Stochastic equations, flows and measure-valued processes (Q414291)

Strong existence is proved for two classes of stochastic flows. The first, flows of continuous-state branching processes with immigration, is defined by stochastic equations of the form \[ \begin{multlined} Y_t(v)= v+ \sigma \int^t_0 \int^{Y_{s-}(v)}_0 W(ds,du)+ \sum^t_0 [\gamma(v)- bY_{s-}(v)]\,ds +\int^t_0 \int^\infty_0 \int^{Y_{s-}(v)} z\widetilde N(ds,dz,du),\end{multlined} \] where \(W\) is a white noise on \((0,\infty)^2\) and \(\widetilde N\) a compensated Poisson random measure on \((0,\infty)^3\). The second, generalized Fleming-Viot flows, is given by equations of the form \[ \begin{multlined} X_t(v)= v+\sigma \int^t_0 \int^1_0 [1_{\{u\leq X_{s-}(v)\}}- X_{s-}(v)] B(ds,du)+ b \int^t_0 [\gamma(v)- X_{s-}(v)\,ds +\int^t_0 \int^1_0 \int^1_0 z[1_{\{u\leq X_{s-}(v)\}}- X_{s-}(v)]\,M(ds,dz, du),\end{multlined} \] where \(B\) is a white noise on \((0,\infty)\times(0,1]\) and \(M\) a Poisson random measure on \((0,\infty)\times(0,1]^2\). For the latter, scaling limit results are obtained, which lead to sub-critical branching immigration superprocesses. The work is related to a series of papers on stochastic flows associated to coalescent processes by \textit{J. Bertoin} and \textit{J.-F. Le Gall} [Probab. Theory Relat. Fields 126, No. 2, 261--288 (2003; Zbl 1023.92018); Ann. Inst. Henri Poincaré, Probab. Stat. 41, No. 3, 307--333 (2005; Zbl 1119.60024); Ill. J. Math. 50, No. 1--4, 147--181 (2006; Zbl 1110.60026)]. Using techniques differing from theirs, some of their treatments of flows are unified and their limit theorem for finite point motions of flows is generalized.