This page contains the data and scripts corresponding to the analysis presented in the 2017 HAWC paper in the Astrophysical Journal Observation of the Crab Nebula with the HAWC Gamma-Ray Observatory .

Provided here is a file with time-integrated HEALPix equal-area maps of signal and background counts after gamma-hadron cuts in a 3 degree region around the Crab position, for each of the nine analysis bins described in Abeysekara et al., 2017. The detector response file, including the gamma-ray point spread function and true energy distribution for each of the 9 bins is provided as well. This detector response is different from the one used in Abeysekara et al., 2017 due to improvements in the detector modeling. This figure shows a projection of the 3 degree healpix ROI from the counts map onto a flat 2D plane, smoothed with an 0.17 degree gaussian kernel.

We recommend the use of the threeML/astromodels framework (Vianello et al 2015) with the HAL (HAWC accelerated likelihood) plugin for further analysis. Instructions to install an environment to run the analysis may be found here: https://github.com/henrikef/threeML-analysis-workshop

We have prepared an example analysis script which shows how to fit the energy spectrum of the Crab to a log-parabola shape. The script will also output some plots, such as those shown on the webpage and some others, including a spectral energy density plot, as well as a yml file containing the results of the fit, and a likelihood results fits file which also contains more detailed information such as the covariance matrix of the fit. The analysis is performed in python, and you should be sure to put the data and detector response file in the same directory.

Once you have a working environment, you may download and run the analysis here, by running "python crab_fit_logparabola.py". This (gzipped) tarball contains the analysis script linked to above, as well as the expected output files it produces inside a folder called "results".

Fitting the spectrum to a log parabola: dN/dE∝E^{α-βln(E/E₀)}. The best fit results of the analysis will converge to:

• K: 2.54 ± 0.07 × 10^-13 TeV^-1cm^-2s^-1
• α: -2.646 ± 0.019
• β: 0.104 ± 0.015

The binned counts residuals are an important figure of merit for HAWC analysis. Because it is forward-folded, at each iteration of fitting the spectrum is convolved with the detector response to estimate the number of counts in each bin. The best-fit parameters are determined from the maximized likelihood function of the data counts and the model+background.