我有一篇 Latex 科学论文，其中一页的内容很正常，有 2 列方程式。下一页包含 3 个表格，底部有一些空间，可以再次以 2 列模式放置一些文本。
不幸的是，当我在这个页面上添加文本时，无法返回到 2 列模式。在下一页，也就是第三页，可以返回到这种 2 列格式。
这里是 Latex 代码：
We assume that $\langle B_{b,int}^{2}\rangle$ is small compared to $\langle (b_{b}^{2}\big[\mathcal{D}_{\text{DM}}+B^{C}\big])^2\rangle$ since the huge number of botometric density of galaxies.\\
Then, we can rid of the term in the first ratio term above. Using \eqref{na} and \eqref{nb} :
\begin{align}
\sigma_{o}^{2}&\simeq \dfrac{\langle (b_{a}^{2}\cancel{\big[\mathcal{D}_{\text{DM}}+B^{C}\big]})^2\rangle}{\langle (b_{b}^{2}\cancel{\big[\mathcal{D}_{\text{DM}}+B^{C}\big]})^2\rangle}+\dfrac{\langle B_{a,int}^{2}\rangle}{\langle (b_{b}^{2}\big[\mathcal{D}_{\text{DM}}+B^{C}\big])^2\rangle}-\bigg(\dfrac{b_{a}^{2}}{b_{b}^{2}}\bigg)^{2}\\
&=\dfrac{\langle B_{a,int}^{2}\rangle}{\langle (b_{b}^{2}\big[\mathcal{D}_{\text{DM}}+B^{C}\big])^2\rangle}
\end{align}
\begin{equation}
\text{with the relation :}\langle (b_{b}^{2}\big[\mathcal{D}_{\text{DM}}+B^{C}\big])^2\rangle\,=\, \langle [\mathcal{D}_{\text{b}}\big]^{2}\rangle
\end{equation}
one can write :
\begin{align}
\sigma_{o}^{2}&\,\simeq\, \dfrac{\langle B_{a,int}^{2}\rangle}{\langle (b_{b}^{2}\big[\mathcal{D}_{\text{DM}}+B^{C}\big])^{2})\rangle}\,\simeq\,\sum_{\ell=1}^{n}\dfrac{2}{(2 \ell_{i}+1)\Delta\ell f_{\mathrm{sky}}}\,\dfrac{1}{\langle \big[\mathcal{D}_{\text{b}}\big]^{2}\rangle}\,\dfrac{1}{\bar{n}^{2}_{a}}
\label{sig_2}
\end{align}
Then, if one comes back to the notation of variance on $\hat{O}$ for current redshift bin $i$, one has :
\begin{equation}
\sigma_{o,i}^{2}\,\simeq\,\sum_{j=1}^{n}\dfrac{2}{(2 \ell_{j}+1)\Delta\ell f_{\mathrm{sky}}}\,\dfrac{1}{\langle \big[\mathcal{D}_{\text{b},i}\big]^{2}\rangle}\,\dfrac{1}{\bar{n}^{2}_{a,i}}
\label{sig_3}
\end{equation}
with $\bar{n}_{a,i}$ is the density of galaxies for the current $i$-th bin considered. The values for each bin are located below in Table \ref{table:density2}. The sum of densities of all 5 bins is equal to $0.35$ galaxies/arcmin$^2$.
\subsubsection{aecific binning for probes combination}
From IST paper \cite{IST}, we should normally have 4 bins and not 5. Below the official table of density of galaxies, redshift bins and redshift bin widths :
\begin{table*}[h!]
