AI 中文总结
该研究基于Swarm卫星数据，采用广义线性建模方法，针对不同纬度和空间尺度的顶部电离层电子密度及变化性建立统计模型，模型在大尺度表现良好，但小尺度存在物理过程缺失的问题。
AI 中文摘要
电离层是一种高度复杂的等离子体，包含空间尺度范围广泛的电子密度结构。电离层与地球磁层、太阳风以及中性大气的耦合，使其具有高度动态性，且高度依赖驱动过程。因此，对电离层进行建模并考虑所有时空尺度以捕捉其完整动态范围是一项挑战。Swarm是欧洲空间局（ESA）首个用于地球观测的星座任务，由多颗低地球轨道卫星组成。在Swarm-VIP-Dynamic项目期间，利用Swarm的观测数据和日地物理过程的代理变量，开发了一套统计模型。研究采用广义线性建模（Generalised Linear Modelling）的统计建模技术，针对水平空间尺度在7.5 km至100 km之间的电子密度及等离子体结构的变化性分别建立模型，并针对低纬度、中纬度、极光区和极区建立了独立模型。这些模型基于作为潜在物理过程代理变量的解释变量进行预测。电子密度模型的部分拟合优度统计量接近理论最优值，表明该建模方法适用于所开展的任务。空间尺度较大（约100 km）的电离层变化性模型表现良好，但在较小空间尺度下模型性能下降，这表明模型中缺失了部分物理过程，可能的候选过程是不稳定性过程或来自下方的波动活动对电离层的驱动力，而这两者均未被模型捕捉到。
英文摘要
The ionosphere is a highly complex plasma containing electron density structures with a wide range of spatial scales. Coupling of the ionosphere with the Earth's magnetosphere and the solar wind, as well as to the neutral atmosphere, makes the ionosphere highly dynamic and highly dependent on the driving processes. Thus, modelling the ionosphere and capturing its full dynamic range considering all spatiotemporal scales is challenging. Swarm is the European Space Agency's (ESA) first constellation mission for Earth Observation, comprising multiple satellites in low Earth orbit. During the Swarm-VIP-Dynamic project, a suite of statistical models has been developed using observations from Swarm and proxies for heliogeophysical processes. The statistical modelling technique of Generalised Linear Modelling was used to create models for both the electron density and the variability of the plasma structures at horizontal spatial scales between 7.5 km and 100 km. Separate models were created for low, middle, auroral and polar latitudes. The models make predictions based on explanatory variables, which act as proxies for the underlying physical processes. The performance of the models of the electron density approached the theoretical best values for some of the goodness-of-fit statistics. This suggests that the modelling method is appropriate for the task undertaken. The models of ionospheric variability at larger spatial scales (about 100 km) also perform well, however the model performance decreases at smaller spatial scales. This suggests that there are physical processes missing from the models. Possible candidates are instability processes or driving forces of the ionosphere by wave activity from below, neither of which are captured by the models.
Comments32 pages, 3 figures, 4+1 tables, Annex A, and Supplementary materials
