Watershed basins - peak discharges & water surface elev.

I have been assigned the task of determining a constant numerical multiplier or factor that can be used for determining the peak discharges and flood water elevations from watersheds in West Virginia, USA, or even charts or tables.

The WV watersheds are of various sizes, soil groupings, weather areas, geomorphology, etc.

I have failed to convince my supervisors that there are too many parameters associated with the broad area watersheds to allow a "common" or constant numerical multiplier or graphs/tables, etc. to be developed or established for all WV watersheds for peak discharges or flood water elevations.

I would appreciate any direction or sources of information concerning accessible published data and/or studies of basins that have tried to do the same that my bosses are requesting or sufficent information that I could present to them to demonstrate that it is not possible.

Thanks for any and all comments.
Mike Reese, Tech. Assa. III, WV

The easiest way to estimate peak flow rates is using the Regression Equations developed by the USGS for each state. Here is a link to the latest version of the National Flood Frequency software that contains all of the regression equations for the U.S.
http://water.usgs.gov/software/nff.html

Another source of the regression equation data is the West Virginia DOT drainage manual/criteria.

A quick program for estimating flood elevations after determining the flow using the regression equation is the Quick2 program from FEMA. Here is a link to where you can get the quick2 program.
http://www.fema.gov/mit/tsd/frm_soft.htm

John G. Morgan
john.morgan@cte-eng.com

Mike Reese,

It appears that your boss is talking about PI factor.

Van Duong
VDuong@Saintcharlescounty.org

To Mike Reese:

You are absolutely right. There are too many factors involved in hydrology and floodplain hydraulics (and in meteorology) to consider using any single factor to apply to several different watersheds for consistently determining peak discharges (and especially flood elevations) with any degree of accuracy. The most simplistic methods in hydrology include the Rational Method and Regression equations to determine peak discharges, but even then, the accuracy is very limited, eventhough these approximating methods may include more than one factor. Also, there are definite limits to the accuracy of such methods due to circumstances within a watershed that may not be taken into account in the method used (for example, a flood detention basin, channelization, losses in floodplain valley storage, or a gate-operated reservoir). Also, wet moisture soil conditions versus dry moisture conditions, and type of soils, and degree and type of vegetative cover (and seasonal changes) can produce a wide range of peak discharges and flood elevations. To derive a single factor (such as a drainage area ratio, for example) all watersheds considered would need to be extremely similar in all aspects of hydrologic and hydraulic parameters in order to hope for any appreciable degree of accuracy. Even the use (and continued misuse) of the Rational Method and Regression Equations have led to poor design, property damages, loss of life, and lawsuits. In my opinion, misconceived notions like these (i.e., that hydrologic and hydraulic processes can be "boiled down" one factor) is not realistic and not prudent engineering. It will only lead to problems.

There are several texts available on the subject of hydrologic processes, including several "Handbook of Hydrology", by such authors as Ven Te Chow and by David Maidment, to name a few, plus numerous other texts on water resources engineering and hydraulics.

Dwayne Stubblefield, P.E.
dwayne@nwrservices.com

Mike,

It sounds to me like you are looking for the procedures implemented in the USGS National Flood Frequency software. The software and associated documentation are available from the following site:
http://water.usgs.gov/software/nff.html

Make sure you get the documentation for the West Virginia equations from the link on that page.

The NFF methods may or may not be applicable to your watersheds depending on their parameters (such as size, etc.), and depending on the purposes of your study. Pay attention to the standard errors of the results - they can be quite large.

The NFF software does not compute water surface elevations, but the flows can be used in hydraulic computations using HEC-RAS or other software tools.

To my knowledge a method applying a multiplying factor to area as the only variable in a linear equation for predicting streamflow has never been in accepted use, but perhaps others on this list have longer memories than I do. The West Virginia equations implemented in NFF use area with a variable exponent so that:

Q(n) = C*A^e

where

Q(n) is the flow for return period n
C is a variable factor, and
e is a variable exponent

The values of C and e vary depending on return frequency and hydrologic
region.

Paul Bijhouwer, PE
Paul.Bijhouwer@parsons.com

Mike,

You are definitely between a rock and a hard place. You are also correct in your argument against a "one size fits all" approach. The more "generic" you make your investigation, the larger the probability for error in your calculations.

However, assuming that your arguments against such an approach will continue to be debated, you might take a look at the USGS Regional Regression equations to get an estimate of your flows. These divide WV into three areas and give equations for small and large watersheds. The breakpoints for small and large are contained in the text. While this is a "broad brush" view, it will at least give you an order of magnitude. However, the text itself states that "The standard errors of estimate of the regression equations range from 37 to 54 percent". There are additional stipulations as to the validity of the results, so I would recommend viewing the entire report before using any of the results. Again, this will only give you a broad brush view of the flow quantity and will not give you a flow depth. Flow depth will be based on channel geometry, slope, roughness, et. al. There are several reasonably simple equations to use (you can probably locate one in a hydraulics book). I can think of no-way to estimate flood depths without doing any analysis on the channel; you will have to do some cursory analysis even at a 50,000 foot view.

The report and computer program (National Flood Frequency program) is available from the USGS at the following URL:
http://water.usgs.gov/software/nff.html

There are several sections of the report, including state specifics, so be sure and review everything carefully before using your results.

Good Luck!
Jeff Brizendine, P.E.
jbrizendine@burnsmcd.com

You might want to look at the USGS's National Flood Frequency program.
http://water.usgs.gov/software/nff.html

Good luck
Ron Kilmartin
ronkilmartin@attbi.com