Machine Learning Methods Applied to the Global Modeling of Event-Driven Pitch Angle Diffusion Coefficients During High-Speed Streams
Abstract
Whistler-mode waves in the inner magnetosphere cause electron precipitation in the atmosphere through the physical process of pitch angle diffusion. The computation of pitch angle diffusion relies on quasi-linear theory and becomes extremely time-consuming as soon as it is performed at high temporal resolution from satellite measurements of ambient wave and plasma properties. Such an effort is nevertheless required to capture accurately the variability and complexity of atmospheric electron precipitation, which are involved in various Earths ionosphere-magnetosphere coupled problems. In this work, we build a global machine-learning model of event-driven pitch angle diffusion coefficients for storm conditions based on the data of a variety of storms observed by the NASA Van Allen Probes. We first proceed step-by-step by testing different machine learning methods that have been proved to be efficient. This way we reduce the size of the data set to a minimal unknown set for a given accuracy that is quantified. We provide a comparison of these methods for the March 2013 storm and evolve toward a strategy to apprehend a more general problem with multiple storms. Three methods are retain for their accuracy/efficiency: the neural network, the spline, and the radial basis methods. We then build a much larger data set of the diffusion coefficients computed from the measurements of whistler-mode waves and plasma density for 32 high-speed streams during the first 3 days of the storms. Using a neural network model, we achieve a first global mean event-driven model in which we keep the time dependence of the storm history and introduce a KP-index dependence. This global model will eventually be usable in further various studies of the ionosphere-magnetosphere system that involve electron precipitation. We will show the model allows fast data exploration of the variations of the pitch angle diffusion among its multiple variables and can be used for the prediction of pitch angle diffusion from hiss waves in future high-speed stream storms.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2021
- Bibcode:
- 2021AGUFMNG44A..06K