PROGRAM USER GUIDE FOR RECESS
By A.T. Rutledge
CONTENTS
INTRODUCTION........................................................................................................................................2 PURPOSE..................................................................................................................................................2 BACKGROUND ABOUT RECESSION CHARACTERISTICS.............................................................2 GENERAL DESCRIPTION OF THE METHOD......................................................................................3 PREPARING TO USE THE PROGRAM.................................................................................................5 OBTAINING STREAMFLOW DATA FILES..........................................................................................5 THE SCREEN PROGRAM.......................................................................................................................7 THE STATION.TXT FILE........................................................................................................................8 RUNNING THE RECESS PROGRAM.....................................................................................................9 GRAPHICAL DISPLAY OF RECESSION SEGMENTS WITHOUT MATHEMATICAL ANALYSIS9 DETERMINATION OF THE RECESSION INDEX..............................................................................11 THE MASTER RECESSION CURVE....................................................................................................12 Output Files for the RECESS Program............................................................................................12 The CURV Program...........................................................................................................................14 ADDITIONAL CONSIDERATIONS FOR PROGRAM USE................................................................16 REFERENCES...........................................................................................................................................18
The program user should have these files…. curv.exe Holiday.txt Indian.txt index.txt ranges.txt recess.exe recplot.exe recsum.txt screen.exe station.txt winter.ico
Note –
The RECESS program was first adapted for reading web data in 2002. In 2007,
adaptations were made in the code to accommodate changes in the way USGS web data handle gaps in streamflow record. This newer version of the program was posted on the RECESS web site in January 2007. The newer version can read data in the new or old format.
1
INTRODUCTION
PURPOSE
The RECESS program is used to determine the recession index and to define the master recession curve (MRC) from analysis of streamflow records. This manual describes a version of the program that reads data in a format that is available from USGS Web sites. This manual also describes the use of a graphics application RECPLOT which can be used to view recession segments while the user is executing RECESS. RECESS can be used independently of this application which requires the windows operating system. This manual does not include complete explanation of the computer program; the reader is referred to the earlier report (Rutledge, 1998, pages 9-32).
BACKGROUND ABOUT RECESSION CHARACTERISTICS
The method described in this manual is intended for analysis of streamflow recession during times when no ground-water recharge is occurring, when all flow is from ground-water discharge, and when the profile of the ground-water head distribution is nearly stable. According to a model developed by Rorabaugh (1964), recession will be linear on a graph of the logarithm of flow as a function of time. The rate of recession can be expressed as the recession index, in days per log cycle. This variable is needed when the “Rorabaugh Method” is used to estimate recharge (Wilder and Simmons, 1978; Bevans, 1986; Gerhart and Lazorchick, 1988; Hoos, 1990).
Although the linear model of recession can be a reasonable approximation, there are a number of processes that will effect recession-curve shape, some causing departures from linearity (Horton, 1933; Werner and Sundquist, 1951; Riggs, 1964; Ineson and Downing, 1964; Toebes and Strang, 1964; Ibrahim and Brutsaert, 1965; Singh, 1969; Hornberger and others, 1970; Singh and Stall, 1971; Trainer and Watkins, 1974; Daniel, 1976; Nutbrown and Downing, 1976; Anderson and Burt, 1980; Petras, 1986; Szilagyi, 1999, Rutledge, 2002). Wood and others (1972) developed nonlinear master recession curves (MRC’s) for several streams in eastern Pennsylvania, and Rutledge and Mesko (1996) used the RECESS program to derive a variety of MRC shapes in the Appalachian and Piedmont regions of the eastern United States. The description of recession can have a variety of applications in hydrologic studies, including the evaluation of aquifer properties and the sustainability of ground-water discharge.
Theoretically, characteristics of flow recession can be evaluated on the basis of ground-water-level information. This may require designation of the altitude of the outflow boundary. Because this altitude can vary along the length of the boundary (the stream), ground-water-level recession analysis may lead to a variety of results (Rutledge, 2002; 2003).
