论文标题
部分可观测时空混沌系统的无模型预测
Milky Way's eccentric constituents with $Gaia$, APOGEE $\&$ GALAH
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 report the results of an unsupervised decomposition of the local stellar halo in the chemo-dynamical space spanned by the abundance measurements from APOGEE DR17 and GALAH DR3. In our Gaussian Mixture Model, only four independent components dominate the halo in the Solar neighborhood, three previously known $Aurora$, $Splash$ and Gaia-Sausage/Enceladus (GS/E) and one new, $Eos$. Only one of these four is of accreted origin, namely the GS/E, thus supporting the earlier claims that the GS/E is the main progenitor of the Galactic stellar halo. We show that $Aurora$ is entirely consistent with the chemical properties of the so-called Heracles merger. In our analysis in which no predefined chemical selection cuts are applied, $Aurora$ spans a wide range of [Al/Fe] with a metallicity correlation indicative of a fast chemical enrichment in a massive galaxy, the young Milky Way. The new halo component dubbed $Eos$ is classified as $in situ$ given its high mean [Al/Fe]. $Eos$ shows strong evolution as a function of [Fe/H], where it changes from being the closest to GS/E at its lowest [Fe/H] to being indistinguishable from the Galactic low-$α$ population at its highest [Fe/H]. We surmise that at least some of the outer thin disk of the Galaxy started its evolution in the gas polluted by the GS/E, and $Eos$ is an evidence of this process.