-
Deep Learning for Space Weather Prediction: Bridging the Gap between Heliophysics Data and Theory
Authors:
John C. Dorelli,
Chris Bard,
Thomas Y. Chen,
Daniel Da Silva,
Luiz Fernando Guides dos Santos,
Jack Ireland,
Michael Kirk,
Ryan McGranaghan,
Ayris Narock,
Teresa Nieves-Chinchilla,
Marilia Samara,
Menelaos Sarantos,
Pete Schuck,
Barbara Thompson
Abstract:
Traditionally, data analysis and theory have been viewed as separate disciplines, each feeding into fundamentally different types of models. Modern deep learning technology is beginning to unify these two disciplines and will produce a new class of predictively powerful space weather models that combine the physical insights gained by data and theory. We call on NASA to invest in the research and…
▽ More
Traditionally, data analysis and theory have been viewed as separate disciplines, each feeding into fundamentally different types of models. Modern deep learning technology is beginning to unify these two disciplines and will produce a new class of predictively powerful space weather models that combine the physical insights gained by data and theory. We call on NASA to invest in the research and infrastructure necessary for the heliophysics' community to take advantage of these advances.
△ Less
Submitted 26 December, 2022;
originally announced December 2022.
-
Multiplicative perturbation bounds for the Generalized block Cholesky downdating problem
Authors:
Mahvish Samara,
Aamir Farooq
Abstract:
The explicit expressions for the strong and the weak rigorous multiplicative perturbation bounds for the Generalized block Cholesky downdating problem are obtained. By bringing together the modified matrix-vector equation approach with the method of Lyapunov majorant function and the Banach fixed point theorem, we derived the strong rigorous multiplicative perturbation bounds. By using the matrix-…
▽ More
The explicit expressions for the strong and the weak rigorous multiplicative perturbation bounds for the Generalized block Cholesky downdating problem are obtained. By bringing together the modified matrix-vector equation approach with the method of Lyapunov majorant function and the Banach fixed point theorem, we derived the strong rigorous multiplicative perturbation bounds. By using the matrix-equation approach the weak rigorous multiplicative bounds are presented. Numerical experiments are provided to illustrate the obtained results.
△ Less
Submitted 25 June, 2021; v1 submitted 9 June, 2021;
originally announced June 2021.
-
On the Persistent Shape and Coherence of Pulsating Auroral Patches
Authors:
B. K. Humberset,
J. W. Gjerloev,
I. R. Mann,
R. G. Michell,
M. Samara
Abstract:
The pulsating aurora covers a broad range of fluctuating shapes that are poorly characterized. The purpose of this paper is therefore to provide objective and quantitative measures of the extent to which pulsating auroral patches maintain their shape, drift and fluctuate in a coherent fashion. We present results from a careful analysis of pulsating auroral patches using all-sky cameras. We have id…
▽ More
The pulsating aurora covers a broad range of fluctuating shapes that are poorly characterized. The purpose of this paper is therefore to provide objective and quantitative measures of the extent to which pulsating auroral patches maintain their shape, drift and fluctuate in a coherent fashion. We present results from a careful analysis of pulsating auroral patches using all-sky cameras. We have identified four well-defined individual patches that we follow in the patch frame of reference. In this way we avoid the space-time ambiguity which complicates rocket and satellite measurements. We find that the shape of the patches is remarkably persistent with 85-100% of the patch being repeated for 4.5-8.5 min. Each of the three largest patches has a temporal correlation with a negative dependence on distance, and thus does not fluctuate in a coherent fashion. A time-delayed response within the patches indicates that the so-called streaming mode might explain the incoherency. The patches appear to drift differently from the SuperDARN-determined $\stackrel{\rightarrow}{E}$ X $\stackrel{\rightarrow}{B}$ convection velocity. However, in a nonrotating reference frame the patches drift with 230-287 m/s in a north eastward direction, which is what typically could be expected for the convection return flow.
△ Less
Submitted 25 June, 2018;
originally announced June 2018.
