Topographic potential for preventive and decentralized flood measures

[Max Brunke]

Dear SWAT Community, 

i would like to share my ideas / workflow on a study that I am currently planning. 

My Catchment will be the low mountain Rur-Catchment in Germany (approx. 2.360 km^2).

Working with different scenarios and comparing them to a baseline run i would like to evaluate the potential of my study area to reduce Extreme Runoff events implementing preventive flood measures. 

At first i would like to examine the topographic potential of the catchment. 

Mainly i am thinking about identifying natural depressions and parameterising them. 

In other words, assuming that all depressions in the catchment area, or their summed area and maximum storage volume, are available for the application of flood measures.

My first idea for a simple flood measure scenario looks like this and i would love some feedback:

(1) Topographic analysis of DEM: Identifying all local Depressions

(2) Calculating maximum area and storage volume for each subbasin

(3) Using area and volume to parameterise a lumped pond/wetland for each subbasin

(4) Maybe allowing some high Sat_K values for additional infiltration to simulate gw-recharge

I think the general and possibly somewhat abstract purely topographical natural potential of my catchment area for implementing flood protection measures can be investigated quite well in this way. 

But since i wasn't able to find a whole lot of studys trying similar methods i am somewhat insecure.

What are your thoughts?

[Natalja C.]

Hello,

For any flood-related studies I would use dedicated hydraulic models (not hydrological). Maybe a coupled hydrological-hydraulic model combo.

Anyway, how would you parametrize these "ponds"? What data would you use as calibration/validation.

Although your described method might work as a "screening', I would still be very cautious when using the "standard" sub-basin discretization. A hillslope would work better, but that one is not a part of the "standart" tools, like Arc/QSWAT.  Meaning you would need to put in extra work in development of custom tools to achieve such discretization.

Anyway, you may try, but I would very carefully assess how the model simulates these processes and adjust my model setup to be as accurate as possible.

Best,

Natalja