Using Deep Learning to Plug an Observational Gap in EUV Irradiance Observations
SouthWest Research Institute
Extreme UV (EUV) radiation from the Sun is pivotal to the energy balance of the Earth?s thermosphere and ionosphere (where most of our satellites reside), driving changes in density that determine satellite orbital decay. To characterize and monitor solar forcing on this system and associated space weather impacts, the EUV Variability Experiment (EVE) instrument onboard NASA's Solar Dynamics Observatory (SDO) was designed to measure solar spectral irradiance (SSI) in the 0.1 to 105 nm wavelength range. However, as the result of an electrical short, one of its components (MEGS-A) failed in May 2014; leaving an observational gap in the most energetic part of the EUV spectrum (5 - 35 nm). In this presentation we will discuss an implementation of convolutional neural networks (CNNs) that uses narrowband UV and EUV images from the Atmospheric Imaging Assembly (AIA) to fill this observational gap. We start with a brief discussion of solar irradiance, its solar cycle variability, and its impact on our upper atmosphere. Then we will discuss CNNs and their capability for abstracting the relevant connection between input and output data, and then we will discuss our implementation and its performance. This work was performed at NASA's Frontier Development Laboratory: a public-private initiative to apply AI techniques to accelerate space science discovery and exploration.
|Date: ||Jeudi, le 8 novembre 2018|
|Lieu: ||Université de Montréal|
| ||Pavillon Roger-Gaudry, Local D-460|
|Contact: ||Paul Charbonneau|