\begin{center}
\begingroup
\setlength{\tabcolsep}{6pt} % Default value: 6pt
\renewcommand{\arraystretch}{1} % Default value: 1
\begin{tabular}{|c|c|c|c|c|}
\hline
\addstackgap{Redshift common central bin} & 1.0 & 1.2 & 1.4 & 1.65 \\ \hline
\addstackgap{$\mathrm{d} N\left(z_{\text {mean }}\right) / \mathrm{d} \Omega \mathrm{d} z\left[\mathrm{deg}^{-2}\right]$} & 1815.0 & 1701.5 & 1410.0 & 940.97 \\ \hline
\addstackgap{$\Delta z$ : width of the central bin} & 0.2 & 0.2 & 0.2 & 0.3 \\ \hline
\addstackgap{Density for current bin (galaxies.arcmin$^{-2}$}) & 0.10083333 & 0.09452778 & 0.07833333 & 0.07841417 \\ \hline \hline
\addstackgap{Total Density gal.arcmin$^{-2}$ - 4 bins} & 0.3521 &
\multicolumn{1}{c}{} &
\multicolumn{1}{c}{} &
\multicolumn{1}{c|}{} \\ \hline
\end{tabular}
\endgroup
\end{center}
\caption{From IST paper \cite{IST} - Number of Galaxies per unit area and redshift interval : 4 bins}
\label{density1}
\end{table*}
We have adapted the binning and the width of each IST bin table in order to get the same $n(z)$ botometric population, and interpolated densities taking into account of the last common bin $z=1.688$. All of this is done to make match the aectroscopic and botometric bins.\\
These modifications are showed in Table below :
\begin{table*}[h!]
\begin{center}
\begingroup
\setlength{\tabcolsep}{6pt} % Default value: 6pt
\renewcommand{\arraystretch}{1} % Default value: 1
\begin{tabular}{|c|c|c|c|c|c|}
\hline
\addstackgap{Redshift common central bin} & 0.9595 & 1.087 & 1.2395 & 1.45 & 1.688 \\ \hline
\addstackgap{$\mathrm{d} N\left(z_{\text {mean }}\right) / \mathrm{d} \Omega \mathrm{d} z\left[\mathrm{deg}^{-2}\right]$} & 1807.76 & 1793.63 & 1655.01 & 1320.51 & 870.13 \\ \hline
\addstackgap{$\Delta z$ : width of the central bin} & 0.119 & 0.136 & 0.169 & 0.252 & 0.224 \\ \hline
\addstackgap{Density for current bin (galaxies.arcmin$^{-2}$)} & 0.05975651 & 0.06775935 & 0.07816296 & 0.09243570 & 0.05365801 \\ \hline \hline
\addstackgap{Total Density (gal.arcmin$^{-2}$) - 5 bins} & 0.3517 &
\multicolumn{1}{c}{} &
\multicolumn{1}{c}{} &
\multicolumn{1}{c}{} &
\multicolumn{1}{c|}{} \\ \hline
\end{tabular}
\endgroup
\end{center}
\caption{Number of Galaxies per unit area and redshift interval : 5 bins}
\label{table:density2}
\end{table*}
\subsection{Results of aectroscopic/botometric combination}
\begin{table*}[h!]
