论文标题
部分可观测时空混沌系统的无模型预测
Formation of Quiescent Prominence Magnetic Field by Supergranulations
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。
To understand the formation of quiescent solar prominences, the origin of their magnetic field structures, i.e., magnetic flux ropes (MFRs), must be revealed. We use three-dimensional magnetofriction simulations in a spherical subdomain to investigate the role of typical supergranular motions in the long-term formation of a prominence magnetic field. Time-dependant horizontal supergranular motions with and without the effect of Coriolis force are simulated on the solar surface via Voronoi tessellation. The vortical motions by the Coriolis effect at boundaries of supergranules inject magnetic helicity into the corona. The helicity is transferred and accumulated along the polarity inversion line (PIL) as strongly sheared magnetic field via helicity condensation. The diverging motions of supergranules converge opposite magnetic polarities at the PIL and drive the magnetic reconnection between footpoints of the sheared magnetic arcades to form an MFR. The magnetic network, negative-helicity MFR in the northern hemisphere, and fragmented-to-continuous formation process of magnetic dip regions are in agreement with observations. Although diverging supergranulations, differential rotation, and meridional flows are included, the simulation without the Coriolis effect can not produce an MFR or sheared arcades to host a prominence. Therefore Coriolis force is a key factor for helicity injection and the formation of magnetic structures of quiescent solar prominences.