Difference between revisions of "SGR1935+2154"

From my_wiki
Jump to: navigation, search
(High Level Analysis)
(High Level Analysis)
Line 415: Line 415:
 
'''The Nightly SEDs and the Light Curve can be seen in the presentation PDFs. The Source is never detected.'''
 
'''The Nightly SEDs and the Light Curve can be seen in the presentation PDFs. The Source is never detected.'''
 
[[File:DL4_Theta2.png|none|thumb|500px|Theta2 plot for the stacked analysis (25.8 h)]]
 
[[File:DL4_Theta2.png|none|thumb|500px|Theta2 plot for the stacked analysis (25.8 h)]]
 +
[[File:paper_dnde_SED.png|none|thumb|500px|Stacked SED ULs DNDE (25.8 h). HESS is taken from their 2021 paper on SGR1935+2154 (2h data, 5 stereo telescopes).]]
 +
[[File:paper_e2dnde_SED.png|none|thumb|500px|Stacked SED ULs E2DNDE (25.8 h). HESS is taken from their 2021 paper on SGR1935+2154 (2h data, 5 stereo telescopes).]]
  
 
==Cuts optimization for a short-scale emission==
 
==Cuts optimization for a short-scale emission==

Revision as of 06:46, 21 June 2024

General Information

  • Name of the source: SGR 1935+2154
  • Brief description of the source:
    • Object type : Soft Gamma Repeater, Galactic Magnetar
    • Other relevant information: Its simultaneous X-ray bursts (INTEGRAL, AGILE) and Fast radio Bursts (CHIME) on April 28th, 2020 were associated, providing the first evidence that FRBs can originate from magnetars.
    • RA: 19 35 00.0 (hh mm ss), Dec: +21 53 59.1 (dd mm ss)
    • RA, Dec in deg (ICRS): 293.75, 21.90
    • Galactic Coordinates l=57.26, b=0.81

SGR 1935+2154 is a Galactic magnetar that presents two types of high-energy emission:

  • the persistent emission at few keV
  • the transient emission of short nonthermal bursts at keV−MeV. The transient emission is characterized by occasional activity period lasting days-weeks with tens-hundreds of random, short (≈0.1s) and irregular bursts.

Repository for the Burst Analysis: gabriele-panebianco-inaf/ShortTimeScaleAnalysis (Please contact me to be added if not visible.)

People involved

  • Gabriele Panebianco (gabriele.panebianco@inaf.it)
  • Rubén López-Coto
  • Andrea Bulgarelli
  • Pol Bordas
  • Paolo Da Vela
  • Tarek Hassan
  • Alicia López-Oramas
  • Ambra Di Piano
  • Nicolò Parmiggiani
  • Irene Jimenez Martinez

Presentations

Data-taking Information

SGR 1935+2154 was observed during July 2021, September 2021, June 2022, October 2022 as follow up to external science alerts. SGR 1935+2154 was observed for a total of 37.7 hours over 118 runs in 17 nights.

Roughly half of the observations were conducted with MAGIC.

Observations are quasi-simultaneous to other facilities observations, both at VHE and MWL. Observations are simultaneous to 9 time of alerts of high-energy satellites triggered by SGR 1935+2154 bursts. See March 2024 slides for comparison between LST-1, MAGIC observations and the Time of Alerts published by external facilities in GCN and ATels (they trace SGR 1935 activity periods). All the available science alerts and time of alerts (trigger times) are recorded here

Most observations are in Wobble mode, though some runs (about 5 hours) were taken in ON mode but without OFF observations (this is the case for all October 2022 runs and 4 July 2021 runs).


