Satellite technology and in particular GPS- or GNSS-based systems are becoming vital for our society. Plasma density structures in the near-Earth space can significantly influence the propagation of GPS signals and hence influence the accuracy of GPS navigation. Moreover, space plasmas can also damage satellites. To carefully evaluate these effects of the space environment, it is important to develop an accurate model of the plasma density based on a variety of direct and indirect measurements.
In this initial project we will demonstrate how machine learning tools can be used to produce a real-time global empirical model of the near-Earth plasma density based on a variety of measurements. The model that will be developed as a follow-up to this current project will then utilize all available data and will be used by a broad range of stakeholders for GPS navigation and satellite operations.
DLR-SO has developed an advanced tomographic reconstruction technique for assimilating GNSS topside total electron content (TEC) data recorded on board Low Earth Orbiting (LEO) satellites such as COSMIC-1/FORMOSAT-3 and METOP-A. Among the difficulties of ionospheric/plasmaspheric imaging based on GNSS measurements, the development of procedures to invert TEC into electron density (Ne) distributions still remains as a challenging task. In this study, a new tomographic reconstruction technique is developed to solve the electron density estimations based on the integral TEC data. The proposed method is evaluated during four geomagnetic storms to check the capabilities of the method to be used as a tool for space weather monitoring. The investigation shows that the developed method can successfully capture and reconstruct the well-known enhancement and depletion in the TEC and electron density during storms main and recovery phases, respectively.
DLR Jena works on data management solutions for large-scale data sets from data-intensive sciences. In this context, DLR Jena develops a data cube solution, which scales to TBs of data while exploiting massive parallelism offered by modern NVMe solid state drives (SSD). Such drives expose different characteristics with respect to reading and writing, which is oftentimes ignored by the corresponding software stacks. Oftentimes this leads to undesired and poor I/O performance. Data cubes by themselves can serve as a scalable, high-performance storage backend infrastructure and are used and further developed in the context of the MAP project to allow fast interactive data explorations on large datasets from ionospheric satellite missions. The goal here is to enable low response times, i.e., less than 300 ms for any scale of data.
Dec 2019 - Aug 2023
Helmholtz Incubator Information & Data Science Pilot-Project