I'm doing a sensitivity analysis for my model, but I'm not sure what simulation time window I should use. I'm using unsteady flow analysis, with diffusion wave equation and time step based on courant. Since I'm doing the sensitivity testing I'm using the unsteady flow model with steady state flow conditions (a hydrograph with a constant flow value over time). This means I'm making the hydrograph myself, and it is not a real incident (meaning I have to create the starting and ending time myself). The model is going to be used for flood mapping.
Should the time window be long or short, or doesn't it matter? Does anyone have any experience on this field?
Re: Simulation time window for sensitivity analysis
The model run should be long enough for the results to stabilize. For example, if you are simulating for say 48 hours and you don't see changes in your water surface elevations for the last 24 hours then you can shorten your simulation time window.