rnugent3
commented
1 year ago We're going to drop any notion of using exterior-interior functions for this calculation. It does not make sense that it should be used. 2D hydraulic modeling will have to be done to model interior drainage problems or levee flanking, right, @Brennan1994? We're talking about overland flow. In which case, we should just use that 2D hydraulic modeling.
@codymccoy please disregard the exterior-interior stuff I noted in the screen shots I shared. However, I will need you to not allow the user to select an exterior-interior function if they have checked the box to include non-fail risk in the compute.
For impact areas with levees, we need a way to calculate total risk, which is the sum of breach risk and non-breach risk, but the HEC-FDA methodology assumes that non-breach risk is zero.
Why this matters: there are plenty of cases where interior drainage issues result in damage before the levee is overtopped. Take for example a likely flood scenario in Sacramento from the American River. I might have the details wrong but the premise is correct. The left bank of the river just north-east of midtown is expected to be the first place to overtop. The flooding would inundate south sac and pocket, while the pocket and south land park levees may still be in tact. In the current Freeport example, a levee is expected to be flanked before it is overtopped.
When interior drainage issues have been serious, users find a way to use our square peg to fit their round hole. Instead, we need to do this calculation internally.
This will involve accepting a second pair of stage-damage functions as part of the scenario compute, so that we can do breach-scenario and non-breach scenario. We'll need to know if the user plans to model non-breach risk (and make sure to communicate in the results). Interior-exterior functions are possible for both scenarios. Interior-exterior functions simplify the analysis by not requiring additional hydraulic and stage-damage computes (although some hydraulic analysis must occur to develop the interior-exterior function).
The change to the computational code is straightforward. All existing code should stay the same. We add logic for the scenario where a user wants to model non-breach risk, and if so, apply total probability.