[Mark]

Hi Scott.

Awesome work. This knowledge really takes off what would have been a load of effort to get realistic D*V max values. As I don’t have spatial analyst, I went with QGIS and classified my D, V and D*V grids between 1 and 5 according to the FEMA document’s hazard graph, and then combined them preserving max values in each to derive the composite hazard as Mike mentioned. So I think I’m good.
For my needs the FEMA approach is fine which benefits me as I don’t need to use the D(V+0.5) approach from the UK.

Thanks again for checking that RAS uses the correct approach to D*V.

[Mark]

Thanks to Scott, Cameron and Mike for your replies.

One final check with regards to DxV – Scott, you mentioned that RAS uses stored maps (max depth and max velocity). When adding a DxV grid to results, I have no ‘stored maps’ other than default (D, V, WSE) which aren’t necessarily set to max. So when calculating DxV, how do I know that its using max values independent of time step?
The manual only says “the maximum value is the maximum of depth times velocity” based on the user mapping interval.

It would be great for some inbuilt results calculator to create grids like hazard based on interrogation of all time steps. Until then I will go with what you have recommended – batch raster calculations for depth and velocity outputs for all time steps (when I need detailed), and a dual hazard criteria using DxV and D when not as much detail is necessary (with a caveat of course).

[Mark]

Hi Luis

Thanks for that feedback. Yes – that makes sense. It may not be an option in my study due to the substantial variation in elevation… but good to know for future model builds. Could even be worth double checking cell faces on a slope grid.

[Mark]

Thanks Scott. The 5m “low hazard” comes from the next page (Table 1) in the document you referenced.
Replicating the hazard classification in Figure 31 or applying the Formula (D*(V+0.5)+Debris Factor) would be ideal – but I don’t know how to do that. In the other thread you responded about using basic logical conditions?

I understand the raster math approach. My problem is extracting a relevant velocity and depth grid that represents maximum hazard, since max hazard for a cell wouldn’t be specific to a single time step. I suppose you could export depth and velocity for each time step and make individual time step hazard grids and then merge max values. Problem is that it would be exceptionally labour intensive if you have hundreds of time steps.

Any advice?

[Mark]

I’m just promoting this as its a good question which i don’t have the answer for

[Mark]

Thanks Jarvus

By wave action i mean modelling the impact of wave over-topping (not just a tidal level that steadily fluctuates but the influence of waves on top of the tidal level)?

[Mark]

Thanks for that input Luis

Is the conclusion then that you cannot create a multi domain (distributed) 2D model using land-cover/soils as your domain (mesh) divider and that you HAVE to go the hydraulic route where your outflow boundary for an upslope 2d area is higher or the same as the inflow boundary of the downslope 2d area across the full length of the SA/2D Area Connector?

If so, this will be problematic in modelling precipitation derived surface water flooding of a large site since hydrologically homogeneous areas which would have the same effective infiltration (total rain less infiltration and depression storage) would either need to be lumped into ‘hydraulic areas’ thereby losing spatial detail or otherwise broken down into much smaller units.

[Author]

Posts