[Math] “Support” of a measure

measure-theory

  1. I was wondering if there is some
    definition for "support" of a
    measure in the sense that one or
    both of the following can be true:

    • one measure is absolutely continuous with respect to
      another measure, if and only if the support
      of the former is inside the
      support of the latter?
    • two measures are mutually singular (as in Rudin's Real and
      Complex Analysis), if and only if the
      supports of the two measures are
      disjoint?
  2. The definition for support
    of a measure in Wikipedia
    relies
    on that the measurable space is also a
    topological space. I would like to
    know if it makes sense to define
    support of a measure solely on a
    measurable space?

Thanks and regards!

Best Answer

  1. No way: Think of Lebesgue measure and a point measure.

  2. You can often speak of the support up to a null-set, but in order to single out a specific support you need further structure on your measure space.

Added: In 2. I was deliberately a bit sloppy. A sufficient condition for the existence of a good notion of support is a class of null sets closed under arbitrary unions. The union of those null-sets is then the largest such $\mu$-null set and its complement deserves the name of support of $\mu$. For instance, if $\mu$ happens to be a Radon measure on a locally compact space, you can take the class of open $\mu$-null sets and the union of those is precisely the complement of the (closed) support of $\mu$.