# Gamma Convergence of functionals on Probability measures

Would be grateful if someone could provide a hint or an appropriate reference for the following.

Notation:

$\mathcal{P}(\mathbb{R}^n)$- Space of probability measures on $\mathbb{R}^n$

$g_n\stackrel{\Gamma}{\longrightarrow}g$ denotes gamma convergence of appropriate space

Notion of $\Gamma$ convergence: $\Gamma$ convergence is a form of convergence for functionals. For definition see http://en.wikipedia.org/wiki/%CE%93-convergence

Question

Let $(\rho_n)$ be a sequence in $\mathcal{P}(\mathbb{R}^n)$ and $f_n:\mathbb{R}^n\rightarrow\mathbb{R}$ a sequence of functions. Define, $F_n(\rho):=\int\limits_{\mathbb{R}^n}f_n(x)\rho(dx)$.

Prove that: $f_n\stackrel{\Gamma}{\longrightarrow}f$ in $\mathbb{R}^n \Longleftrightarrow F_n\stackrel{\Gamma}{\longrightarrow}F$ in $\mathcal{P}(\mathbb{R}^n)$.

Remarks: (1) I am looking to use narrow convergence of probability measures ($\rho_n$) in context of proving $\Gamma$ convergence of $F_n$.
(2) This usually pops up in existence of solution to some Fokker-Planck type of equation. More specifically it pops in Sandier-Serfaty framework for Gradient flows.

-
This is a friendly suggestion. Perhaps you could provide more details about the notion of $\Gamam$-convergence. It is not such a popular concept, and maybe some of the MO raders know it under a different name. –  Liviu Nicolaescu Mar 24 '13 at 11:27