Heavy rain or flood events have recurrence intervalls. I.e. 100 year, 50 year or 25 year events. However, with increasing data and global warming and other factors these intervalls change.

How can I check wether a heavy rain event from 2015 with a recurrence intervall of 50 years still has the same intervall? Or wether it already has a 48 year intervall.

  • $\begingroup$ I doubt this is generally possible and i don't see a reason for 100/50/25 year intervals. Could you provide a source ? There are seasonally driven events connected to recurring circulation patterns that have changed fundamentally in the past 30/40 years. And intervals between significant weather events generally get shorter due to global warming. Clarification could make answering more on the point. Or interval :-) $\endgroup$
    – user18411
    Dec 8 '19 at 12:09
  • $\begingroup$ For example, I remember in 2002 in Germany there was a huge flood. On the news they called it a "century flood" because statistically a flood event with this extent happens only once every 100 years. Are there methods to figure out wether this event still happens statistically once every 100 years or perhabs now every 90 years? $\endgroup$
    – DGIS
    Dec 8 '19 at 12:20
  • $\begingroup$ I don't think so. Statistics does not say that an event happens at a certain time, it is just about the probability of an event of a magnitude happening in a frame of time and geographic extent. $\endgroup$
    – user18411
    Dec 8 '19 at 12:27
  • 2
    $\begingroup$ MAybe try "probability of precipitation" and "rainfall trends" as search terms ? $\endgroup$
    – user18411
    Dec 8 '19 at 12:43
  • 1
    $\begingroup$ You could collect historical flood events and do some sort of wavelet/frequency analysis on them, but I think you need to define the question better. There's plenty of historical weather data available, so if you can define what you mean by "heavy rain event", that data would be a place to get started. $\endgroup$
    – user967
    Dec 11 '19 at 2:18

In engineering it is common practice to build intensity-duration-frequency (IDF) curves, which basically map out the frequency of a given rainfall event based on its average intensity (mm/hr) and duration of the event. This addresses the issue of needing to classify a 'heavy' rainfall event. In general, the duration and intensity are inversely related (more intense rainfall events tend to last a shorter time).

There are many examples of calculation and software packages to actually build these curves, which are used often for engineering design to help decide the storm a given piece of infrastructure is being designed to (e.g. 5 year storm, 25 year storm, etc.). Most municipalities will have developed their own IDF curves to set standards for design.

There are also examples of online tools to help generate IDF curves for a given location, such as this one in Ontario. This tool is neat as it calculates the IDFs for a given year, and uses trend analysis to predict future changes to the IDF curve, allowing you to also pick the target year for generating the IDF curve.

If you see that the intensity/duration changes for a given frequency storm event (e.g. 50 year), that would indicate a shift in the rainfall patterns due to climate change. You may need additional statistical tests to check if these changes are statistically significant (examples of this type of analysis applied here and here). You will likely not see much change in these curves over a few years as the datasets overlap quite a bit and the trends will not change rainfall patterns in that short a time frame, however, over a 25 or 50 year period you will likely see some changes in the IDF curves, depending on the location.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.