SGR 1935+2154 Observations with LST-1
Run Number Night Run Start Time [UTC] Run Elapsed Time [s] Source Offset [deg] Mean pointing azimuth [deg] Mean pointing zenith [deg] Used in stacked analysis mode
5154 20210706 2021-07-06 21:35:40.008366 967.415 0.348 81.331 56.930 False wobble
5155 20210706 2021-07-06 21:55:39.538852 1016.190 0.451 83.681 51.829 False wobble
5156 20210706 2021-07-06 22:18:14.314086 1053.514 0.372 85.327 47.133 True wobble
5157 20210706 2021-07-06 22:41:17.329760 1205.443 0.434 88.929 42.344 True wobble
5158 20210706 2021-07-06 23:03:40.273986 1205.356 0.357 90.827 37.672 True wobble
5161 20210706 2021-07-06 23:38:21.458475 1206.221 0.443 96.226 29.316 True wobble
5162 20210706 2021-07-07 00:00:21.764677 1205.398 0.442 100.261 24.532 True wobble
5163 20210706 2021-07-07 00:25:00.965802 1443.322 0.368 104.802 19.462 True wobble
5164 20210706 2021-07-07 00:49:25.963768 1145.974 0.436 116.293 14.803 True wobble
5165 20210706 2021-07-07 01:10:44.258733 1203.290 0.365 127.337 10.904 True wobble
5166 20210706 2021-07-07 01:32:50.001943 1203.480 0.437 158.649 7.453 True wobble
5167 20210706 2021-07-07 01:54:27.456639 1206.198 0.367 195.814 6.820 True wobble
5168 20210706 2021-07-07 02:16:10.395605 1205.429 0.436 224.205 9.957 True wobble
5169 20210706 2021-07-07 02:38:19.418482 1205.540 0.370 241.561 13.290 True wobble
5170 20210706 2021-07-07 02:59:59.232267 1205.947 0.432 252.307 18.372 True wobble
5171 20210706 2021-07-07 03:22:42.369718 1203.333 0.367 258.984 22.682 True wobble
5175 20210706 2021-07-07 04:15:42.475526 1205.977 0.434 266.547 34.509 True wobble
5176 20210706 2021-07-07 04:35:08.352232 905.936 0.377 269.512 38.264 True wobble
5192 20210707 2021-07-08 01:19:53.546984 1204.101 0.022 142.520 8.584 False ON
5193 20210707 2021-07-08 01:41:14.972905 1205.786 0.021 177.694 6.946 False ON
5194 20210707 2021-07-08 02:02:34.873506 1204.280 0.020 214.208 8.269 False ON
5195 20210707 2021-07-08 02:16:24.810369 292.236 0.019 230.309 10.271 False ON
5196 20210707 2021-07-08 02:25:59.058815 605.461 0.381 236.500 11.662 True wobble
5197 20210707 2021-07-08 02:38:06.497275 606.417 0.418 245.497 14.658 True wobble
5198 20210707 2021-07-08 02:51:02.354297 602.274 0.398 251.387 16.760 True wobble
5199 20210707 2021-07-08 03:03:00.516614 605.874 0.402 252.977 19.611 True wobble
5211 20210709 2021-07-09 22:50:58.651400 1205.470 0.370 90.657 37.872 True wobble
5212 20210709 2021-07-09 23:15:16.868188 1205.343 0.428 94.403 31.785 True wobble
5213 20210709 2021-07-09 23:46:33.454934 1205.261 0.399 98.543 25.197 True wobble
5214 20210709 2021-07-10 00:10:33.631798 1201.304 0.402 106.053 20.354 True wobble
5215 20210709 2021-07-10 00:34:57.851661 1204.504 0.376 112.827 15.470 True wobble
5216 20210709 2021-07-10 01:00:10.746264 1206.832 0.422 131.291 10.101 True wobble
5217 20210709 2021-07-10 01:24:30.383948 1205.869 0.398 160.492 6.922 True wobble
5218 20210709 2021-07-10 01:49:41.124249 1206.164 0.403 205.604 8.073 True wobble
5253 20210712 2021-07-13 00:58:23.075673 1205.001 0.379 138.150 9.080 True wobble
5254 20210712 2021-07-13 01:23:14.131849 1206.654 0.420 184.029 6.975 True wobble
5255 20210712 2021-07-13 01:47:16.445606 1203.150 0.398 221.435 8.531 True wobble
5256 20210712 2021-07-13 02:11:20.228025 1201.871 0.403 239.169 13.172 True wobble
5257 20210712 2021-07-13 02:33:55.047752 1205.168 0.383 250.357 17.114 True wobble
5258 20210712 2021-07-13 02:58:36.310374 1206.582 0.415 258.359 23.067 True wobble
5281 20210716 2021-07-17 01:58:45.359084 606.296 0.382 241.680 13.200 True wobble
5284 20210716 2021-07-17 02:56:34.882860 1205.573 0.414 261.193 26.029 True wobble
5285 20210716 2021-07-17 03:20:04.027781 1205.415 0.399 265.644 30.642 True wobble
5286 20210716 2021-07-17 03:37:31.978401 486.510 0.401 266.766 34.734 True wobble
5294 20210717 2021-07-18 00:00:54.795768 1205.661 0.376 111.738 15.997 False wobble
5295 20210717 2021-07-18 00:24:54.352744 1205.205 0.422 127.894 10.745 False wobble
5299 20210717 2021-07-18 01:13:54.687818 1206.271 0.398 200.712 6.971 True wobble
5300 20210717 2021-07-18 01:37:59.408685 1206.359 0.403 229.105 10.732 True wobble
5301 20210717 2021-07-18 02:03:47.384962 1205.794 0.383 246.089 14.978 True wobble
5320 20210718 2021-07-19 00:20:59.428245 1206.284 0.378 125.137 11.347 False wobble
5321 20210718 2021-07-19 00:45:18.481347 1206.659 0.421 157.894 7.455 False wobble
5322 20210718 2021-07-19 01:10:06.164715 1203.304 0.398 200.913 6.987 False wobble
5326 20210718 2021-07-19 01:53:36.855329 1206.637 0.402 242.509 14.294 True wobble
5327 20210718 2021-07-19 02:18:57.552751 1202.194 0.384 253.242 18.914 True wobble
5328 20210718 2021-07-19 02:43:21.359871 1201.198 0.414 260.142 24.866 True wobble
5329 20210718 2021-07-19 03:07:08.815895 1206.109 0.399 264.874 29.535 True wobble
5330 20210718 2021-07-19 03:30:44.162188 1206.552 0.401 266.901 34.969 True wobble
6125 20210907 2021-09-07 20:59:10.755258 1205.524 0.378 124.154 11.577 True wobble
6126 20210907 2021-09-07 21:23:12.990901 1205.299 0.421 155.315 7.588 True wobble
6127 20210907 2021-09-07 21:47:15.248035 1205.164 0.398 196.709 6.816 True wobble
6128 20210907 2021-09-07 22:10:35.035322 1205.448 0.403 226.375 10.253 True wobble
6129 20210907 2021-09-07 22:33:47.974373 1205.418 0.383 243.388 13.879 True wobble
6198 20210910 2021-09-10 21:00:37.748895 1157.028 0.421 139.966 8.843 True wobble
6199 20210910 2021-09-10 21:22:28.734336 1159.121 0.398 171.222 6.611 True wobble
6200 20210910 2021-09-10 21:43:20.062237 1172.043 0.403 207.669 8.199 True wobble
6201 20210910 2021-09-10 22:06:16.572737 1400.066 0.382 232.938 10.951 True wobble
6202 20210910 2021-09-10 22:38:58.567869 1208.935 0.416 251.895 18.029 True wobble
6203 20210910 2021-09-10 23:00:16.579873 1166.859 0.398 258.398 22.026 True wobble
6204 20210910 2021-09-10 23:21:18.495585 1202.413 0.402 260.967 26.864 True wobble
6205 20210910 2021-09-10 23:42:55.358996 1211.386 0.387 265.112 31.062 True wobble
6206 20210910 2021-09-11 00:02:55.673674 1217.310 0.411 268.387 36.205 False wobble
6208 20210910 2021-09-11 00:47:32.362278 1206.405 0.399 272.881 45.754 True wobble
6223 20210911 2021-09-11 22:47:39.440023 1188.808 0.384 254.708 19.958 True wobble
6224 20210911 2021-09-11 23:09:07.997514 1188.388 0.414 260.555 25.306 True wobble
6225 20210911 2021-09-11 23:30:04.701359 1148.914 0.399 264.754 29.356 True wobble
6226 20210911 2021-09-11 23:49:53.932634 1204.881 0.401 266.265 33.963 True wobble
6227 20210911 2021-09-12 00:12:02.313915 1172.191 0.390 269.596 38.313 False wobble
6228 20210911 2021-09-12 00:30:34.922076 866.843 0.408 272.167 43.141 False wobble
6229 20210911 2021-09-12 00:47:58.089294 906.975 0.401 274.398 46.445 True wobble
6244 20210912 2021-09-12 21:53:20.006110 1284.220 0.382 228.547 10.084 False wobble
6245 20210912 2021-09-12 22:15:07.177227 1181.297 0.417 245.585 14.756 False wobble
6246 20210912 2021-09-12 22:36:24.544463 1197.066 0.398 254.190 18.612 False wobble
6247 20210912 2021-09-12 22:58:01.640121 1232.768 0.402 257.764 23.537 False wobble
6248 20210912 2021-09-12 23:19:36.515968 1208.833 0.386 262.595 27.689 False wobble
6249 20210912 2021-09-12 23:50:33.302675 1328.482 0.411 267.791 35.214 True wobble
6250 20210912 2021-09-13 00:05:53.763888 328.248 0.400 270.158 38.076 True wobble
6251 20210912 2021-09-13 00:19:51.266180 1236.984 0.400 270.625 41.402 True wobble
6252 20210912 2021-09-13 00:41:51.959164 1153.771 0.393 273.459 45.724 True wobble
6253 20210912 2021-09-13 00:57:51.387474 595.861 0.405 275.503 49.979 True wobble
6269 20210913 2021-09-13 22:58:57.497780 1562.563 0.385 259.350 24.064 False wobble
6271 20210913 2021-09-13 23:30:41.016183 312.418 0.387 265.064 30.965 False wobble
6272 20210913 2021-09-13 23:49:12.585701 28.083 0.388 267.697 35.026 False wobble
6273 20210913 2021-09-14 00:00:57.099488 1229.859 0.410 269.620 38.362 False wobble
6288 20210914 2021-09-15 00:04:45.977484 1164.339 0.400 269.760 39.817 False wobble
6289 20210914 2021-09-15 00:25:30.855964 1163.854 0.392 272.532 43.863 False wobble
6290 20210914 2021-09-15 00:46:18.103774 1208.983 0.406 275.120 49.172 False wobble
6291 20210914 2021-09-15 01:08:00.772730 1193.965 0.402 277.627 53.399 False wobble
8607 20220602 2022-06-03 02:18:17.116211 1168.144 0.375 100.131 24.574 True wobble
8608 20220602 2022-06-03 02:39:12.544475 1189.029 0.423 106.122 19.362 True wobble
8610 20220602 2022-06-03 03:12:23.674925 1175.487 0.399 117.749 12.744 True wobble
8611 20220602 2022-06-03 03:32:43.771255 1134.869 0.402 137.193 9.706 True wobble
8612 20220602 2022-06-03 03:52:58.185166 1134.483 0.380 161.859 7.264 True wobble
8613 20220602 2022-06-03 04:12:50.695077 1130.355 0.418 204.330 7.555 True wobble
8614 20220602 2022-06-03 04:28:44.929503 630.429 0.398 225.667 8.973 True wobble
8891 20220624 2022-06-25 01:38:48.786781 1239.415 0.402 116.787 14.563 True wobble
8892 20220624 2022-06-25 02:04:18.557745 1173.345 0.379 131.935 9.980 True wobble
8893 20220624 2022-06-25 02:24:22.669043 1151.907 0.419 164.241 7.185 True wobble
8894 20220624 2022-06-25 02:44:34.480588 1152.283 0.398 199.396 6.903 True wobble
8895 20220624 2022-06-25 03:04:38.363385 1154.168 0.402 224.591 9.965 True wobble
8896 20220624 2022-06-25 03:22:02.053649 841.338 0.383 239.084 12.399 True wobble
9722 20221019 2022-10-19 20:30:27.600677 1794.876 0.015 257.908 22.698 False ON
9723 20221019 2022-10-19 21:00:12.800844 1206.477 0.012 263.753 29.174 False ON
9724 20221019 2022-10-19 21:21:36.919730 1205.612 0.011 266.964 33.859 False ON
9725 20221019 2022-10-19 21:43:03.569971 1206.730 0.009 269.741 38.567 False ON
9726 20221019 2022-10-19 22:04:13.933069 1206.449 0.008 272.208 43.217 False ON
9727 20221019 2022-10-19 22:26:16.460377 1205.381 0.006 274.589 48.049 False ON
9728 20221019 2022-10-19 22:43:13.303530 605.614 0.004 276.339 51.755 False ON
9752 20221020 2022-10-20 21:02:33.395669 5316.643 0.012 264.317 30.563 False ON

