1: \begin{abstract}

2: We study variants of one-dimensional $q$-color nearest-neighbor voter models in discrete time. In addition to the usual voter model transitions in which a color is chosen from the left or right neighbor of a site there are two types of noisy transitions. One is bulk nucleation where a new random color is chosen. The other is boundary nucleation where a random color is chosen only if the two neighbors have distinct colors. We prove under a variety of conditions on $q$ and the magnitudes of the two noise parameters that the system is ergodic, i.e., there is convergence to a unique invariant distribution. The methods are percolation-based using the graphical representation of the model which consists of coalescing random walks combined with branching (boundary nucleation) and dying (bulk nucleation).  

3: \end{abstract}

4: