How much oscillation in flow is acceptable, if the model does not go unstable & oscillations resolve themselves?

[Geo_Wick]

For dam breach simulations specifically. For example, this breach hydrograph:

To me, it doesn’t seem “that bad” since the oscillations are relatively small, the model is stable, peak flow is appropriate for the given breach parameters, and the general shape of the hydrograph appears typical. But, is that wishful thinking?

Unfortunately I do not have access to the model files, only the hydrograph.

[cameron]

The question would be, due the oscillations impact results downstream. As the hydrograph travels downstream, it will attenuate and volume from the hydrograph is more important than the peak flow.

You could try a test where you either cut off the hydrograph at say time 0.4 hours or adjust the flow after 0.4 hours and run the simulation. Then do a test with the unmodified hydrograph to see how results compare. If the depths are within a few tenths or arrival time is similar, then probably not a big deal.

[Geo_Wick]

Ah, that makes sense. The total volume released is appropriate (manually integrated, I don’t have access to the model files, only selected results).

Excellent suggestion to modify the hydrograph to see if the downstream areas are affected, I am keeping that one in my back pocket!! In this case, the downstream area is mainly a channel in a deep canyon. Usually I see maximum depths at the peak of the hydrograph for that type of terrain, so I’m not too concerned about depth increasing after the peak flow has passed.

In general, would you say that if “minor” oscillations occur after the peak flow, one can reasonably assume the peak flow is “accurate”? i.e., if I took the time to fully stabilize and resolve the oscillations, do you think I would be likely to see the same or very similar peak flow in the modified model?

[Author]

Posts