The IBEX Data Products used for the ISN temperature analysis presented in Figure 9 and Table 2 of Möbius et al. (2015) are compiled in in an Excel Workbook, which is organized as follows:

There is one worksheet for each ISN season that was analyzed (2009 – 2014)

Shown are 5 data points for each IBEX Orbit arc under investigation. Based on the Start and Stop Times in the “ISN List” the Histogram Corrected Counts of ISN flow distributions in the Level 3 Data of this Release were accumulated for each Orbit arc and then subdivided into 5 time intervals of equal accumulation time.

The data consist of:

- Year Ecl Long: Center Ecliptic Longitude of each time bin in degrees
The center longitude refers to the center time of each of the 5 time intervals. Some intervals may consist of smaller discontiguous intervals. In any case the actual accumulation times are subdivided equally between the left and right half of the time interval.

- Year Sigma: Width of angular distribution
Sigma width of a Gaussian that fit to the angular distribution, including a convolution of the IBEX-Lo collimator function

- Year ∆Sigma: Fit uncertainty of angular distribution
Purely statistical uncertainty from the fit routine

Sigma and ∆Sigma were obtained through a Maximum Likelihood Fit of a Gaussian to the observed angular distributions of the Histogram Corrected Counts of the ISN Flow, after convoluting the collimator function over the Gaussian.

Before using the data for the analysis, two additional corrections were applied:

1) For the analysis the Counts were accumulated in 6^o bins. The difference between an angular distribution at highest angular resolution and 6^o binning was not implemented in the fit, but corrected for afterwards in the second set of columns (E – G). The necessary correction factors had been obtained as a quadratic fit to the difference between model Gaussians with Sigma width of 5^o – 10^o when analyzed with 1^o and 6^o binning. The fit parameters (m0, m1, m2) are shown above the columns with the corrected Sigma widths.

2) The distributions for years 2009 through 2012 were corrected for small flux dependent data transfer suppression between IBEX-Lo and CEU (that is described in Möbius et al. (2015) and Swaczyna et al. (2015)). We corrected all width for an average increase of 1.5% (equivalent to an apparent temperature increase of 3%). 1013 and 2014 data do not show the suppression and thus do not need a correction.

2012 through 2014 data are also subdivided into ascending (a) and descending (b) arcs for the separate analysis shown in Table 2.