Professional Consultants...

Dec 3, 2010 at 2:34 PM

Im a recent EIT grad.  My FIRM does a lot of floodplain mapping and detention design & analysis.  Im curious what you professional consultants are discovering concerning the contributions of this application to your productivity.  Virtually all of our rainfall data used in our designs come from synthetic storms.  Rarely do I employ stream gage data. 

I would love to use the app, just dont know if its going to be worth my time. 

Hope to hear some opinions.

Jan 21, 2011 at 2:19 PM

HydroDesktop is a great discovery tool. When I was consulting, my standard augmentation to a rainfall-runoff model was to conduct a LPIII analysis on the 2 or 3 nearest gages with a "decent" record, say 15+yrs, and HydroDesktop is an ideal tool for that search. I also went to the FEMA Map Service Center to get the published flood values (despite being 20-25 years old in some cases) and finally, if the state I was working in had a USGS Streamstats site, here too, I would add the statewide regression. So then when TR-55 or HEC-HMS generated a discharge for a given rain event, I at least had some context from the other sources.  One of the biggest problem with a rainfall-runoff model is 50-yr synthetic storm is entered, it does not necessarily mean a 50-yr runoff event will occur.  I've seen relatively small rainfall events lead to large runoff events because the soil is saturated and conversely, I've seen pretty big rainfall event lead to relatively small runoff events because the ground is so dry.  Finally, I would divide all the runoff values by the drainage area at the point of interest and ask myself, is it reasonable that say 5inches of rain for a 24 hours storm would generate 200 cfs per square mile (csm)? More often than not, the answer was no. If the other three sources (gages, FIS, Streamstats) were generating a Q100 in the 80-120 csm range, and the rainfall runoff model was generating 200+csm, then it seemed as though the model was in excess of a reasonable estimate.   If the exercise is to map a floodplain, then it seems as though the 200csm estimate is much more conservative and would lead to more land falling within the floodplain, and more homeowners being required to purchase flood insurance., which may in the end be a good thing.

If you are starting out as an EIT,, I suspect that the odds of really calibrating a rainfall runoff model are pretty low. As such, I always thought the additional information provided some useful context, and asking if a model is producing reasonable results was pretty much standard practice in my experience.

Bill McDavitt