论文标题
部分可观测时空混沌系统的无模型预测
Discovery of three new near-pristine absorption clouds at $z=2.6$-4.4
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 discovery of three new "near-pristine" Lyman Limit Systems (LLSs), with metallicities ~1/1000 solar, at redshifts 2.6, 3.8 and 4.0, with a targeted survey at the Keck Observatory. High resolution echelle spectra of eight candidates yielded precise column densities of hydrogen and weak, but clearly detected, metal lines in seven LLSs; we previously reported the one remaining, apparently metal-free LLS, to have metallicity <1/10000 solar. Robust photoionisation modelling provides metallicities [Si/H] = -3.05 to -2.94, with 0.26 dex uncertainties (95% confidence) for three LLSs, and [Si/H] >~ -2.5 for the remaining four. Previous simulations suggest that near-pristine LLSs could be the remnants of PopIII supernovae, so comparing their detailed metal abundances with nucleosynthetic yields from supernovae models is an important goal. Unfortunately, at most two different metals were detected in each new system, despite their neutral hydrogen column densities (10^{19.2-19.4} cm^{-2}) being two orders of magnitude larger than the two previous, serendipitously discovered near-pristine LLSs. Nevertheless, the success of this first targeted survey for near-pristine systems demonstrates the prospect that a much larger, future survey could identify clear observational signatures of PopIII stars. With a well-understood selection function, such a survey would also yield the number density of near-pristine absorbers which, via comparison to future simulations, could reveal the origin(s) of these rare systems.