Statistics – Applications
Scientific paper
Dec 2004
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2004agufmsm41a1130r&link_type=abstract
American Geophysical Union, Fall Meeting 2004, abstract #SM41A-1130
Statistics
Applications
2720 Energetic Particles, Trapped, 2722 Forecasting, 2784 Solar Wind/Magnetosphere Interactions
Scientific paper
Linear state-space models, most common in engineering applications, offer a more flexible alternative to the familiar finite impulse response (FIR) linear prediction filters commonly used to predict radiation belt electron fluxes based on solar wind input. They can be designed to be mathematically equivalent to FIR models, but may also incorporate dynamic feedback and allow cross-coupling between multiple system outputs, thereby providing an empirically derived description of a system's dynamics that is more consistent with reality. In addition, their numerical structure is ideally suited for use with the Kalman Filter as a form of data assimilation, and/or the so-called extended Kalman Filter for adaptive identification of optimal model parameters. A brief overview of linear state-space models is given, followed by a demonstration of their ability to predict 2-6 MeV electron fluxes based on historical solar wind data and SAMPEX electron observations binned by geomagnetic L-shell (1-8 RE).
Baker Daniel N.
Joshua Rigler E.
Klimas Alex J.
Vassiliadis Dimitris
Weigel Robert S.
No associations
LandOfFree
Predicting Radiation Belt Electron Flux with Adaptive Linear State-Space Models does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Predicting Radiation Belt Electron Flux with Adaptive Linear State-Space Models, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Predicting Radiation Belt Electron Flux with Adaptive Linear State-Space Models will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1469160