Referee 2's Review (Dave Smith):

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication the manuscript has been revised in light of these comments and to address other editorial requirements.
**********

This paper looks at edge effects in the geography of measles dynamics. Human populations have a metapopulation structure, measles dynamics are stochastic, the patches are coupled by the movement of measles cases, and the strength of coupling between patches is determined by a gravity model. These have been explained elsewhere in detail, and they have been deployed in a straightforward and justifiable way in this manuscript.

The paper does two things to look at edge effects. First, it creates artificial landscapes to simulate edge effects; a clever design allows the authors to compare stochastic fadeout on edge cities with inland cities that are comparable in every other way, except that they are on the edge. The experimental design is novel, but I'm convinced that the design of it was sound. Second, it looks at actual stochastic fadeout patterns in the UK measles reporting data. The dataset here has been explored in other papers in similar ways, and the methods are, again, sound.

The findings of the paper are very interesting. To put it in colloquial terms, coastal cities are slightly heavier than their inland counterparts, presumably because they are tourist destinations. The observed pattern of stochastic fadeout in coastal cities doesn't seem to be strongly affected by the seasonal migration out of cities to the coast. Thus, being on the edge would tend to isolate cities, except that they find themselves to be less isolated than expected because they are near the oceans. The paper illustrates how this is played out for the dynamics of measles. '