Abstract:
Background The numerical assessment of delayed radiation ionization from high-altitude extreme explosions involves complex processes, including debris evolution (as surface or volume sources), radiation transport, and atmospheric ionization. Traditional empirical formulas have limited applicability under certain high-altitude conditions, while Monte Carlo methods are inefficient for hour-scale simulations.
Purpose This study aims to develop an efficient assessment method by constructing a data-driven debris-cloud model using deep learning and coupling it with standard radiation transport theory and atmospheric attenuation mechanisms, so as to calculate the spatial distribution of the delayed-radiation-induced electron production rate.
Methods The debris model was developed using deep learning techniques. Then, an analytical-semi-empirical assessment model was developed by combining radiation transport theory and atmospheric attenuation mechanisms. This model was used to compute the spatial distribution of the electron production rate under high-altitude explosion conditions.
Results The results indicate that the spatial distribution of debris significantly affects the ionization range and intensity. As the debris height and radius increase, the ionization area expands, but the peak electron production rate decreases. Delayed γ-rays exhibit a wide ionization range, with a maximum radius exceeding 3,000 km. In contrast, β-particles have a limited range due to geomagnetic confinement, yet their local ionization intensity is one order of magnitude higher than that of γ-rays.
Conclusions The proposed deep learning-based debris model combined with the analytical-semi-empirical approach provides an efficient and accurate alternative to traditional empirical formulas and computationally intensive Monte Carlo methods. This method effectively captures the key impacts of debris geometry and radiation type, offering a practical tool for assessing delayed radiation ionization from high-altitude extreme explosions over extended time scales.
下载: