A document by Shelley Stall. Click on the document to view its contents.

Data science refers to the set of tools, technologies, and teams that alter the paradigm by which data are collected, managed and analyzed. Data science is, therefore, decidedly broader than ‘machine learning,’ and includes instead the full data lifecycle. Never has the need for effective data science innovation been greater than now when at every turn data-driven discovery is both burdened and invigorated by the growth of data volumes, varieties, veracities, and velocities. This growing scale of science requires dramatic shifts in collaborative research, requiring projects to climb the gradations of collaboration from unidisciplinary, to multi-, inter-, and transdisciplinary (Figure 1, [Hall et al., 2014; NRC, 2015]), and perhaps even to an entirely new level that defies any traditional boundary, or antidisciplinary (https://joi.ito.com/weblog/2014/10/02/antidisciplinar.html). We will discuss the cutting-edge efforts advancing collaborative research in Space Physics and Aeronomy, highlight progress, and synthesize the lessons to provide a vision for future innovation in data science for Heliophysics. We will specifically focus on three trail-blazing initiatives: 1) the NASA Frontier Development Laboratory; 2) the HelioAnalytics group at the Goddard Space Flight Center in cooperation with the NASA Jet Propulsion Laboratory’s Data Science Working Group; and 3) an International Space Sciences Institute project. References: Hall, K.L., Stipelman, B., Vogel, A.L., Huang, G., and Dathe, M. (2014). Enhancing the Ef- fectiveness of Team-based Research: A Dynamic Multi-level Systems Map of Integral Factors in Team Science. Presented at the Fifth Annual Science of Team Science Confer- ence, August, Austin, TX. NRC (National Research Council) (2015). Enhancing the Effectiveness of Team Science. Washington, DC: The National Academies Press. https://doi.org/10.17226/19007.

We are at a unique time in the study of our place in space. On one hand, we operate in the same paradigm that has guided the study of space science for the past couple of decades, and on the other a rising dependence of our economic and social well-being on space demands a shift. Everywhere in our society ‘big data’ (defined by four V’s: volume, variety, veracity, and velocity) and the advent of sophisticated and efficient methods to explore these data (i.e., data science) present new opportunities for discovery, and the time is ripe for these methods to shift how we study the physics of space. We will first discuss the meaning of data science in the context of space science, and then demonstrate the potential for new discovery through a power use case: leveraging Global Navigation Satellite Systems (GNSS) signals for space weather prediction. In this use case, we take advantage of a large volume of data from GNSS signals, data science-driven technologies, and a machine learning algorithm known as the Support Vector Machine (SVM) to develop a novel predictive model for high-latitude ionospheric phase scintillation. This talk will conclude with a perspective on opportunities in space science through ‘big data’ and creating new scientific discovery at the intersection of traditional approaches and data science-driven innovation.

Field-aligned currents (FACs), or the system of currents flowing along Earth’s magnetic field lines, are the dominant form of energy and momentum exchange between the magnetosphere and ionosphere. FACs are ubiquitous across the high-latitude region and have unique characteristics depending on the magnetospheric or solar wind source mechanism, and, therefore, mapping location in the ionosphere (i.e. auroral zone, polar cap, cusp). Further complicating the picture, FACs also exhibit a large range of spatial and temporal scales. In order to create new understanding of FAC spatial and temporal scales, their cross-scale effects, and the impact on the polar region, including on critical technologies, new data analysis approaches are required. This talk addresses a coherent progression of investigation in three parts: 1) an exploration of the characteristics, controlling parameters, and relationships of multiscale FACs using a rigorous, comprehensive analysis across multiple spacecraft observations; 2) augmentation of these statistical results with detailed case studies, fusing observations from diverse platforms and incorporating critical information about the high-latitude electrodynamics across scales; and 3) a quantitative investigation of the impact on Global Navigation Satellite System (GNSS) signals. We find that the relationships between FAC scales are complex and reveal new information about the connection between multiscale FACs and irregular space weather activity. Additionally, there are observable signatures of multiscale FACs and resultant electrodynamic activity in ionospheric data from GNSS signals, suggesting that these signals are affected distinctly according to scale size of the coupling process. Our results indicate that GNSS data may be a powerful source of information about the multiscale near Earth space environment.