|
<br />-
<br />
<br />APPENDIX B -- DESCRIPTION OF THE NATIONAL
<br />FLOOD FREQUENCY PROGRAM
<br />
<br />The National Flood Frequency (NFF) computer program
<br />evaluates regression equations for estimating T-year flood-
<br />peak discharges for rural and urban watersheds, As many as
<br />7 multiple regression equations (2-, 5-, 10-, 25-, 50-, 100-,
<br />and 500-year) are defined for each of 200 or more flood
<br />regions. Methods are also available for estimating a typical
<br />flood hydrograph corresponding to a given T-year peak
<br />discharge,
<br />
<br />The NFF computer program is composed of two components
<br />.. a state-by-state data base of regression coefficients, stan-
<br />dard errors, etc, for about 1,500 multiple regression equa-
<br />tions and a calculation routine for rural and urban flood
<br />characteristics including tabling and graphing capabilities,
<br />The fonnat of the state-by-state data base is described below.
<br />As noted earlier, the NFF program is written in the "C" pro-
<br />gramming language and is designed to run on a microcom-
<br />puter with at least 640K bytes of user memory,
<br />
<br />Figure B I is a flow chart of the NFF computation options. A
<br />State may be selected by a two-<:haracter code. Each State
<br />will have from I to 12 flood regions, When a flood region is
<br />selected, the program will prompt the user for the required
<br />watershed and climatic characteristics and other infonnation
<br />to make the flood computations. Options include the compu-
<br />tation of regional regression estimates of the rural flood-peak
<br />discharge for a given station, computation of a weighted esti-
<br />mate of the station and regional estimates (if equivalent
<br />years of record are provided for the regional equations),
<br />computation of urban flood-peak discharges for a given sta-
<br />tion, the ability to plot and save any computed frequency
<br />curve, computation of a flood hydrograph corresponding to a
<br />given T-year peak discharge, and the ability to plot and save
<br />the computed flood hydrograph. The normal sequence of
<br />these computations and plots is shown in figure B I.
<br />
<br />An example of a logfile showing the sequence of questions
<br />and input data needed for computing a flood-frequency curve
<br />for the Fenholloway River near Foley, Aorida is illustrated
<br />in figure B2. As can be determined by inspection of figure
<br />B2, the Fenholloway River near Foley watershed is con-
<br />tained in one hydrologic region - Region B. The NFF pro-
<br />gram numbers the regions numerically so Region B is
<br />identified in NFF as hydrologic region 2. The watershed
<br />characteristics input by the user are Drainage Area = 120
<br />square miles and Lake Area = 0.37% of the drainage area,
<br />The watershed of interest is contained in maximum flood
<br />region 3 as defined by Crippen and Bue (1977) and shown
<br />earlier (fig, 3), The Maximum Aood Envelope value of
<br />101,000 cubic feet per second is an estimate of the maximum
<br />
<br />flow ever experienced for a 120-square-mile watershed in
<br />Crippen and Bue's flood region 3,
<br />Given the above input values, a rural flood-frequency curve
<br />is then computed and a table of flood-frequency values, stan-
<br />dard errors of estimate and equivalent years of record are dis-
<br />played in figure B2. The flood-frequency curve was
<br />computed without using the 500-year equation, therefore the
<br />500-year value shown in figure B2 was detennined by
<br />extrapolation as defined in the section entitled Estimation of
<br />Extreme Hoods. The regional flood-frequency curve is
<br />shown earlier (fig, 2). In reality, 500-year equations do exist
<br />for Aorida (Bridges, 1982) and the extrapolated 500-year
<br />flood was compared to the value computed from the pub-
<br />lished 500-year equation, This example was provided to
<br />illustrate the applicability of the 500-year extrapolation pro-
<br />cedure.
<br />Finally, NFF allows weighting of observed and regionall
<br />regression flood estimates (if equivalent years of record
<br />available), computation of a urban flood-frequency curve,
<br />plotting a flood-frequency curve, and computation of a flood
<br />hydrograph, In figure B2, the response N (no) was provided
<br />for all these questions.
<br />The flood-frequency curve ordinates and the flood-
<br />hydrograph ordinates can be output to a flat file for further
<br />analysis with another program.
<br />The flood-frequency curve ordinates are output in the fol-
<br />lowing fonnat:
<br />
<br />Nltion.1 Flood Frequency Program
<br />
<br />Flood Frequency Data
<br />Date: 091211199410:30
<br />Basin: Hypothetical River near Example
<br />Consult the log file for the input data
<br />Recurrence Rural
<br />Interval, years Discharge
<br />2 8120
<br />5 13200
<br />10 17400
<br />25 22100
<br />50 26700
<br />100 29800
<br />500 39900
<br />
<br />The flood-hydrograph ordinates can be output in two formats
<br />.. generic and HYDRAIN, The generic fonnat is simply a
<br />listing of the time, in hours, since runoff began and the cor-
<br />responding discharge for selected points on the hydrograph,
<br />The points are listed from 0,25 LT (lagtime) to 2.40 LT in
<br />increments of 0.05 LT. An example of the file follows:
<br />
<br />190 Nationwide Summary of U.S. Geologlcol Survey Raglonal Regression Equatlona for Estimating Msgnltude and Frequency of
<br />Flood. for Ungaged Sit.., 1993
<br />
|