\documentclass[11pt]{article}
\usepackage[a4paper,margin=2cm,landscape]{geometry}
\usepackage{amsmath,amssymb,booktabs}
\usepackage[labelfont=bf,labelsep=space,justification=raggedright,singlelinecheck=false]{caption}
\newcommand{\cy}{\overline{c}^{\,y}}
\newcommand{\cmax}{c_{\max}}
\newcommand{\stk}[2]{\begin{tabular}[c]{@{}c@{}}#1\\{\scriptsize$[#2]$}\end{tabular}}
\pagestyle{empty}
\begin{document}
\setcounter{table}{5}
\begin{table}[h]
\centering\footnotesize
\caption{Statistical indices for the stable Prairie Grass experiment (crosswind-integrated ground-level concentration, 25 runs, 75 pairs, and the clean set of 23 runs without runs 13 and 14; this work: Eqs.~3 and~8). Values in brackets are block-bootstrap 95\% confidence intervals for the proposed model. $^{*}$Lagrangian model.}\label{tab:pgs_stat}%
\setlength{\tabcolsep}{2pt}%
\begin{tabular}{@{}llllll@{}}
\toprule
Model & Cor & NMSE & FS & FB & FA2\\
\midrule
This work, $\alpha=1$ (classical limit) & 0.742 & 0.783 & +0.386 & +0.434 & 0.693\\
\quad 95\% CI & $[0.66,\,0.97]$ & $[0.64,\,0.93]$ & $[+0.12,\,+0.77]$ & $[+0.26,\,+0.59]$ & $[0.61,\,0.77]$\\
This work, $\alpha=0.91$ & 0.734 & 1.052 & +0.451 & +0.566 & 0.560\\
\quad 95\% CI & $[0.66,\,0.97]$ & $[0.75,\,1.36]$ & $[+0.17,\,+0.87]$ & $[+0.37,\,+0.74]$ & $[0.44,\,0.67]$\\
\quad clean set (69 pairs), $\alpha=1$ & 0.950 & 0.833 & +0.747 & +0.563 & 0.696\\
\quad 95\% CI & $[0.90,\,0.98]$ & $[0.67,\,0.98]$ & $[+0.67,\,+0.82]$ & $[+0.47,\,+0.63]$ & $[0.61,\,0.78]$\\
\quad clean set (69 pairs), $\alpha=0.91$ & 0.957 & 1.240 & +0.842 & +0.714 & 0.580\\
\quad 95\% CI & $[0.91,\,0.98]$ & $[1.02,\,1.43]$ & $[+0.77,\,+0.91]$ & $[+0.63,\,+0.78]$ & $[0.46,\,0.70]$\\
\midrule
Kumar and Sharan (2012) & 0.71 & 0.72 & -1.05 & -0.57 & 0.68\\
Moreira et al. (2002) & 0.77 & 0.49 & -0.26 & -0.22 & 0.82\\
Carvalho et al. (2007)$^{*}$ & 0.81 & 0.39 & -0.18 & -0.06 & 0.95\\
\bottomrule
\end{tabular}
\end{table}
\end{document}
