Physically Accurate Large Dynamic Range Pseudo Moments for the MMS Fast Plasma Investigation

摘要

The Fast Plasma Investigation onboard the Magnetospheric Multiscale (MMS) mission captures full‐sky particle data over a range of 10?eV to 30?keV every 30?ms for electrons and 150?ms for ions. Only about 5% of the count data at this time resolution are sent to the ground due to telemetry limitations; however, a complete set of onboard trigger terms are sent due to their small data volume. These trigger terms are composed of simplified integrals that scale proportionally to plasma moments. They were initially conceived as a mechanism to rank scientific data for priority downlink; however, they can be adapted to represent pseudo moment quantities that are accurate to within several percent of fully corrected plasma moments. Tuning the dynamic range per plasma region and correcting via a trained neural network allow for far greater accuracy than previous missions. Corrections are performed based on statistical sampling of data from the first MMS day side pass. The corrections are validated by comparing pseudo moments to a published event showcasing detection of kinetic Alfvén waves. Physically accurate pseudo plasma moment quantities are available for the entirety of the MMS prime science regions, and can be a valuable asset for future scientific analysis of the MMS data repository. Plain Language Summary Space missions are generating more and more data as instrumentation becomes more advanced. It is becoming increasingly difficult to downlink all of this data due to telemetry concerns, meaning that many missions now only downlink a fraction of their high resolution data. This work describes a new method of using onboard, simplified integral approximations of scientific quantities. The dynamic range of the environment is preserved by tuning the limited range of the system to an appropriate interval using a scaling system. These simplified data undergo a robust correction process on the ground leading to physically meaningful quantities that are accurate to within several percent of the raw high resolution science data. Because these simplified onboard data integrals are significantly smaller in data volume, all of them can be downlinked leading to a dramatic increase in available scientific data from the mission.