2018 Workshop: Big Data Challenges

In the near future we can expect to have ever larger volumes of many types of observational data available to the CEDAR and space science community due to the continual addition of ground-based sensors (e.g. photometers, ionosondes, MST radar, etc.) and the growing popularity of CubeSat constellation concepts. One of the most striking example of a rapidly increasing dataset is GNSS observations. The relative easy and affordable placement of ground receivers and their suitability as small satellite payloads has resulted in a high volume of data, both spatially and temporally. However, GNSS data limits the type of phenomena that can be investigated in detail. For example, dense ground-based GNSS receivers are ideal to explore Travelling Ionospheric Disturbances (TIDs) but are limited to land masses. The reverse is true for space-based GNSS observations which provides global coverage but because of the spatial ambiguity may not be the most appropriate to study phenomena like TIDs.

In order to justify future mission/facility support or the purchase of data from commercial resources, the science community must begin to quantify the optimal amount of coverage needed per data type to address specific science questions. Further, the community needs to begin to identify and address the unique challenges associated with "Big Data". The intent of this workshop is to promote discussion on various aspects of this topic such as: 1) Linking available large ground and/or space-based data sets to their ability to address specific science topics alone or in combination, 2) Present ideas for a global coverage metrics for specific types of observations, 3) Address the potential of cumulative errors resulting from ingestion of large volume data from sensor networks (e.g. Absolute TEC from ground GPS receivers with various biases) including their impact on global models, 4) Applying "Big Data" approaches to future constellations and ground sensor networks.