Help decoding the Met Office Data re:2012 extreme rain

[Chimune]

Can someone who undertands these things, please explain what the graph in the Met Office press release is showing:

http://www.metoffice.gov.uk/news/releases/archive/...

Defra says:

"changes in climate extremes are based on the 99th percentile. Assuming each season is 100 days long, the 99th percentile roughly corresponds to 1 day per season. Hence, this variable can be thought of as the change in the precipitation of the wettest day of each season"

The Met Office article states that the extreme event - originaly expected every 100 days, is now expected every 70 days.

Yet that event is - "the change in the precipitation of the wettest day of each season".
How can 'the wettest day of the season' happen more often than it used to ????


What is the vertical axis actually measuring ?
My brain hurts....

[Chimune]

trashbat said:

Forget about the Defra bit, as it means you're trying to conflate two definitions.

The Met site says, under the graph, "frequency of what climate averages tell us should be roughly 1 in 100 day heavy rainfall events in each year". I imagine this means that their '1 in 100 day level' is historically based, and not about the current 100 days.

They're saying that reality shows those events happening more than the model.

But the vertical axis isnt frequency - it looks like percentile. Someone else said it was probability...

[Chimune]

Ok, so what is the 99th %ile in this graph?