Multiple Day Extreme Events
Clarification on using average vs consecutive when projecting multiple day thresholds
This note is included for the record only as it is useful for understanding the workings of the extreme event tool.
In the extreme value analysis box for Tmax, you have a choice of two buttons under the “multiple sample event definition” option.
Query from a user - the left button gives the return period “x” days where the average between the two days is over threshold temperature. The right gives the return period where the individual temperature for where each of the “x” consecutive days is over the threshold temperature.
In this way, with 35 degree days if there were 2 consecutive days 32 and 39 degrees ((32+39)/2=35.5), they would be picked up by the left button but not the right.
Because we are interested in only days over 35 degrees in this case I should use the right button?
CLIMsystems:You are right on all accounts. I would probably do both, as heatwaves are often defined as the average high temperature over a period of time. In the "consecutive days" approach you could have 5 days with the first and last 2 (way) over 35, and the middle just below, which would not be picked up as an extreme event.
Anecdotally – for Sydney the return period using the average above was half (twice as frequent) of that given by using the “on consecutive” (absolute) button.