Identify rain events from time series

[fenclmar]

Function for Identifying starts and ends of individual rainfall events based on (local) rain rate time series (e.g. single rain gauge).

arguments:

• rain rate (or attenuation?) time series (as xarray?)
• minimum length of dry spell separating two events
• cumulative rainfall threshold to consider interval dry (by default zero, but for longer aggregation times it might be some positive number, e.g. 0.5 mm)

output:

• table with start and end times (enabling easy sub setting of time series)

alternately the function might return:

• segmented time series in a form of list (not really modular solution)
• array of event ids having length of original time series indicating e.g. by nan (0,0,0,0, nan,nan,nan, 1,1,1, nan,nan,nan, 2,2,2, ...)

[cchwala]

Sounds good!

If I would have to decide now, I would say this is the best solution for the output:

array of event ids having length of original time series indicating e.g. by nan (0,0,0,0, nan,nan,nan, 1,1,1, nan,nan,nan, 2,2,2, ...)

I can easily be used to subset the events (using time_series_array[event_ids.where(event_number)] or something similar) or to get all of the via a groupby (not sure, but I think this will work seamlessly in xarray)