2
GENERAL DESCRIPTION OF THE METHOD
The RECESS program is an empirical method for describing recession characteristics. A general description follows. The program scans the streamflow dataset, finding periods of continuous recession; such as the period from day 93 to day 121 in figure 1. When a segment is found, the program user can decide whether it should be analyzed and, if so, which parts represent near-linear conditions (on the semi-log plot) that can be used to quantify the recession index. In this case (fig. 1), the selection might be the period from day 100 to day 121, as prior data indicate the effects of direct-surface runoff or instability of the ground-water-head profile, or both. After the user has selected a segment, the program will calculate a mathematical expression in this form:
t = K1 x LogQ + K2 (1)
where t is time in days, LogQ is the logarithm of the flow in ft3/s, and K1 and K2 are coefficients that are determined by linear regression. This expression is used to derive the recession index (days/log cycle), which is the absolute value of K1. After this
solution is obtained the program proceeds to the next period of continuous recession, and repeats this process until a number of segments have been analyzed.
1000FLOW, IN CFS100101050100TIME, IN DAYS150200
Figure 1. Example period of continuous recession (day 93 to day 121) that RECESS might detect depending on program-user options. (The streamflow-gaging station is Holiday Creek near Andersonville, Virginia. The time scale represents days in the 1996 calendar-year.)
3
After the program has found all recession segments in the period of interest, it can perform calculations to determine the expression for the MRC. These calculations are illustrated in figure 2. Frame A represents the near-linear segments that were selected. The program uses equation 1 to derive the recession index (K) for each, represented by the dots in frame B. The program determines the best-fit linear equation for K as a function of LogQ, represented by the straight line in frame B. Coefficients of this
equation are used to obtain the MRC (frame C), which is a polynomial expression of time as a function of LogQ:
t = A x (LogQ)2 + B x (LogQ) + C (2)
where t is time and A, B, and C are coefficients. The MRC (frame C) is the result of integration of the best-fit linear equation for K as a function of LogQ (frame B). It is unique because at time=0, LogQ is the maximum value of the LogQ from all selected recession segments. In this example (frame C) the MRC is concave, but one regional application of RECESS demonstrated a variety of MRC shapes among basins (Rutledge and Mesko, 1996, p. 7-15). This empirical expression requires the assumption of a linear relation between K and LogQ. If this assumption is not valid for a particular dataset, the user should exercise caution or consider other methods for deriving the MRC from data displayed in frame B.
A B C
LOG OF FLOWTIME AFTER PEAKRECESSION INDEX (K)TIMEFigure 2. Schematic representation of the method used to determine the master recession curve: (A) selected recession segments, (B) recession index (K) (days per log cycle of recession), and best-fit line, and (C) the master recession curve, obtained from coefficients in B.
4
PREPARING TO USE THE PROGRAM
This section describes data files that the program user may need to obtain or prepare prior to executing the program. The first-time user experimenting with the data files included with this report may skip this section and proceed to the instructions for running the program (page 10). The section about the SCREEN program (page 8) may be helpful to the first-time user.
