Comment on Article by Bradley, Holan, and Wikle.

We provide a discussion of the article "Computationally Efficient Multivariate Spatio-Temporal Models for High-Dimensional Count-Valued Data" by Bradley, Holan, and Wikle. In our opinion, this work constitutes a major contribution to the field of spatio-temporal statistics and contains distribution theory that should be broadly applicable. In this note, we reflect on modeling decisions made by the authors. We include a small set of simulation results to illustrate the effect of one aspect of the proposed model.