Data analysis Information

SGR 1935+2154 is a Galactic magnetar that presents two types of high-energy emission: the persistent emission and the transient emission, hence we performed two types of analysis.

Common starting point are the DL1b produced by lstosa.

To reconstruct the events properties and obtain DL2 data, we employed a set of Random Forests trained with Monte Carlo (MC) simulations tuned to the Night Sky background (NSB) level of our data.

  • Monte Carlo simulations of DL2 and trained RF were requested to lstmcpipe team, NSB tuned. MC prod name: 20230428_src_dec2276_tuned_nsb_az_tel
  • lstchain version: 0.9.13

NSB Tuning and MC production

  • We evaluated the NSB on a sub-sample of our data and obtained the following image modifier parameters:
"image_modifier":{
"increase_nsb": true,
"extra_noise_in_dim_pixels": 1.309,
"extra_bias_in_dim_pixels": 0.54,
"transition_charge": 8,
"extra_noise_in_bright_pixels": 1.79}

We use these parameters for the whole dataset. The true distribution of the parameters of the dataset is shown in the images below:

Parameter distribution for every subrun of the dataset
Parameter distribution for subruns of the good runs (no moon, no bad weather, only wobble)


We produced the configuration files for the lstMCpipe and lstchain, added the image modifier parameters and produced the Monte Carlo simulations as described in the links below.

(Preliminary) MC production = 20230315_src_dec2276_tuned_nsb

... Then we figured out we used sin_az_tel instead of az_tel so the RF would not work with our data, and we had to update the MC production just for this parameter...

Final MC production = 20230428_src_dec2276_tuned_nsb_az_tel

Types of analysis

  • The persistent emission analysis is a "standard" stacked spectral analysis.
    • We obtained DL3 and IRFs with a set of "standard" cuts (details below).
    • We selected "good quality" runs (25h of data).
    • Results are upper limits on the SED and Light Curve on 3 time scales: all stacked data (25h), night-wise (approx 2h), run-wise (approx 20m)
  • The transient analysis is energy-integrated due to poor statistics, and focuses on 100ms time scale.
    • The cuts were chosen to optimise the short-scale sensitivity (details below). This method can be used for the detection of other short transients too.
    • We obtained a new set of DL3 and IRFs with the short scale cuts. We placed upper limits to SGR flux on the 9 time intervals (100ms) centered around the 9 simultaneous Time of Alerts, plus the stacked upper limit
    • We searched for non-simultaneous bursts in all our data, without any detection.