OBTAINING STREAMFLOW DATA FILES
A streamflow data file can be obtained from the following USGS web site:
http://waterdata.usgs.gov/nwis/dv/?referred_module=sw. From this site, the user is first asked to “Choose Site Selection Criteria”. For example, if the site of interest is Holiday Creek near Andersonville, Virginia, yet the site identification number is not known, the user might check the boxes for State and Site Name and then hit the “Submit” button. At the subsequent “Select sites that meet all of the following criteria” page, the user would then enter Holiday under site name, select the state of Virginia from the state list, check the “Streamflow, ft3/s” box in the “Water Level/Flow Parameters” group of the Available parameters list, and then issue the request using the “Submit” button near the very bottom of the page. The subsequent page will be a list of all sites found that meet the specified criteria. Clicking on the site number will display a page including graph(s) of the
previous 365 days of data. The final step to obtain a file of streamflow data is to check the “00060 Discharge (Mean)” item under the Available parameters heading, choose “Tab-separated” from the Output format list, enter the Begin and End dates of interest, and then hit GO. Once the data are displayed in your browser select “Save as” from the file pull-down menu of your browser and designate the directory where the data are to be written on the user’s computer, and specify a file name. On some systems this might require copying and pasting into Notepad. File names should be in the format xxxxx.txt, where xxxxx is an abbreviated station name. In this case the file name might be
designated as Holiday.txt. (Note: the length of the file name, including .txt, should be 12 characters or less.) Designate file type should be “Text File (.txt)”. An example streamflow data file, obtained as described above, is illustrated in figure 1.
A few comments about data files obtained from a USGS web site. Beginning in 2006, some files may include text qualifiers about streamflow data. These qualifiers, which may be evident in recent data (2006+), can appear immediately after a streamflow quantity and in some cases the qualifier may be located in the place of the streamflow quantity. The program may not execute properly due to these text qualifiers. For this reason the program user may need to edit streamflow data files, scanning for these qualifiers, and removing them if they exist.
In some applications, the program user may choose to generate a data file from a source other than a USGS web site. In this case, a few items about file format may be helpful: (1) Most of the header information (fig. 3) is not read by the program, and may not be necessary in a user-generated file. (2) The program starts reading data, two lines after it detects the string “agency” in the first five spaces of a line. (3) The columns of
5
information in each data line are tab-separated in a data file from a USGS web site (these could be substituted with single blank spaces in a user-generated data file).
# ---------------------------------- WARNING ---------------------------------------- # The data you have obtained from this automated U.S. Geological Survey database # have not received Director's approval and as such are provisional and subject to # revision. The data are released on the condition that neither the USGS nor the
# United States Government may be held liable for any damages resulting from its use. # Additional info: http://waterdata.usgs.gov/nwis/help/?provisional #
# File-format description: http://waterdata.usgs.gov/nwis/?tab_delimited_format_info # Automated-retrieval info: http://waterdata.usgs.gov/nwis/?automated_retrieval_info #
# Contact: gs-w_support_nwisweb@usgs.gov # retrieved: 2007-01-08 12:44:58 EST #
# Data for the following site(s) are contained in this file # USGS 02038850 HOLIDAY CREEK NEAR ANDERSONVILLE, VA
# ----------------------------------------------------------------------------------- #
# Data provided for site 02038850
# DD parameter statistic