Predictive Analytics of Cold Ion Outflow from the Earth's Ionosphere

Nicolas Doepke¹Elena A Kronberg^1*Kun Li²Artem Smirnov¹Raluca Ilie³Fabian Scheipl⁴

• ¹Ludwig Maximilian University of Munich, Munich, Germany
• ²Planetary Environmental and Astrobiological Research Laboratory, School of Atmospheric Sciences, Sun Yat-sen University, Zhuhai, China, Zhuhai, Guangdong, China
• ³Department of Electrical and Computer Engineering, University of Illinois, Urbana-Champaign, IL, USA, Illinois, USA, United States
• ⁴Ludwig-Maximilians-Universitat Munchen, Munich, Germany

In this study, we investigate the cold ions (<70 eV) originated in the high-latitude ionosphere of the Earth entering the magnetosphere towards the magnetotail. We analyze measurements from Cluster spacecraft along with solar irradiance, solar wind (SW), and geomagnetic observations. Two machine learning models driven by solar irradiance and solar wind measurements are derived to predict the cold ion flux. With the linear baseline model, we provide an empirical formula. The nonlinear model (Extra-Trees Regressor) yields 17% better Doepke et al. Frontiers performance. The total cold ion escape rate from the polar cap ranges between ∼1.1·1024 and ∼2.7·1026 s−1. The upper limit is comparable to the neutral escape rate. The results show that spatial location is the most important predictor. Solar EUV irradiance is also among the top predictors, followed by the solar wind electric field, the interplanetary magnetic field (IMF), and solar wind dynamic pressure. These results can help to evaluate the influence of the stellar wind-magnetospheric interaction on the ion outflow at Earth-like exoplanets. They indicate the importance of such an interaction for the atmospheric escape during active geomagnetic conditions. Stronger outflow from the Northern Hemisphere than from the Southern Hemisphere hints that the magnetic field strength can impact the amount of ionospheric outflow.