Analysis of Persistent Emission

Bad quality runs and runs taken in ON mode were discarded from the analysis.

The SGR 1935+2154 analysis covers 25.8 hours of observations over 83 runs and 13 nights.

Plots can be found in early presentation pdf.

Low Level Analysis

1 - The DL1 were processed into DL2 sbatching one job per run, using the same path-models and config.

  • We used long jobs and mem-per-cpu 120g:
conda activate lstchain-v0.9.13
lstchain_dl1_to_dl2 \
 --input-files /fefs/aswg/data/real/DL1/20210706/v0.9/tailcut84/dl1_LST-1.Run05163.h5 \
 --path-models /fefs/aswg/data/models/AllSky/20230428_src_dec2276_tuned_nsb_az_tel/dec_2276 \
 --config /fefs/aswg/data/models/AllSky/20230428_src_dec2276_tuned_nsb_az_tel/dec_2276/lstchain_config.json \
 --output-dir /fefs/aswg/workspace/gabriele.panebianco/analysis/SGRJ1935+2154/DL2 

2 - We prepared a DL2->DL3 configuration file (energydep_intensity80_gheff0.7_thetacont0.68.json) to get IRFs and DL3 data.

  • We modified a template existing in lstchain docs by setting:
    • GH-efficiency: 0.7.
    • Theta Containment: 0.68
    • Event selector filter: INTENSITY>80 P.E.

3 - We computed the IRFs for every node as point-like and with energy-dependent cuts.

MCDL2DIRECTORY=/fefs/aswg/data/mc/DL2/AllSky/20230428_src_dec2276_tuned_nsb_az_tel/TestingDataset/dec_2276
lstchain_create_irf_files \
 -g "$MCDL2DIRECTORY/node_theta_23.630_az_100.758_/dl2_20230315_src_dec2276_tuned_nsb_node_theta_23.630_az_100.758__merged.h5" \
 -o "/fefs/aswg/workspace/gabriele.panebianco/analysis/SGRJ1935+2154/IRF/energydep_intensity80_gheff0.7_thetacont0.68/pointlike_Theta23.630_Az100.758/irf.fits.gz" \
 --point-like \
 --energy-dependent-gh \
 --energy-dependent-theta \
 --config "/home/gabriele.panebianco/Workspace/SGRJ1935+2154/DL3configurations/energydep_intensity80_gheff0.7_thetacont0.68.json" 

4 - We processed DL2 files into DL3 with the same configuration file used for the IRFs and associated every run with its closest IRF (in terms of azimuth and zenith).

IRFDIRECTORY=/fefs/aswg/workspace/gabriele.panebianco/analysis/SGRJ1935+2154/IRF
lstchain_create_dl3_file \
 -d "/fefs/aswg/workspace/gabriele.panebianco/analysis/SGRJ1935+2154/DL2/dl2_LST-1.Run05163.h5" \
 -o "/fefs/aswg/workspace/gabriele.panebianco/analysis/SGRJ1935+2154/DL3/energydep_intensity80_gheff0.7_thetacont0.68/pointlike/Run05163/" \
 --input-irf "$IRFDIRECTORY/energydep_intensity80_gheff0.7_thetacont0.68/pointlike_Theta23.630_Az100.758/irf.fits.gz" \
 --source-name "SGRJ1935+2154" \
 --source-ra "293.75deg" \
 --source-dec "21.9deg" \
 --overwrite \
 --config "/home/gabriele.panebianco/Workspace/SGRJ1935+2154/DL3configurations/energydep_intensity80_gheff0.7_thetacont0.68.json" 

5 - We created the DL3 index files to take advantage of gammapy DataStore class in the High Level Analysis.

lstchain_create_dl3_index_files \
 --input-dl3-dir "/fefs/aswg/workspace/gabriele.panebianco/analysis/SGRJ1935+2154/DL3/energydep_intensity80_gheff0.7_thetacont0.68/pointlike" \
 --file-pattern "Run*/dl3*.fits" \
 --overwrite

High Level Analysis

We processed every run in the Low Level Analysis, but we used only a subset for the High Level Analysis made up by good runs.