Description
# 02 00060 00003 Discharge, cubic feet per second (Mean) #
# Data-value qualification codes included in this output:
# A Approved for publication -- Processing and review completed. #
agency_cd site_no datetime 02_00060_00003 02_00060_00003_cd 5s 15s 16s 14s 14s USGS 02038850 1966-04-27 5.6 A USGS 02038850 1966-04-28 7.0 A USGS 02038850 1966-04-29 4.8 A USGS 02038850 1966-04-30 4.2 A USGS 02038850 1966-05-01 4.2 A USGS 02038850 1966-05-02 23 A USGS 02038850 1966-05-03 11 A USGS 02038850 1966-05-04 6.1 A USGS 02038850 1966-05-05 4.9 A USGS 02038850 1966-05-06 4.3 A USGS 02038850 1966-05-07 3.7 A USGS 02038850 1966-05-08 3.5 A USGS 02038850 1966-05-09 3.2 A USGS 02038850 1966-05-10 2.9 A USGS 02038850 1966-05-11 2.9 A USGS 02038850 1966-05-12 2.9 A USGS 02038850 1966-05-13 2.9 A USGS 02038850 1966-05-14 3.8 A USGS 02038850 1966-05-15 3.2 A USGS 02038850 1966-05-16 2.8 A USGS 02038850 1966-05-17 2.6 A USGS 02038850 1966-05-18 2.5 A USGS 02038850 1966-05-19 5.8 A USGS 02038850 1966-05-20 3.6 A USGS 02038850 1966-05-21 2.8 A USGS 02038850 1966-05-22 2.5 A USGS 02038850 1966-05-23 2.3 A USGS 02038850 1966-05-24 2.3 A USGS 02038850 1966-05-25 4.5 A USGS 02038850 1966-05-26 3.6 A USGS 02038850 1966-05-27 2.9 A USGS 02038850 1966-05-28 4.2 A USGS 02038850 1966-05-29 2.9 A USGS 02038850 1966-05-30 2.4 A USGS 02038850 1966-05-31 3.6 A USGS 02038850 1966-06-01 2.8 A USGS 02038850 1966-06-02 2.4 A USGS 02038850 1966-06-03 2.2 A USGS 02038850 1966-06-04 2.0 A USGS 02038850 1966-06-05 1.8 A USGS 02038850 1966-06-06 1.7 A USGS 02038850 1966-06-07 1.7 A USGS 02038850 1966-06-08 1.7 A USGS 02038850 1966-06-09 1.5 A USGS 02038850 1966-06-10 1.5 A USGS 02038850 1966-06-11 2.1 A USGS 02038850 1966-06-12 1.5 A USGS 02038850 1966-06-13 1.4 A USGS 02038850 1966-06-14 1.4 A USGS 02038850 1966-06-15 1.3 A USGS 02038850 1966-06-16 1.2 A
Figure 3. Example streamflow data file obtained from a USGS web site. (This is truncated from the data file Holiday.txt, which is available with this manual.)
6
As noted on the first page, efforts have taken place to maintain the functionality of the program using data in a particular format. Nonetheless, there may be some cases in
which the program user will need to modify format statements in the code and re-compile the program. The fortran source code is available on the program web site.
THE SCREEN PROGRAM
Before analyzing a streamflow data file the user may execute the SCREEN program to display the period of record. To demonstrate, execute the program by double-clicking the application in windows or by entering the command “screen” in the MS-DOS command window. Designate the streamflow data file Indian.txt. The output file screenout.txt should look like figure 4. This illustrates data are available from 1970 to 2000.
THIS IS OUTPUT OF THE SCREEN PROGRAM, VERSION OF JANUARY 2007
READING FILE NAMED Indian.txt FIRST YEAR IN RECORD = 1958 LAST YEAR IN RECORD = 1986 MONTH
YEAR J F M A M J J A S O N D 1958 X X X X X X X X X . . . 1959 . . . . . . . . . . . . 1960 . . . . . . . . . . . . 1961 . . . . . . . . . . . . 1962 . . . . . . . . . . . . 1963 . . . . . . . . . . . . 1964 . . . . . . . . . . . . 1965 . . . . . . . . . . . . 1966 . . . . . . . . . . . . 1967 . . . . . . . . . . . . 1968 . . . . . . . . . X X X 1969 X X X X X X X X X X X X 1970 X X X X X X X X X . . . 1971 . . . . . . . . . . . . 1972 . . . . . . . . . . . . 1973 . . . . . . . . . . . . 1974 . . . . . . . . . . . . 1975 . . . . . . . . . . . . 1976 . . . . . . . . . . . . 1977 . . . . . . . . . . . . 1978 . . . . . . . . . . . . 1979 . . . . . . . . . . . . 1980 . . . . . . . . . . . . 1981 . . . . . . . . . . . . 1982 . . . . . . . . . . . . 1983 . . . . . . . . . . . . 1984 . . . . . . . . . . . . 1985 . . . . . . . . . . . . 1986 . . . . . . . . . X X X
COMPLETE RECORD = . INCOMPLETE = X
Figure 4. Example tabular display of a streamflow record, showing periods of
complete record and periods of incomplete record. (This display is generated onto a file by the program SCREEN, and on the monitor by RECESS. Streamflow data file is Indian.txt.)
