juliensimon/fermi-gbm-triggers

Fermi GBM All-Trigger Catalog

Credit: NASA/JPL-Caltech/STScI/CXC/SAO

Part of a dataset collection on Hugging Face.

Dataset description

All triggers from the Fermi Gamma-ray Burst Monitor -- GRBs, solar flares, SGRs, terrestrial particles, and more. Updated daily from NASA HEASARC.

The Fermi GBM detects transient events across the full unocculted sky in the 8 keV to 40 MeV energy range. While the confirmed GRB catalog contains only verified gamma-ray bursts, this all-trigger catalog includes every trigger the instrument recorded: GRBs, solar flares, soft gamma repeaters (SGRs), terrestrial gamma-ray flashes, particle events, and unclassified triggers.

The GBM trigger system operates continuously on multiple timescales (16 ms to 4.096 s), flagging statistically significant count-rate increases above background in any of its 12 NaI detectors (8 keV - 1 MeV) or 2 BGO detectors (200 keV - 40 MeV). The resulting trigger population is a rich zoo of astrophysical and non-astrophysical transients.

For multi-messenger astrophysics, the complete trigger catalog is essential because sub-threshold events can become significant when combined with external coincidences. The landmark GW170817 / GRB 170817A detection demonstrated that even a weak, off-axis GRB can produce a marginal GBM trigger that becomes unambiguous only in the context of a gravitational-wave detection.

This dataset is suitable for tabular classification tasks.

Schema

Column	Type	Description	Sample	Null %
__row	str		1670	3.0%
version	str		1	3.0%
trigger_name	str		bn260404998	3.0%
name	str	Trigger identifier in the format 'bnYYMMDDFFF' (bn + UTC date + fraction of day, e.g. bn170817529)	GRB260804130	0.0%
ra	float64	Right ascension of best-fit localization, ICRS J2000.0 (degrees, 0-360); null if localization failed	93.83	0.0%
dec	float64	Declination of best-fit localization, ICRS J2000.0 (degrees, -90 to +90); null if localization failed	18.04	0.0%
galactic_lon	float64	Galactic longitude of trigger localization (degrees, 0-360)	192.8313073	0.0%
galactic_lat	float64	Galactic latitude of trigger localization (degrees, -90 to +90)	0.48122292	0.0%
error_radius	float64	1-sigma statistical localization error radius (degrees); GBM typical ~1-10 degrees	3.61	0.0%
time	str		61134.996461846	3.0%
end_time	str		61135.003573068	3.0%
trigger_time	datetime64[ns]	Trigger UTC timestamp, converted from MJD; spans 2008-July to present	2026-08-04 03:07:09.188630244	0.0%
trigger_type	str	On-board trigger algorithm type (e.g. 'long', 'short', 'soft'); reflects the detector timescale and energy range that fired	GRB	0.0%
reliability	float64	Ground-based reliability score (0-1) from automated classification; higher = more likely astrophysical	0.9294	0.0%
trigger_timescale	float64	Trigger accumulation timescale in milliseconds (e.g. 16, 64, 256, 1024, 4096 ms); shorter scales identify short GRBs and TGFs	256.0	0.0%
trigger_algorithm	str		4	3.0%
channel_low	str		3	3.0%
channel_high	str		4	3.0%
adc_low	str		259	3.0%
adc_high	str		1352	3.0%
detector_mask	str		00000000100100	3.0%
geo_long	str		70.9167	4.4%
geo_lat	str		-25.5333	4.4%
ra_scx	str		359.5529	3.4%
dec_scx	str		13.2489	3.4%
ra_scz	str		263.4062	3.4%
dec_scz	str		24.4548	3.4%
theta	str		113.0	6.8%
phi	str		210.0	6.8%
localization_source	str	Origin of the reported sky position: 'flight' (on-board), 'ground' (refined post-downlink), or 'IPN' (triangulation)	Fermi, GBM	15.6%
__x_ra_dec	str		0.162481058116705	3.0%
__y_ra_dec	str		-0.641553847541895	3.0%
__z_ra_dec	str		-0.749672306049428	3.0%
is_grb	bool	True if classification or trigger_type contains 'GRB'; derived column for filtering	True	0.0%

Quick stats

• 12,440 triggers (2008-07-14 to 2026-08-04)
• 4,357 classified as GRBs
• 8,083 non-GRB triggers (solar flares, SGRs, particles, etc.)

Trigger type breakdown

• 4,357 GRB
• 2,776 SFLARE
• 1,778 LOCLPAR
• 1,665 TGF
• 830 UNCERT
• 683 SGR
• 214 TRANSNT
• 131 DISTPAR

Usage

from datasets import load_dataset

ds = load_dataset("juliensimon/fermi-gbm-triggers", split="train")
df = ds.to_pandas()

from datasets import load_dataset

ds = load_dataset("juliensimon/fermi-gbm-triggers", split="train")
df = ds.to_pandas()

# Filter to confirmed GRBs
grbs = df[df["is_grb"] == True]
print(f"{len(grbs):,} GRBs out of {len(df):,} total triggers")

# Non-GRB triggers
non_grb = df[df["is_grb"] == False]
print(non_grb["trigger_type"].value_counts())

# Triggers per year
import matplotlib.pyplot as plt
df["year"] = df["trigger_time"].dt.year
df.groupby("year").size().plot(kind="bar", title="Fermi GBM Triggers per Year")
plt.show()

Data source

https://heasarc.gsfc.nasa.gov/W3Browse/fermi/fermigtrig.html

Update schedule

Daily at 20:00 UTC via GitHub Actions

Related datasets

• juliensimon/gamma-ray-bursts

• juliensimon/fermi-4fgl-dr4

• juliensimon/solar-flare-events

• juliensimon/nicer

• juliensimon/maxi

• juliensimon/auger

• juliensimon/crdb

If you find this dataset useful, please consider giving it a like on Hugging Face. It helps others discover it.

About the author

Created by Julien Simon — AI Operating Partner at Fortino Capital. Part of the Space Datasets collection.

Citation

@dataset{fermi_gbm_triggers,
  title = {Fermi GBM All-Trigger Catalog},
  author = {juliensimon},
  year = {2026},
  url = {https://huggingface.co/datasets/juliensimon/fermi-gbm-triggers},
  publisher = {Hugging Face}
}

License

CC-BY-4.0