Postprocessing

An improvement of the Borealis system is the storage of lower-level data products, specifically antenna-level I&Q samples. The generation of these files allows higher-level data products such as RAWACF files to be re-processed, for example, to fix a mistake in the original processing or to beamform in new directions after the data has already been collected. A Python package, borealis_postprocessors, has been written for exactly this purpose. The package contains an identical processing chain to Borealis, as well as additional processing classes for novel post-processing. It is written in an easily extensible way to facilitate the development of new processing capabilities.

As an example of the usefulness of this package, we have used it to implement bistatic experiments using pairs of SuperDARN radars. In order to analyze bistatic data received by a radar, we wanted to first ensure that the transmitting radar in the pair was indeed operating for a given time. To do so, we extended borealis_postprocessors to extract the timestamps from each data file of the transmitting radar during the bistatic experiment, then transferred these timestamp files to the receiving radar computer. Next, we created another class in borealis_postprocessors to read in the timestamp files and only post-process data with matching timestamps from the receiving radar files. This would have been impossible to do in real-time given the limited bandwidth and internet speeds of the radar sites.

For further information on borealis_postprocessors, check out the GitHub page here. Any comments, suggestions, or contributions to the package are encouraged!