7
THE STATION.TXT FILE
An input data file “station.txt” includes drainage area of the stations of interest (fig. 5). This file includes an entry for the streamflow data files that are provided with this report (Indian.txt and Holiday.txt). The user should place additional entries in this file for each station analyzed. The RECESS program does not use the drainage area, but an auxiliary program CURV uses this variable in calculations of the MRC.
A few items about the format of station.txt. The file should have 10 header lines which are not read by the program. Beginning with line 11, each line should include streamflow file name (such as Indian.txt) and the drainage area of that station. The CURV program is formatted to read these two fields as A12, 1F8.0. All drainage areas below are given with two decimal spaces after the decimal point, in which case the programmed format is ignored, and the two decimal places are read. When new lines are added to station.txt, no tabbing should be used. The program is sensitive to the case of the letters in the name of the streamflow file. For example one of the example data files included is “Indian.txt”. The case of letters in file station.txt (figure 5) is the same, including the capital I. The program user should be aware of this when executing CURV.
File \"station.txt\"
This file is read by programs PREP, RECESS, RORA, and PART, to obtain the drainage area. Note: This file should have ten header lines. The streamflow file name should be 12 characters or less. ---------------------------------------------------------------- Drainage
Name of area The space below, after drainage area, is streamflow (Square for optional information that is not read file miles) by the programs. This is free-form.
---------------------- ------------------------------------------ Indian.txt 8.88 02371200 Indian Creek near Troy Alabama Holiday.txt 8.53 02038850 Holiday Cr near Andersonville, VA Sewee.txt 117.00 Sewee Creek near Decatur, TN
LitAndro.txt 73.50 01057000 Lt Androscoggin R nr S.Paris, ME BigHill.txt 37.00 Big Hill Cr near Cherryvale KS
Figure 5. Example file “station.txt”, including the drainage area for each of the data files that are included with the program.
8
RUNNING THE RECESS PROGRAM
This section gives specific instructions for running RECESS for a streamflow data file that is included with this report. The first sub-section describes the use of RECESS in conjunction with the RECPLOT application to generate a graph of all recession segments that are detected by the program. This exercise is performed without mathematical analysis. The first-time program user should try these steps as a way of gaining
familiarity with the program. The second sub-section describes the determination of the median recession index. The operations in the second sub-section require user
designation of segments to be analyzed based upon near-linear recession and demonstrate calculations performed by the program. The third sub-section describes the derivation of the master recession curve (MRC). These three sub-sections describe specific use of the program for a specific dataset. The fourth sub-section includes considerations for program use in hydrologic studies that involve other datasets.
GRAPHICAL DISPLAY OF RECESSION SEGMENTS WITHOUT MATHEMATICAL ANALYSIS
The program can be executed by double-clicking the RECESS application in Windows, or by opening the MS-DOS window and entering the command “recess”. In either case, a command window is displayed. Expand this window downward so that more than 40 lines are displayed. Enter Holiday.txt as the streamflow data file. The program generates a tabular display of the completeness of the record. Designate the starting year as 1970 and the ending year as 2000, to analyze the entire period available. At this point the program will ask the number of months of interest; enter 12, then enter “n” to designate that no particular season is being analyzed. The program asks how many days of recession are required in order to detect a recession period; enter 15.
