1

We know that a saturated measure is a measure space $(X,\mathcal{M},\mu)$ such that $\mu:\mathcal{M}\rightarrow [0,\infty]\;$ and $\;\mathcal{M}=\widetilde{\mathcal{M}}\;$

where

$\;\widetilde{\mathcal{M}}=\{E\subseteq X\;|\;E\cap M\in \mathcal{M}\;,\;\forall M\in\mathcal{M} \;\text{with}\;\mu(M)<\infty\}\;$ is the set of all locally measurable sets.

Are Haar measures defined on borel $\sigma$-algebra ${\frak{B}}_G$ of locally compact group $G$ saturated?

$\textbf{EDITED:}$

I found a solution, But I doubt whether it is true or not! Can anyone tell me if my solution is correct or not?

$\textbf{my solution:}$

if ${\frak{B}}_G$ be borel $\sigma$-algebra on $G$, then we know that ${\frak{B}}_G\subseteq\widetilde{{\frak{B}}_G}$ and it's enough that prove $\widetilde{{\frak{B}}_G}\subseteq {\frak{B}}_G$.

let $E\in \widetilde{{\frak{B}}_G}$, then we have $E\cap M\in {\frak{B}}_G$ for every $M\subseteq G$ with $\mu(M)<\infty$. since Haar measures are $\;$decomposable$^*$ (for definition, see below ($*$) and for proof, see Haar measures are decomposable), then according to (i) in below $(*)$ there exists $\{X_i\}_{i\in I}\subseteq {\frak{B}}_G$ with $\mu(X_i)<\infty$. since $\mu(X_i)<\infty$, $\forall i\in I$ and $E\cap M\in {\frak{B}}_G$ for all $M$ with finite measure, then we have $E\cap X_i\in {\frak{B}}_G$, $\forall i\in I$ and according to (iii) in below $(*)$, we have $E\in {\frak{B}}_G$. hence $\widetilde{{\frak{B}}_G}\subseteq{\frak{B}}_G$ and $\widetilde{{\frak{B}}_G}={\frak{B}}_G$.

$\color{\red}{\bigg[}$actually, I think that is proved that every decomposable measure is saturated.$\color{\red}{\bigg]}$

$\textbf{is my answer correct?}$

$(*)$ A measure space $(X,\mathfrak{M},\mu)$ is decomposable if:

(i) $X$ is a disjoint union of measurable subsets, $X=\bigcup_{i\in I}X_{i}$, with $\mu(X_{i})<\infty$ for all $i$.

(ii) $\mu(E)=\sum_{i\in I}\mu(E\cap X_i)$ for every measurable set $E$ of finite measure.

(iii) if $E\subseteq X$ and $E\cap X_i\in \frak{M}$ for all $i$, then $E\in\frak{M}$.

  • 1
    You want $E\subset X$ for $X$ the measure space, not $E\in \mathcal{M}$, right? Otherwise this is tautological. – SomeCallMeTim Feb 20 '23 at 12:47
  • 1
    The previous commenter means that you presumably want $;\widetilde{\mathcal{M}}={E\subset X;|;E\cap M\in \mathcal{M};,;\forall M\in\mathcal{M} ;\text{with};\mu(M)<\infty};$. $\sigma$-algebras are closed under intersection so that for every $E, M \in \mathcal{M}$, $E \cap M \in \mathcal{M}$. This means that as currently written in the question, all measures are saturated. – Rhys Steele Feb 20 '23 at 14:44
  • Perhaps spell it out more. $G$ is a locally compact group. $\mu$ is a Haar measure on $G$. So we could have either $\mu$ is defined on the Baire sets, or $\mu$ is regular and defined on the Borel sets. Then do we use Caratheodory's definition to define measurable sets $\mathcal M$? What happens when you try to prove this is saturated? – GEdgar Feb 20 '23 at 14:46
  • thanks, I revised this problem. actually we know that $\mathcal{M}\subseteq\widetilde{\mathcal{M}}$ for arbitrary $\sigma$-algebra $\mathcal{M}$ in measure space $(X,\mathcal{M},\mu)$. now problem is that: is this true that ${\frak{B}}_{G}$=$\widetilde{{\frak{B}}_G}$ for borel $\sigma$-algebra ${\frak{B}}_G$ in haar measure space $(G,{\frak{B}}_G,m)$?$;$ where $G$ is a locally compact group. – Amirhossein Haddadian Feb 20 '23 at 15:43
  • 2
    I’m voting to close this question because this is at least the fourth question by same OP in the resent days that came up while reading Folland and not showing any self-effort. – Kurt G. Mar 22 '23 at 07:42

0 Answers0