Data-driven Applications for the Sun-Earth System
Abstract
Advances in observational and data mining techniques allow extracting information from the large volume of Sun-Earth observational data that can be assimilated into first principles physical models. However, equations governing Sun-Earth phenomena are typically nonlinear, complex, and high-dimensional. The high computational demand of solving the full governing equations over a large range of scales precludes the use of a variety of useful assimilative tools that rely on applied mathematical and statistical techniques for quantifying uncertainty and predictability. Effective use of such tools requires the development of computationally efficient methods to facilitate fusion of data with models. This presentation will provide an overview of various existing as well as newly developed data-driven techniques adopted from atmospheric and oceanic sciences that proved to be useful for space physics applications, such as computationally efficient implementation of Kalman Filter in radiation belts modeling, solar wind gap-filling by Singular Spectrum Analysis, and low-rank procedure for assimilation of low-altitude ionospheric magnetic perturbations into the Lyon-Fedder-Mobarry (LFM) global magnetospheric model. Reduced-order non-Markovian inverse modeling and novel data-adaptive decompositions of Sun-Earth datasets will be also demonstrated.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2016
- Bibcode:
- 2016AGUFMSM32A..01K
- Keywords:
-
- 1910 Data assimilation;
- integration and fusion;
- INFORMATICSDE: 2753 Numerical modeling;
- MAGNETOSPHERIC PHYSICSDE: 7833 Mathematical and numerical techniques;
- SPACE PLASMA PHYSICSDE: 7924 Forecasting;
- SPACE WEATHER