\begin{center}
\begingroup
\setlength{\tabcolsep}{3pt} % Default value: 6pt
\renewcommand{\arraystretch}{3.5} % Default value: 1
\begin{tabular}{|c|c|}
\hline
\addstackgap{$\mathlarger{\sum_{i=1}^{n}}\dfrac{2}{(2 \ell_{i}+1) f_{\mathrm{sky}}\Delta\ell} \quad (\text{with}\,f_{\mathrm{sky}}=15000./41253.0 \simeq 0.363$}) \text{and with $\Delta\ell=4990/60\simeq83.16$} : 0.0020022 & \\ \cline{1-1}
\addstackgap{$\mathcal{D}_{\text{b,$j$}}=\Delta\ell\, \mathlarger{\sum_{i=1}^{n}}$}\,
$C_{\ell,\text{b},j} (\ell_{i}) : 0.00120448, 0.0010552, 0.00091048, 0.00073706, 0.00066718, $ & \\ \cline{1-1}
\addstackgap{$\dfrac{1}{\langle \big[\mathcal{D}_{\text{b,$j$}}\big]^{2}\rangle}$} : 6.89286e+5, 8.98117e+5, 1.206320e+6, 1.840745e+6, 2.246532e+6
& \\ \cline{1-1}
\addstackgap{$\dfrac{1}{\bar{n}_{a,i}^{2}}$ $\text{(str$^{2}$)}$
=$\text{Var}(B_{a,i})$} : 2.0050e-12, 1.5594e-12, 1.1719e-12, 8.3796e-13,
2.4867e-12 & \\
\cline{1-1}
\addstackgap{$\dfrac{1}{\bar{n}_{b,i}^{2}}$ \text{(str$^{2}$)}$=\text{Var}(B_{b,i})$} : (3.0/8.46159e-08)$^{-1}$ = 7.9553e-16\,\,\(\ll\)\,\,$\langle \big[\mathcal{D}_{\text{b}}\big]^{2}\rangle$ & FoM = 1556 \\ \cline{1-1}
\addstackgap{$\text{Central redshift common bins}$} : 0.9595,\,\, 1.087,\,\, 1.2395,\,\, 1.45,\,\, 1.688\,\, & \\ \cline{1-1}
\addstackgap{$\sigma_{o,i}$}\,\, :\,\,0.00479017, 0.00482285, 0.00484123, 0.00504779, 0.00958328 & \\ \cline{1-1}
\addstackgap{$\dfrac{1}{\sigma_{o,i}}$ \,\, :\,\, 208.76079553, 207.34622092, 206.55917251, 198.10655604, 104.3484116} & \\ \cline{1-1}
\hline
\end{tabular}
\endgroup
\end{center}
\caption{Numerical values of quantities appearing in the computation of $\sigma_{o,i}^{2}$ \eqref{sig_2}}
\label{table_quantities}
\end{table*}
\subsection{Final results}
Our cross-correlation method with the using of observable $O$ as described from the beginning gives constraints correaonding to a FoM = 1556. \\
\section{Introduction of a new second observable $O$}
\subsection{Definition of new second observable}
We introduce a new second observable $O$ :
\begin{equation}
O=\dfrac{\sum\limits_{\ell=\ell_{min}}^{\ell_{max}}\sum\limits_{m=-\ell}^{\ell} a_{\ell m}^2}{\sum\limits_{\ell=\ell_{min}}^{\ell_{max}}{\sum\limits_{m=-\ell}^\ell (a'_{\ell m})^2}}
\end{equation}
The sign prime $\prime$ refers to botomeric qantities and without $\prime$ for aectroscopic quantities.
Theoritically, this new second observable is equal to the squared ratio between the aectroscopic and botometric bias :
\begin{equation}
O=\dfrac{\sum\limits_{\ell=\ell_{min}}^{\ell_{max}}C_{\ell,a}}{\sum\limits_{\ell=\ell_{min}}^{\ell_{max}}C'_{\ell,b}}
\end{equation}
As we have the equality $C_{\ell,a}=\bigg(\dfrac{b_{a}}{b_{b}}\bigg)^{2}\,C_{\ell,b}$, we can simplify to get :
\begin{equation}
O=\bigg(\dfrac{b_{a}}{b_{b}}\bigg)^{2}
\end{equation}
We find the same constant value than with our first observable but
with different quantities involved which are sums of $a_{lm}^{2}$.
