\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}{8}
\begin{table}[h]
\centering\footnotesize
\caption{Sensitivity to the wind-profile convention (2-D, $\cy$). Mast: median ratio of the model wind to the Prairie Grass mast wind over all levels $0.5$--$16$~m (convective runs / clean stable runs). FA2$_{\alpha=1}$: stable Prairie Grass (75 pairs) / Hanford~83 (30) / pooled (228). Gain: paired difference $\alpha=0.91$ minus $\alpha=1$ within each convention, pooled and for the two Prairie Grass sets, with 95\% block-bootstrap intervals; $\star$ marks intervals excluding zero. Copenhagen saturates at $1.000$ and convective Prairie Grass moves by at most one pair in every variant; Kincaid is examined against its measured 100-m wind in Section~5.1.}
\label{tab:windconv}
\setlength{\tabcolsep}{2pt}%
\begin{tabular}{lcccc}
\toprule
convention & mast & FA2$_{\alpha=1}$ & gain, pooled & gain, PGu\,/\,PGs\\
 & conv.\,/\,stab. & PGs\,/\,Han\,/\,pooled & $\alpha=0.91$ & $\alpha=0.91$\\
\midrule
reference (cut) & 0.97\,/\,0.92 & 0.693\,/\,0.700\,/\,0.737 & \stk{$+0.070^\star$}{+0.02;+0.12} & $+0.22^\star$\,/\,$-0.13^\star$\\[6pt]
uncut & 0.97\,/\,0.92 & 0.693\,/\,0.700\,/\,0.737 & \stk{$+0.066^\star$}{+0.01;+0.12} & $+0.22^\star$\,/\,$-0.15^\star$\\[6pt]
anchored, cut & 0.99\,/\,0.80 & 0.787\,/\,0.600\,/\,0.750 & \stk{$+0.040$}{-0.00;+0.08} & $+0.23^\star$\,/\,$-0.23^\star$\\[6pt]
anchored, uncut & 0.99\,/\,0.80 & 0.893\,/\,0.633\,/\,0.789 & \stk{$+0.027$}{-0.02;+0.07} & $+0.23^\star$\,/\,$-0.25^\star$\\
\bottomrule
\end{tabular}
\end{table}
\end{document}