After the operations described above, the program will detect the first period of recession. At this point the program should indicate the “date of new peak” to be 1970-11-20 (year, month, day). The “peak” is the day that precedes the period of recession. The program displays, within the command window, a number of possible operations, most of which will not be needed during this exercise. Before responding to these, move the command window to the left side of the monitor. Click on the RECPLOT application in Windows. This should generate a new window entitled RECPLOT, which includes a graph of the period of recession that has been detected. The X-axis is days since the peak and the Y-axis shows the logarithm of flow (Log-10 of ft3/s). The plot should be similar to the bottom curve in figure 6A. Leaving the new window open, return to the command window and respond “a” to advance to the next period of recession. Then, in the
9
RECPLOT window, click on “open” then the file name “recdata.txt”, and another recession segment should be generated. When these operations are repeated two more times there should be four recession segments in the RECPLOT window, similar to figure 6A. The program user might repeat these operations until RECESS has reached the end of the period of record. When all segments have been illustrated, the RECPLOT window should appear similar to figure 6B. At this point the program displays the following
question at the bottom of the command window: “Do you want to analyze the recessions selected?” Enter “n” then click “enter” one more time to terminate the RECESS program.
No mathematical analysis has been performed at this point. Most of these segments include departures from the linear model of recession at early time. These departures may be caused by direct surface runoff or instability of the ground-water-head profile. The magnitude of these departures depends on the magnitude of the storm event (and the amount of recharge) that occurred during or near the time of the peak. The mathematical analysis summarized earlier (pages 2-4) should take place only after these early-time sections have been removed. The following section explains this process. Click “Exit” in the RECPLOT window when ready to go to the next exercise.
A B
1.51.51LOG OF FLOWLOG OF FLOW010TIME, IN DAYS2010.50.500-0.5-0.5010TIME, IN DAYS20
Figure 6. Example graphical display of recession segments without mathematical analysis, similar to a graph that can be generated with the RECPLOT application. (Frame A shows the plot after 4 segments have been illustrated. Frame B shows the plot after all segments have been illustrated.)
10
DETERMINATION OF THE RECESSION INDEX
The use of the program for this purpose begins with steps that are similar to the previous section. Execute the program by double-clicking the RECESS application in Windows Explorer, or by opening the MS-DOS window and entering the command “recess”. In either case, a command window is displayed. Expand this window downward so that more than 40 lines are displayed. Enter Holiday.txt as the streamflow data file. The program generates a tabular display of the completeness of the record. Designate the starting year as 1970 and the ending year as 2000, to analyze the entire period available. At this point the program will ask the number of months of interest; enter 12, then enter “n” to designate that no particular season is being analyzed. The program asks how many days of recession are required in order to detect a recession period; enter 15.
At this point the program will require user selection of the recession periods, and
designation of the segment for analysis. The first time the program is executed, the user should specify program input as noted below. This will be helpful for explanation of the program on this page and the next four pages. (Later, the user should read the section “Additional Considerations for Program Use.”) For the sake of this exercise, all
recession periods will be selected. Also, it is noted that most departures from linearity occur within the first 6 days of each period of recession (fig. 6) and that all recession periods are at least 15 days long. To simplify this exercise, the same options are selected whenever RECESS detects a new recession period:
g(return)….. c(return)…..7(return)……15(return)…….r(return)
The option “g” will generate a graphic of the period of recession, and is optional for this exercise. The option “c” will designate the user is to choose limits of the segment, and the next two entries, “7” and “15” are the limits. The option “r” will instruct the program to perform linear regression to solve equation 1, and thus determine the recession index. The five entries are given 20 times (the number of recession periods that are detected by RECESS).
After operations on the last recession segment, the program asks whether you want to analyze the recessions selected. Enter “y.” Then the program displays the recession indexes (K) for each segment selected. The value of K is shown numerically and by the position of the asterisk. At this point the program user could eliminate outliers, but enter “0” to indicate no recession segments are to be eliminated. Then enter “2” to indicate you want to describe the MRC (this will be discussed in the next section). Then, when prompted, “enter” one more time to terminate the program.
If the RECESS program has been executed as described above, there should be a new line at the bottom of file “index.txt”, giving the recession index = 67.43 days per log cycle. This represents the median value among all segments analyzed. In hydrologic studies the recession index might be reported with viewer significant figures (for example 67 or 70).