-
Convergence of the Freely Rotating Chain to the Kratky-Porod Model of Semi-flexible Polymers
Authors:
Philip Kilanowski,
Peter March,
Marko Samara
Abstract:
The freely rotating chain is one of the classic discrete models of a polymer in dilute solution. It consists of a broken line of N straight segments of fixed length such that the angle between adjacent segments is constant and the N-1 torsional angles are independent, identically distributed, uniform random variables. We provide a rigorous proof of a folklore result in the chemical physics literat…
▽ More
The freely rotating chain is one of the classic discrete models of a polymer in dilute solution. It consists of a broken line of N straight segments of fixed length such that the angle between adjacent segments is constant and the N-1 torsional angles are independent, identically distributed, uniform random variables. We provide a rigorous proof of a folklore result in the chemical physics literature stating that under an appropriate scaling, as N tends to infinity, the freely rotating chain converges to a random curve defined by the property that its derivative with respect to arclength is a brownian motion on the unit sphere. This is the Kratky-Porod model of semi-flexible polymers. We also investigate limits of the model when a stiffness parameter, called the persistence length, tends to zero or infinity. The main idea is to introduce orthogonal frames adapted to the polymer and to express conformational changes in the polymer in terms of stochastic equations for the rotation of these frames.
△ Less
Submitted 10 November, 2017;
originally announced November 2017.
-
Scale size-dependent characteristics of the nightside aurora
Authors:
B. K. Humberset,
J. W. Gjerloev,
M. Samara,
R. G. Michell
Abstract:
We have determined the spatiotemporal characteristics of the magnetosphere-ionosphere (M-I) coupling using auroral imaging. Observations at fixed positions for an extended period of time are provided by a ground-based all-sky imager measuring the 557.7 nm auroral emissions. We report on a single event of nightside aurora ($\sim$22 magnetic local time) preceding a substorm onset. To determine the s…
▽ More
We have determined the spatiotemporal characteristics of the magnetosphere-ionosphere (M-I) coupling using auroral imaging. Observations at fixed positions for an extended period of time are provided by a ground-based all-sky imager measuring the 557.7 nm auroral emissions. We report on a single event of nightside aurora ($\sim$22 magnetic local time) preceding a substorm onset. To determine the spatiotemporal characteristics, we perform an innovative analysis of an all-sky imager movie (19 min duration, images at 3.31 Hz) that combines a two-dimensional spatial fast Fourier transform with a temporal correlation. We find a scale size-dependent variability where the largest scale sizes are stable on timescales of minutes while the small scale sizes are more variable. When comparing two smaller time intervals of different types of auroral displays, we find a variation in their characteristics. The characteristics averaged over the event are in remarkable agreement with the spatiotemporal characteristics of the nightside field-aligned currents during moderately disturbed times. Thus, two different electrodynamical parameters of the M-I coupling show similar behavior. This gives independent support to the claim of a system behavior that uses repeatable solutions to transfer energy and momentum from the magnetosphere to the ionosphere.
△ Less
Submitted 6 June, 2017;
originally announced June 2017.
-
Temporal characteristics and energy deposition of pulsating auroral patches
Authors:
B. K. Humberset,
J. W. Gjerloev,
M. Samara,
R. G. Michell,
I. R. Mann
Abstract:
We present a careful statistical analysis of pulsating aurora (PA) using all-sky green line (557.7 nm) images obtained at 3.3 Hz. Six well-defined individual PA patches are identified and extracted using a contouring technique. Quantitative parameters such as the patch duration (on-time and off-time), peak intensity, and integrated intensity are determined for each patch and each pulsation. The re…
▽ More
We present a careful statistical analysis of pulsating aurora (PA) using all-sky green line (557.7 nm) images obtained at 3.3 Hz. Six well-defined individual PA patches are identified and extracted using a contouring technique. Quantitative parameters such as the patch duration (on-time and off-time), peak intensity, and integrated intensity are determined for each patch and each pulsation. The resulting characteristics serve as strict observational constraints that any of the many competing theories attempting to explain PA must predict. The purpose of this paper is to determine the characteristics of PA patches in order to provide better observational constraints on the suggested mechanisms. All aspects of the temporal behavior of the individual patches appear to be erratic. Historically, PA has been defined very loosely and we argue that the use of the term "pulsating" is inappropriate since our findings and other published results are not regularly periodic and thus a more appropriate term may be fluctuating aurora. Further, we find that the observational constraints do not fit well with the flow cyclotron maser theory, which in particular is suggested to create PA patches. There is no clear candidate of the suggested mechanisms and drivers to explain the observational constraints set by the PA patches in a satisfactory manner.
△ Less
Submitted 2 December, 2016;
originally announced December 2016.
