论文标题
部分可观测时空混沌系统的无模型预测
Electron acceleration from the interaction of three crossed parallel Alfvén waves
论文作者
论文摘要
储层计算是预测湍流的有力工具，其简单的架构具有处理大型系统的计算效率。然而，其实现通常需要完整的状态向量测量和系统非线性知识。我们使用非线性投影函数将系统测量扩展到高维空间，然后将其输入到储层中以获得预测。我们展示了这种储层计算网络在时空混沌系统上的应用，该系统模拟了湍流的若干特征。我们表明，使用径向基函数作为非线性投影器，即使只有部分观测并且不知道控制方程，也能稳健地捕捉复杂的系统非线性。最后，我们表明，当测量稀疏、不完整且带有噪声，甚至控制方程变得不准确时，我们的网络仍然可以产生相当准确的预测，从而为实际湍流系统的无模型预测铺平了道路。
We study the non-linear interaction of three parallel Alfvén wave packets (AWP) in an initially uniform plasma using 2.5D particle-in-a-cell (PIC) numerical simulations. We aim to help to explain the observation of suprathermal electrons by multiple Alfvén waves collision in regions where these waves are trapped like IAR (Ionospheric Alfvén Resonator), Earth radiation belts or coronal magnetic loops. In the context of APAWI process that have been described by Mottez (2012, 2015), the interaction of two parallel Alfvén waves (AW) generates longitudinal density modulations and parallel electric field at the APAWI crossing region that can accelerate particles effectively in the direction of the background magnetic field. Our simulations show that when a third parallel AWP of different initial position arrives at the APAWI crossing region, it gives rise to a strong parallel electron beams ($V \sim 5--7VTe$) at longitudinal cavity density gradients. We suggest that velocity drift from outgoing AW generates interface waves in the transverse direction, which allows propagating waves to develop parallel electric fields by phase mixing process when the transverse scale of the wavy density gradient (oblique gradient) is in the range of the electron inertial length.