11
THE MASTER RECESSION CURVE
To describe the MRC, the user should execute the RECESS program as described in the previous page, then proceed with the use of data files and the CURV program, as
described below. It is noteworthy that near the end of the previous sub-section, the user elected to derive the MRC, with the number “2”. If the number “1” was designated, then the program would not have generated output as described below. If the “1” were entered the program would only write to file “index.txt”.
Output Files for the RECESS Program
When the RECESS program is executed as described above, results are written to three output files, as described below.
recsum.txt -- Each time RECESS is executed, one new line is written to the bottom of recsum.txt. This line gives a summary of results of the session: name of the streamflow datafile; single-letter designation of the season; time period of analysis; number of
recession segments used; the minimum, median, and maximum recession index, the range of LogQ among all segments used; and the three coefficients of the MRC (eq. 2). The auxiliary program CURV will read recsum.txt for the purpose of generating an x-y file which can be used to generate the MRC.
X-file – A new x-file gives details about one execution of RECESS. The x-file is
assigned a name equal to the name of the original streamflow datafile, except two letters are added to the beginning of the name. The first is the letter “x” and the second is the letter that designates the season. If RECESS has been executed as described above, the file name is xn.Holiday.txt. The x-file includes detailed information about each recession segment that was initially selected—date of the peak and designation of the starting and ending of the recession segment (days after the peak). Also included is a listing of each data pair of LogQ and K for each recession segment and a graphical representation of K for each segment. This listing is shown twice: once before and once after the user
eliminates segments not wanted. The x-file also shows the best-fit linear equation for K as a function of LogQ and the resulting second-order polynomial that is the MRC.
Because of the file name, if RECESS is executed twice for the same streamflow data file and season, the first x-file will be overwritten.
Y-file – A new y-file gives detailed information about the time and flow for each day of each selected recession segment. The y-file is assigned a name equal to the name of the streamflow datafile, except two letters have been added to the beginning of the name. The first is the letter “y” and the second is the letter that designates the season. If RECESS has been executed as described above the file name is yn.Holiday.txt. The y-file can be used to generate a graph of each recession segment. One such graph could be the flow as a function of the time since the last peak (fig. 7). Another could be the flow as a function of time, where the time has been adjusted so that all recession segments are
12
drawn on the same MRC (fig. 8). Because of the file name, if RECESS is executed twice for the same streamflow data file and season, the first y-file will be overwritten.
1.21LOG OF FLOW0.80.60.40.20010TIME, IN DAYS20
Figure 7. Recession segments generated from the Y-file, as described in the text. (The x-axis is the time since the last peak.)
1.21LOG OF FLOW0.80.60.40.20-100102030TIME, IN DAYS40506070
Figure 8. Recession segments generated from the Y-file, as described in the text. (The x-axis is the time adjusted so that all segments will appear roughly on the same master recession curve.)
13
The CURV Program
CURV is an auxiliary program which can be used to display multiple MRC’s on the same graph. These can represent a number of streamflow gaging stations (basins), such as the example in figure 9, or various seasons for the same streamflow gaging station. This program reads file station.txt (see page 8) to obtain the drainage area. The program can display MRC’s using units of specific discharge (cubic feet per second per square mile). The CURV program is used to illustrate the exact solution of the MRC equation, without showing each segment (unlike the Y-file).
The CURV program scans recsum.txt and finds the first entry that satisfies the user-designated combination of streamflow data file and season. For each combination, the program determines 50 evenly-spaced values of LogQ—from the maximum to the
minimum values of LogQ for all recession segments that were selected in the execution of RECESS (these limits are read from recsum.txt). For each LogQ, the time is
calculated according to equation 2, using coefficients in recsum.txt. The program writes 50 lines to curvout.txt.
