For my thesis, the lecturer proposed to make a "sensitivity analysis" with a 2D surface model connected with a 1D drainage Network model for a storm rain event. However, I'm having problems to figure out what that means or could mean.

My study area is about 8km² large. I have a high resolution DEM with 1m grids. I have a weather station about 1-2km from my catchment which measured the precipitation of the storm event in mm/h for every 5min, I have a detailed and current land use map of the area, I have a station which measured the water level at the river and Information on the drainage network.

So, I understand that sensitivity analysis basically means to figure out how the model results change when I modify the input data.

So I guess I could look at precipitation variablility. I could check how the model reacts to diffrent heterogeneous rainfall distributions or diffrent rain intensities. But is that a good approach? The study area seems small, and there are no hills that could cause big differences to the weather station. There shouldn't be a noticable diffrent rain intensity and spatial distribution shouldn't it?

I could change the DEM and see what happens but why would anyone use a coarser DEM if a high resolution DEM is available?

I can play around with the roughness but landuse data is very good and using the manning strickler tables for roughness are standard.

I feel like nothing makes sense. Does anyone have some advice on what kind of sensitivity analysis I could do?

  • 1
    $\begingroup$ 1D drainage model is ok when your basin has no time evolution (you can consider a fix raining) and you have a unidirectional flux. You should have a good Manning characterization and one unique flood plain. But if you have a big flood plain or multiples channels or channel confluence or a non-regular channel section, then you should go for bi-dimensional. It is not a matter of size, is a matter of hydraulic complexity. $\endgroup$ Dec 2 '19 at 13:10

Your Answer

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

Browse other questions tagged or ask your own question.