When the return-interval describes the expected recurrence frequency for a single site/region, we must expect many more occurrences over a wider area. This is explained by maths and mathematics (the Binomial distribution and probabilities). Then a small change in the probabilities (probability density function) can lead to a large increase in the number of observed events. Furthermore, the definition of climate change is indeed a changing probability density function (climate is weather statistics), which means that a past 500-year event is no longer a 500-year event, but perhaps a 100-year event. In other words, this is not surprising and is in accordance with mathematical reasoning. Actually, it is to be expected, especially since a warming leads to a higher evaporation rate and more moisture in the atmosphere. The fact that the return intervals are estimated for single sites/regions means that we can expect a dramatic increase in similar extreme weather events in the future. We can gauge this development by studying the number of record-breaking events: see http://onlinelibrary.wiley.com/doi/10.1029/2008EO410002/pdf