Quick question about cross sections used in flood modeling 1

[dalchri]

I apologize if this request is inappropriate for the forum.

I recently discovered that the BFE for our community is 4' above the historical highest river level crest (which was a 1% event). I though this sort of error was reasonable until I found that all the other communities upstream and downstream have BFE's either at or below the historical highest crest.

When inspecting the river cross sections near our community I saw something odd and just want to ask a general question.

Is it acceptable for either end of a cross section to terminate below BFE or must both ends make it the whole way above BFE at both banks of the river/stream?

Thank you for any advice you are willing to provide.

 

[SMIAH]

is 4' above the historical highest river level crest (which was a 1% event).

1. The historical Highest River Level might not coincide with a Base Flood Elevation.
This historical data is often used to determine the BFE (using a frequency analysis for a x% event).

2. Some juridiction ask to add a safety margin over a BFE. Some place in the world ask 1 m above it.

3. Cross section used to determine a BFE should extend to or above that level. If not, a vertical line will be assumed in most calculation and it could lead to an over-estimation of the velocity ad/or water level elevation. It might be ok in some case though (e.g. bridge with abutment).

 

[dalchri]

Thank you very much for the information SMIAH.

Error is the wrong word in describing the difference between BFE and our historical highest crest. I apologize for that mistake. I do feel that the absence of a difference of this magnitude in any of our neighboring communities warrants some investigation into the modeling though.

A follow up question if you will indulge me...

We recently had a 500 year flood from the tributary that joins the main river at our community. However, the river was barely at flood stage due to a very localized weather event on the tributary. Our community sits in a basin about the same size as the river basin, away from the river and below the tributary. During this 500 year flood event, the level of the tributary was 15' below the BFE of the river at the junction. The FIS profile for this tributary projects that it will be 20' below the BFE of the river during a 1% event.

It is well known that the river was the primary cause of the highest historical crest in our community. The river has a watershed over an order of magnitude greater than the tributary and the tributary just doesn't have enough flow to come anywhere near BFE.

My question is this: Is HEC-RAS smart enough to reduce the river level by transferring water into the large basin that is modeled as part of the tributary? I've read that a "lateral structure" or "weir" is necessary to do this and neither of these are defaults?

It just feels like the modeling is not aware of the significant volume that the river must fill. This would explain why down stream communities are matching well because the flow rates are all correct, its just that the modeling doesn't realize the river can spread out and squeeze back together at our community before heading downstream.

Thank you again for any information you are willing to provide!

 

[SMIAH]

Modeling for a Floodplain Study can vary alot.

One could use a pencil to draw the floodplain based on historical marks or vegetation or a steady state simulation using a model or/and unsteady, then 1D or 2D, Ice jams, etc.
This depend on alot of factors (e.g. data available, site of interest, person in charge, $ & time available, etc.)
Sometimes, the waterlevel might be relatively accurate but the floodplain delineation not at all due to the lack of data betweem cross sections.

You seem to describe a situation where an unsteady flow simulation would ne necessary (attenuation, storage, etc.).
Where a BFE determined using peak flow only based on a steady flow simulation AND an unsteady simulation taking the volume of the storm could lead to different results.

Volume of a flood has a major importance on the results.

 

[dalchri]

Thank you again for the insight SMIAH.

 

[beej67]

"It depends" is really the best answer you're likely to get without hiring someone to do a peer review of the study. Fortunately, doing such a peer review and looking for obvious errors in data entry or in the assumptions (particularly mannings roughness) is a fairly easy (read: affordable) thing to do, for a civil engineering consultant.

To your first question, it's typically not acceptable to cut the section off inside the ponding limits as determined by the BFE. Even as an ME with little civil background, you could add in a few extra points to the section to extend it out to where you think the section should go, based on freely available USGS topo, and rerun the model, to see the difference. It probably won't be four feet unless they've cut a lot of section out of the model, but it could be noticeable, and that could be grounds for a LOMR. You'll need a CE to hold your hand through the LOMR process though, if you go that far with it.



Hydrology, Drainage Analysis, Flood Studies, and Complex Stormwater Litigation for Atlanta and the South East -

http://www.campbellcivil.com

 

[dalchri]

Thankfully our community is nearing a FIRM review. I am hopeful that our borough will reach out to the county and maybe it will be within their purview to have the company they contracted out give the situation a second look.

It is an interesting problem. Our FIRMs are actually drawn to the historic inundation. The earliest FIRMs do not even list a BFE. Then in 1980, the BFE shows up and is a mismatch to the area covered on the map. The mismatch has persisted to this day.