HESE 7.5 year data release


This data release accompanies results published in Physical Review D describing a revisited analysis of the HESE (high-energy starting events) sample with an additional 4.5 years of data, newer glacial ice models, and improved systematics treatment.

For further details, refer to the IceCube publication doi:10.1103/PhysRevD.104.022002 / arXiv:2011.03545.

Data release

Suggested citation for this dataset:

IceCube Collaboration (2021): HESE 7.5 year data release. Dataset. DOI: https://doi.org/10.21234/4EQJ-BB17

Click here to download (.zip, 75 MB)

Provided in this release are several json files. The first contains the 102 data events that pass the HESE selection criterion. The others file contain information for the MC events used to compute the expected data event rates. The contents of these files are described below.

As an instructive example, we provide a python3 script which reproduces the fit of the data to a single power-law astrophysical flux in the same manner as described in the text. The primary goal of these scripts are to provide a working example utilizing the information provided in the data files, and we encourage readers to use these files as a jumping point into their own analyses.

Data files

The zipped data download contains the following:

  • README: A starting point file, describing the full contents of the data release files, including file locations, descriptions of variables provided for the analyzed events, and summaries of provided scripts.
  • LICENSE: The GNU Lesser General Public License.
  • Python scripts: Individual files for scripts described in the README file.
  • “resources” folder: Contains the “data” subfolder, among others, that holds the four json files mentioned above.
  • “test” folder: Contains a single test script.

For additional detail, see the GitHub repository.


In order to run the example scripts, the PHOTOSPLINE software package is required. The package can be found here. In addition to the instructions provided in the photospline README.md, the user must add the -DPYTHON_EXECUTABLE=<full path of python3> option when running cmake. The python3 path can be found by running which python3 on the command line.

Code organization

Out of the box, the provided scripts simply fit the parameters of a single power-law astrophysical flux model to the data. The goal is that these scripts provide an example on how to use the provided information, as well as a starting point for more complex analyses.