The SEPEM reference datasets

In practice, there are only two data sources that can fullfill the above criteria: the IMP8/GME data, and the SEM/EPS data from successive GOES spacecraft (click here for an overview of the data sources that have been ingested in SEPEM). Even these datasets need substantial processing before they can be used in statistical processing, in order to:

• remove data spikes;
• correct or at least remove episodes where problems such as saturation, pulse pile-up, contamination etc. occur;
• fill data gaps (including gaps introduced by removing bad data).

• cross-calibration of the data;
• rebinning of the individual data point spectra into a common set of energy bins;
• merging of the individual datasets without overlaps in time;
• background subtraction.

At present, there are two versions of the SEPEM reference dataset (RDS):

1. v1.0: a H dataset based on GOES/SEM data cross-calibrated with IMP-8/GME data, and IMP-8/GME before 1984;
2. v2.0: a H and He dataset using GOES/SEM/EPS data only (starting in 1974), whereby IMP-8/GME data were used to re-calibrate the GOES instrument energy channels.

The SEPEM reference proton dataset is intended to be used for analysing proton induced radation effects. The energy range covered by the dataset is 5–200 MeV, in ten logarithmically spaced channels, which can be interpreted as a virtual instrument dataset (the energy channels are listed below).

Data selection

For IMP8, data from the GME and CPME instruments have been analysed for suitability in the reference dataset construction. It turned out that the CPME instrument suffered from severe saturation (and possibly other) effects, and that in addition there were numerous gaps in the dataset. Therefore, it was decided not to use these data. The GME data also show saturation effects and data gaps, but to a less severe extent than the CPME data. Nevertheless, especially during the second half of the mission, a number of very large SEP events are insufficiently covered by the GME data. A detailed analysis of the data caveats was compiled for the CPME and GME datasets.

The GOES/SEM proton data do not suffer from saturation, and only have relatively small data gaps, which can be filled with interpolation or usage of the secondary GOES spacecraft data. A detailed description of the GOES/SEM proton datasets and caveats was also compiled.

In order to make maximum use of the strong points of the respective datasets, the following procedure was used to combine the data:

• IMP8/CPME data were not used.
• For the period prior to the GOES05 mission (01-01-1984), IMP8/GME data were used. Older GOES data (dating back to 1974) are available and have been processed and ingested into the SEPEM database. After analysing these additional data, it was concluded that they are too noisy to be used for the reference proton dataset (note that the GOES05 data for years 1984 and 1985 were retrieved from this old FITS file series).
• For the period 01-01-1984 to 28-02-2011, SEM proton data were used from successive GOES spacecraft as specified in the table below. The G data series (uncorrected five minute averaged data) was used.
• From 01-03-2011 onward, GOES13/EPS data were used (again, uncorrected five minute averaged data).

As the energy range of the SEPEM proton reference dataset is limited to 5–200 MeV, not all energy channels of the GME, SEM and EPS instruments are required. The tables below list the channels which were retained for the construction of the reference dataset.

Last modified on: 14 December 2018.