Qianli Ma ^1,2* Xinliang Gao ^3,4,5 Dedong Wang ⁶

• ¹ Center for Space Physics, Boston University, Boston, Massachusetts, United States
• ² Department of Atmospheric and Oceanic Sciences, College of Physical Sciences, University of California, Los Angeles, Los Angeles, California, United States
• ³ Deep Space Exploration Laboratory, School of Earth and Space Sciences, University of Science and Technology of China, Heifei, China
• ⁴ Center of Excellence for Comparative Planetology, Chinese Academy of Sciences (CAS), Hefei, Anhui Province, China
• ⁵ Collaborative Innovation Center of Astronautical Science and Technology, Harbin, China
• ⁶ GFZ German Research Centre for Geosciences, Potsdam, Brandenburg, Germany

Earth's radiation belts are populated by relativistic particles which are highly dynamic due to various source and loss processes (Reeves et al., 2003;Thorne, 2010;Turner et al., 2014). Satellite observations revealed that the outer radiation belt fluxes are strongly affected by solar wind and geomagnetic activities (e.g., Baker et al., 2019). The most important drivers of the radiation belt variability are radial diffusion due to ultra-low frequency waves (e.g., Mann et al., 2016), and local wave-particle interactions due to whistler-mode waves (e.g., Horne and Thorne, 1998), electron cyclotron harmonic waves (e.g., Zhang et al., 2015), and electromagnetic ion cyclotron (EMIC) waves (e.g., Summers and Thorne, 2003). Quasilinear and nonlinear theories were developed to demonstrate and quantify the importance of each process in the radiation belts (e.g., Albert, 1999;Omura et al., 2008). Numerical simulations generally reproduce the overall source and loss of radiation belt particles (e.g., Ma et al., 2016), but detailed quantification of the observed features is challenging. Machine learning technique is proved to be a useful tool in reproducing and forecasting the particle fluxes in radiation belts (e.g., Bortnik et al., 2018). Although the Van Allen Probes provided a great opportunity to improve the understanding of Earth's radiation belt dynamics, many science questions regarding the wave and particle properties, distributions, variability, and evolution remained open after the end of the spacecraft mission (Li and Hudson, 2021). This research topic, "Radiation Belt Dynamics: Theory, Observation, and Modeling", aims to advance the understanding of radiation belt dynamics and improve the capability to model and forecast the energetic particles and plasma waves in the magnetosphere. This research topic collected 11 research articles and 1 mini review article. The published papers address a wide range of topics in the theory, observation, and modeling of radiation belt dynamics.Most of the radiation belt models are drift-averaged and consider the radial transport as one-35 dimensional radial diffusion process. Lejosne Albert developed a theoretical framework to retain 36 the drift phase information, to resolve the effects of trapped particle bulk motion as well and 37 diffusion. The authors derived formulae to evaluate the drift phase resolved diffusion coefficients, 38 and impacts of particle drift as well as radial diffusion. 39Following their theoretical Lejosne, Albert and angles of unducted waves increase rapidly with latitude and are reflected on the ion hybrid frequency. 98The modeled wave fields are useful to study the particle precipitation by EMIC waves in the future. 99 Shao et al. performed one-dimensional particle-in-cell simulations to study the propagation of 100 magnetosonic waves through density structures. The absorption was found to be as important as 101 reflection, both of which strongly depend on the height and width of density structures. The 102 magnetosonic wave power absorption was suggested to be important to understand the wave 103 distribution in Earth's magnetosphere. 104