  • Runs exclusion criteria (info taken from long-term data check files):
    • Source offset not in [0.35,0.45] deg (use only wobble runs, no on runs).
    • Rate of cosmics < 2500 evt/s (avoid problematic runs due to calima or other environmental/instrumental problems).
    • Pedestal charge standard deviation > 2 P.E. (avoid Moon runs).
    • Run 6253 has an asymmetric telescope response (probably due to a humidity patch?)

We performed High-level analysis with gammapy v1.0.

  • Standard Wobble Analysis was performed with 1 off region. Region size is energy dependent and set in the RADMAX IRF.
  • Safe Mask method: aeff-max with aeff-percent=1.
  • Flux Points upper limits were obtained for the source under different conditions:
    • SED ULs in [0.1, 10] TeV, 5 bins per energy decade.
      • 1 set of stacked ULs for total integrated livetime.
      • 13 sets of night-wise ULs.
    • Light Curve UL run-wise and night-wise, energy integrated in [0.1, 50] TeV
  • Flux points obtained with power law spectrum, index=2.4.

Main results: theta2 plot and upper limits on the stacked emission. The Nightly SEDs and the Light Curve can be seen in the presentation PDFs. The Source is never detected.

Theta2 plot for the stacked analysis (25.8 h)
Stacked SED ULs DNDE (25.8 h). HESS is taken from their 2021 paper on SGR1935+2154 (2h data, 5 stereo telescopes).
Stacked SED ULs E2DNDE (25.8 h). HESS is taken from their 2021 paper on SGR1935+2154 (2h data, 5 stereo telescopes).

Cuts optimization for a short-scale emission

See June 2024 slides for the full process.

Optimisation process in a nutshell

  • For a grid of cuts: estimate background rate from an OFF region → get N, the number of counts needed to reject the background (null) hypothesis at 5σ level
  • For a grid of cuts: estimate the cut efficiency on a Monte Carlo file εMC.
  • Divide N by εMC (this is proportional to the sensitivity of the analysis) and find the region with lowest values.
Cuts optimisation. Optimisation region in lower left part of the map.


  • Cuts optimised to detect a 100ms burst above a Poisson background in a single energy bin (>30 GeV):
    • Intensity > 50 p.e.
    • Global Theta2 cut 0.08 deg^2
    • Global Gammaness cut 0.75

Analysis of Transient Emission

VHE Flux Upper Limits on simultaneous X bursts

See June 2024 slides for details.

1 - We obtained a new set of DL3 and IRFs for the transient analysis.

  • We used the IRFs to obtain the effective area and the exposure in a 100ms time bin in the energy range [30 GeV; 10 TeV] assuming a power law with index=2.
  • We counted the ON photons in a 100ms time window centred on the Time of Alert (ToA) provided by ATels/GCNs. The ON radius is given by the theta2 global cut.
  • These counts are either 0 or 1. They are not enough to claim a detection (the threshold is N) → No TeV counterpart of the X bursts was observed.

2 - We can use the ON counts to estimate the upper limit on the transient emission flux.

  • In every bin the expected number of background counts (according to the background rate) is <1, we can consider the arrival of background events negligible in such a short time scale.
  • According to Cowan 2007 (eq. 6), Patrignani 2016 (eq 39.59->39.61, Table 39.3), if we suppose the ON counts are due to a signal Poisson distribution with unknown expected value μ, the ON counts constrain the 95% upper limit on μ (we are estimating the Bayesian upper limit for a Poisson likelihood and uniform prior). In particular:
    • ON = 0 → μUL=3.00
    • ON = 1 → μUL=4.74
    • ON = 2 → μUL=6.30
  • The Flux upper limit can be found by dividing μUL over the bin exposure.

