
- Count Rate Corrections
- Examination in epoch 1
- Event distribution about vetoed photons by shields
J.-L. Chiu (Institute of Astronomy, NTHU)
Method for the Flight Background Study
• “Monte Carlo Simulations” (with detailed “NCT Mass Model”) were performed with MGGPOD, a superset of the GEANT3-based MGEANT (version 3.1) containing additional code and scripts for the simulation of proton and neutron induced activation.
• The approach was to simulate particle intensities averaged over the flight trajectory for each of the flight segments for the observing times given.
• Each of the modeled background particle components includes a reference from which a depth-dependent component spectrum was taken.
• Cosmic particles are assumed to be simply absorbed or scattered out of the beam with an effective mass absorption coefficient, α (90 g cm−2 for protons and 36.6 g cm−2 for electrons/positrons; α is energy dependent for photons).
• Source functions for atmospheric photons were integrated along all lines of sight in the model atmosphere, with scattering and absorption accounted for.
• Source spectra were then corrected for the effects of solar modulation and the geomagnetic field.
• Activation due to the protons and neutrons has been simulated with two radi ation histories for each component. For a given flight segment, both prompt and in situ buildup and decay were modeled, after which radioactive decay was simulated in the remaining segments after propagating all the preceding cooldown periods through to the segments in question.
• Results are then filtered through an instrument response model.
J.-L. Chiu (Institute of Astronomy, NTHU)
[15]. B. Rossi, 1952
[41]. J. C. Ling, 1975
[70]. T. Mizuno, 2004
[71]. J. C. Ling, 1971
[72]. T. W. Armstrong, 1973
[73]. R. R. Daniel, 1974
[74]. M. Boezio, 2000
[76]. J. I. Trombka, 1973
[77]. J. I. Trombka, 1977ISOU (direction dependent intensity)
ISOT (isotropic intensity)
NCT’05 Flight Spectra (simulation vs. observation)
(a) 580 g/cm2 – 380 g/cm2 ; (b) 380 g/cm2 – 250 g/cm2
(c) 250 g/cm2 – 190 g/cm2 ; (d) 190 g/cm2 – 150 g/cm2
(a) 150 g/cm2 – 120 g/cm2 ; (b) 120 g/cm2 – 90 g/cm2
(c) 90 g/cm2 – 60 g/cm2 ; (d) 60 g/cm2 – 30 g/cm2
(a) 30 g/cm2 – 5 g/cm2 ; (b) 5 g/cm2 to float altitude
NCT’09 Flight Spectrum (simulation vs. obsevation)
J.-L. Chiu (Institute of Astronomy, NTHU)
<Simulation:>Spectral lines from
atmospheric neutrons & protons correspond to most
lines from data.
< Discussion > 1. Simulation model, which was built couple decades ago, need to be verified. 2. More background data from NCT flight (i.e. in other epochs) and calibrationsneed to be analyzed. 3. NCT data may help to build a better model for future balloon experiments.
<Simulation:>More counts (~1.5x) in
low-energy part
<Simulation:>Lower counts (<0.5x)from 511-keV line.
1. Spectral lines were fit and identified.
2. Obvious 511-keV line (FWHM=7.5 keV).
3. The rest of spectral lines were mainly related to theactivation.
Simulation vs. Data
Line-identification
(Simulation consulting J. D. Bowen, 2009, N. Gehrels,1985)
Next Step
< Observation >
• Divide data from epoch 1 into segments corresponding to the regions at different atmospheric depths.
• Carry out count rate corrections for data in epoch 1.
• Compare spectra among sub-epochs.
< Simulation >
• Conduct simulations for epoch 1 against the corresponding atmospheric depths.
• Carry out the activation simulations throughout the flight.
• Complete simulations with the rest of components.
• Consider the contribution from Crab Nebula.
Improve the background model & obtain the cosmic background
J.-L. Chiu (Institute of Astronomy, NTHU)