A compactness theorem for approximating the invariant densities of higher dimensional transformations (Q2367659)

Let \(I=[0,1]\). If \(\tau: I\to I\) is a piecewise expanding transformation \textit{A. Lasota} and \textit{J. A. Yorke} [Trans. Am. Math. Soc. 186, 481-488 (1974; Zbl 0298.28015)] proved the existence of an absolutely continuous invariant measure \(\mu\) with respect to Lebesgue measure. This result has subsequently been generalized to higher dimensions in various ways. If \(f\) is the density of \(\mu\) with respect to Lebesgue measure on \(I^ n\), then \(f\) is the fixed point of the Frobenius-Perron operator, but in general it is very hard to solve the resulting functional equation for \(f\). The aim of the present paper is to produce a new method of approximating invariant densities. The authors consider a particular class of transformations which they call \textit{Jablonski} transformations. These are defined on rectangular partitions of \(I^ n\) with the property that on any element of the partition each component of the transformation depends on only one variable. With a piecewise expanding Jablonski transformation the authors show how to construct associated \textit{Markov} transformations. Their main results are that the collection of invariant densities associated with these Markov transformations is weakly compact in \(L^ 1\) and that any weak limit point of norm one is actually an invariant density for the original transformation. This gives the desired approximation.