define a set of events for Neu Darchau and compute simple characteristics

moved from #29 (closed)

changed milestone to %linking climate change and flood similarity workflow

assigned to @christine-nam and unassigned @stefan-luedtke

Scientifically speaking, one needs to keep in mind that a climate model is not driven (at the boundaries) by real-time weather, so it will not produce a flood event over an exact location at for a given time period.

What we can do is to use the physical properties (see below) of the flood event(s) of interest and search through the entire 'historical' period of the climate simulations to determine how often they occur statistically speaking. Then any comparison to the future projections would be regarding how the statistics of these kinds of events change.

By focusing on the physical properties, for example:

• Looking for combinations of X-amount of precipitation preceding a heavy (Y-amount) precipitation event within a given period. These values would come from the Flood Events identified. Translating the Flood Event properties in terms of physical variables found in the regional climate model is key. The rest is "just computations".
• First, do this for one Regional Climate model (entire historical period) and it's three climate projections (rcp2.6, rcp4.5, rcp8.5 entire projections!). While this is computationally heavy, it should be not be difficult to compute.
• Following which one can do this for an ensemble of climate models to get a robust signal.

mentioned in commit 5b88c9c9

mentioned in commit 7ead3bf3

mentioned in commit eef1e3bd