\subsection{Computation for variance of second new observable $O$}
We are interested in computing the variance of this observable $O$
where $(a_{\ell m}, \ell\in \{1, \cdots, N\}, |m|\leq \ell)$ and $(a'_{\ell m}, \ell\in \{1, \cdots, N\}, |m|\leq \ell)$ are independent random variables, with $a_{\ell m}\sim \mathcal{N}(0, C_\ell)$ for each $|m|\leq \ell$ and $a'_{\ell m}\sim \mathcal{N}(0, C_\ell')$ for each $|m|\leq \ell$.\\
We recall the properties of a few basic distributions. We use $\sim $ to denote equality in distribution. We have :
\begin{enumerate}
\item $\mathcal{N}(0, C)^2\sim C\chi^2(1)=\Gamma(\frac{1}{2}, 2C)$,
\item $\langle \Gamma(k, \theta)\rangle=k \theta$ and $\operatorname{Var}(\Gamma(k, \theta))=k \theta^2$, and
\item $\sum_{i=1}^N\Gamma(k_i, \theta)= \Gamma(\sum_{i=1}^N k_i, \theta)$ for independent summands.
\end{enumerate}
\subsection{Distribution followed by second new observable}
Using points 1 and 3 again, we obtain :
\begin{align}
\sum_{m=-\ell}^\ell (a_{\ell m})^2\notag &=
\sum_{m=-\ell}^{\ell} C_{\ell} \cdot\left(\frac{a_{\ell, m}}{\sqrt{C_{\ell}}}\right)^{2}\\
&\sim \sum_{m=-\ell}^{\ell} C_{\ell} \, \mathrm{Chi} \mathrm{Sq}(1) \\
&\sim C_{\ell} \sum_{m=-\ell}^{\ell} \mathrm{ChiSq}(1) \\
&\sim C_{\ell} \, \mathrm{ChiSq}(2 \ell+1)\\
&\sim C_{\ell}\,\mathrm{Gamma}((2\ell+1)/2, 2)\\
&\sim \mathrm{Gamma}((2\ell+1)/2, 2C_\ell),\label{denom}
\end{align}
We have taken the convention (shape,scale) parameters for $\mathrm{Gamma}$ distribution. Given the fact that we consider the random variable :
目前，第 11.3 节“分布后跟第二个新的可观察值”的文本被推到右侧，而不是显示在左侧第一列。
如何让 3 个表格和底部空间采用 2 列标准格式，就像论文的其余部分一样？欢迎提出任何建议。
编辑：以下是使用的相关包：
\documentclass[longauth]{aa}
\usepackage[justification=centering]{caption}
\usepackage[export]{adjustbox}
\usepackage[absolute,overlay]{textpos}
\usepackage{float}
\usepackage{cancel}
\usepackage{stackengine}
\usepackage{relsize}
\usepackage{graphicx}
\usepackage{txfonts}
\usepackage{hyperref}
\usepackage{bm}
\usepackage[dvipsnames]{xcolor}
\usepackage[T1]{fontenc}
\DeclareUnicodeCharacter{2212}{ }
\usepackage{braket}
\usepackage[autostyle]{csquotes}
我无法提供完整的源代码，它太长了。
答案1
这是一个满足您所有需求的模板：
\documentclass[10pt,a4paper]{book}
\usepackage[utf8]{inputenc}
\usepackage[T1]{fontenc}
\usepackage{amsmath}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{graphicx}
\usepackage{scrextend}
\usepackage[english]{babel}
\usepackage{blindtext,multicol,float}
\begin{document}
\noindent \large Normal width\\
\normalsize
\blindtext\\
\begin{addmargin}[-4em]{-4em}
\blindtext\\
\begin{multicols}{2}
\blindtext\\
\end{multicols}
\blindtext\\
\end{addmargin}
\blindtext\\
\begin{addmargin}[2em]{3em}
\blindtext\\
\end{addmargin}
\begin{multicols}{3}
\blindtext\\
\end{multicols}
\begin{center}
\begin{tabular}{|c|c|r|}
\hline
abc & dfg & hij \\
\hline
klm & nop & qrs \\
\hline
\end{tabular}
\end{center}
\[y=\sqrt{\dfrac{x+1}{2x+3}}\]
\begin{multicols}{2}
\blindtext\\
\begin{figure}[H]
\centering
\includegraphics[scale=.2]{logoRS.png}
\caption{Logo}
\end{figure}
\end{multicols}
\end{document}