3 - We can stack the counts to estimate the total upper limit on the transient emission of SGR 1935+2154 (over a 0.9s livetime).

  • The average background rate is 1.098 cts/s, so the Poisson expected number of background counts in a 0.9s bin is 0.99. Can we neglect the background?
  • The Bayesian upper limit with negligible background is ON = 2 → μUL=6.30.
  • The Bayesian upper limit with background is the same - expected background counts, i.e. 5.31.
  • If we divide the counts UL by the exposure, we get the flux upper limit. We considered both the cases of an average effective area and the worst effective area to be conservative.


SGR 1935+2154 VHE Upper limits on external simultaneous X bursts.
Run ToA IRF zenith [deg] IRF azimuth [deg] Eff. Area [108 cm2] Livetime [s] Exposure [108 cm2 s] Background Rate [s-1] N ON counts μUL Flux UL [10-8 cm-2 s-1]
5163 2021-07-07 00:33:31.670 23.630 100.758 24.76 0.1 2.476 1.324 5 0 3 1.212
6205 2021-09-10 23:40:34.460 37.814 270 29.59 0.1 2.959 1.200 4 0 3 1.014
6223 2021-09-11 22:51:41.600 23.630 259.265 24.52 0.1 2.452 1.527 5 1 4.74 1.933
6226 2021-09-11 23:55:45.872 37.814 270 29.59 0.1 2.959 1.102 4 0 3 1.014
6228 2021-09-12 00:34:37.450 47.952 270 33.71 0.1 3.371 0.610 4 0 3 0.890
6229 2021-09-12 00:45:49.400 47.952 270 33.71 0.1 3.371 0.666 4 0 3 0.890
6245 2021-09-12 22:16:36.200 10.0 248.117 21.38 0.1 2.138 1.267 4 1 4.74 2.217
6248 2021-09-12 23:19:32.080 23.630 259.265 24.52 0.1 2.452 1.147 4 0 3 1.223
6251 2021-09-13 00:27:25.200 37.814 270 29.59 0.1 2.959 1.042 4 0 3 1.014
NA stacked 0.9s: avg bkg, avg aeff NA NA 27.93 0.9 25.137 1.098 9 2 6.3 0.251
NA stacked 0.9s: avg bkg, worst aeff 10.0 248.117 21.38 0.9 19.242 1.098 9 2 6.3 0.327

Search for non-simultaneous bursts

See June 2024 slides for details.

We selected the DL3 (with short time scale cuts) in SGR 1935+2154 ON region and binned them in 0.1s bins from the start of every run.

The number of counts in a bin follow the background Poisson distribution -> no detection of non-simultaneous bursts.

In order to claim a detection in a bin, we need a number of counts>N corrected by the number of trials.

We used approx 1E6 trials (because we use 118 runs of approx 20 mins in 0.1s bins).

The post-trial N is the number of counts with a post-trial p-value of 2.87E-7. We convert it first into a pre-trial pvalue, then into a N. Maths is explained in March 2024 slides.

LST-1 rates in the short time scale cuts ranges from 0.2 to 2 cts/s, that imply pre-trial N from 3 to 5, and post-trial N from 6 to 9.

Results recap

  • Persistent Emission Analysis:
  1. SGR 1935+2154 is not detected.
  2. We placed Upper Limits on the SED.
  • Transient Emission Analysis:
  1. No burst is detected in coincidence to the 9 external ToAs.
  2. We placed integrated flux TeV 2sigma upper limits in coincidence to simultaneous X bursts.
  3. No burst is detected in the rest of our dataset (non-simultaeous bursts).