论文标题
最佳1D LY- $α$森林功率谱估计I：Desi-Lite Spectra
Optimal 1D Ly-$α$ Forest Power Spectrum Estimation I: DESI-Lite Spectra
论文作者
论文摘要
1d ly- $α$ forest通量功率谱$ p _ {\ mathrm {1d}} $对比典型的星系调查小的尺度敏感，因此与天际层次培养基的热状态相关，抑制了中性群体和新的暗物质模型。它已经成为研究新物理学的竞争框架，但也面临着各种挑战和系统的分析错误。在这项工作中，我们对$ p _ {\ mathrm {1d}} $的最佳二次估计器进行了重新审视，这对相关问题（例如像素掩码，频谱中的时间演变和quasar contrumum continuum continuum误差）非常有力。我们通过引入基准功率谱来进一步改善估计量，这使我们能够通过减轻频能的离散性来提取更多信息。我们精心将我们的方法应用于合成的DESI光谱，并演示估计器如何克服每个挑战。我们进一步应用了一个优化方案，该方案将Fisher矩阵近似于每排的三个要素，并将计算时间减少60％。我们表明，在没有系统的情况下，使用5年DESI数据可以在$ p _ {\ mathrm {1d}} $中获得百分比精度，并为不同的光谱质量提供预测。
The 1D Ly-$α$ forest flux power spectrum $P_{\mathrm{1D}}$ is sensitive to scales smaller than a typical galaxy survey, and hence ties to the intergalactic medium's thermal state, suppression from neutrino masses and new dark matter models. It has emerged as a competitive framework to study new physics, but also has come with various challenges and systematic errors in analysis. In this work, we revisit the optimal quadratic estimator for $P_{\mathrm{1D}}$, which is robust against the relevant problems such as pixel masking, time evolution within spectrum and quasar continuum errors. We further improve the estimator by introducing a fiducial power spectrum, which enables us to extract more information by alleviating the discreteness of band powers. We meticulously apply our method to synthetic DESI spectra and demonstrate how the estimator overcomes each challenge. We further apply an optimisation scheme that approximates the Fisher matrix to three elements per row and reduces computation time by 60%. We show that we can achieve percent precision in $P_{\mathrm{1D}}$ with 5-year DESI data in the absence of systematics and provide forecasts for different spectral qualities.