To generate the MRC for the streamflow dataset analyzed in previous pages, execute the CURV program by double-clicking the application in windows or by entering the
command “curv” in the MS-DOS command window. Designate the streamflow data file as “Holiday.txt” (note case sensitivity), then specify with the letter “n” there is no
particular season of interest, to be consistent with the application of RECESS described earlier. Then designate no additional MRC’s are to be drawn (with the letter “n”). Exit the CURV program, then generate the MRC using a graphics application such as MS Excel, reading file curvout.txt. It may be necessary to remove some of the header lines at the beginning of the file before plotting. The plot should be similar to figure 10.
If this method is used to generate multiple MRCs, such as figure 9, the recess program should be executed for each dataset and/or season. Each application of RECESS should be represented by one line in recsum.txt. For example, if the user is generating MRC’s for summer and winter for the same streamflow datafile, the user should execute RECESS twice. In one application of RECESS, summer months are selected for
analysis, and the season should be designated as “s”, and in the other application winter months are selected and “w” is designated. When CURV is executed, designate the station name and season “s”, then when asked if additional MRC’s are to be drawn, answer “y” then the station name again, and season “w”.
14
3LOG OF FLOW210020406080100120140TIME, IN DAYS
Figure 9. Example plot of multiple master recession curves, which can be generated from output of the CURV program. (from Rutledge and Mesko, figure 6).
1.21LOG OF FLOW0.80.60.40.20-100102030TIME, IN DAYS40506070
Figure 10. The master recession curve, generated from the file curvout.txt, as described in the text.
15
ADDITIONAL CONSIDERATIONS FOR PROGRAM USE
The exercises described above are provided so the user can gain familiarity with the program. These include the use of one streamflow dataset and specific user options, so results described here can be compared with results generated by the user. When conducting a hydrologic study the program user may consider variations as described below.
The user might select certain months for analysis. For example periods of ground-water evapotranspiration (GWET) might be eliminated because this process will affect recession characteristics. If GWET is considered to be significant in the four months April, May, June, and July, elect to analyze 8 months, then designate them using these numbers: 1, 2, 3, 8, 9, 10, 11, and 12 (each followed by a “return”).
The user will find it necessary to modify the limits for plotting, by editing the file
ranges.txt. The version of this file that is provided with this manual includes ranges of LogQ that are conducive to the analysis of the streamflow datafile Holiday.txt.
The RECPLOT application might be used as a guide in selecting near-linear segments. When a new recession interval has been found by the program, the user clicks on the application outside the command window. Before proceeding to the next recession interval, the user should close the RECPLOT window.
When recession data are displayed (using “g” in the command window, or using the RECPLOT application), the data may exhibit “stepping” which is caused by rounding of daily streamflow in the USGS database. The “t” option can also be used to view recession data.
The exercise described earlier includes selection of all recession intervals, and the same time limits for each: 7 to 15 days (p. 11). When the RECESS program is applied in a hydrologic study the user can apply considerable judgment. For example the author suggests this input for data file Holiday.txt (tabulation below). This includes selection of longer periods of recession, where such data are available, and the selection “a” to advance without analysis of a segment, given some recession intervals may show anomalous recession characteristics.
16
1970-11-20 g c 4 18 r 1971-04-07 g c 7 16 r 1973-05-28 g c 6 18 r 1974-07-07 g c 6 16 r 1976-04-01 g c 6 23 r 1976-08-09 g c 3 18 r 1977-05-04 g c 7 19 r 1977-06-25 g c 3 15 r 1978-03-26 g c 8 15 r 1978-08-07 g c 10 17 r 1981-06-07 g a
1982-04-26 g c 9 22 r 1984-05-06 g c 7 16 r 1984-05-30 g c 7 17 r 1985-08-27 g c 15 25 r 1986-06-12 g r 1993-01-22 g a
1994-09-01 g c 6 15 r 1997-08-20 g a 2000-04-25 g a
….. after this, program options described on p. 11 (paragraph 4) can be repeated. The program displays recession index (K) numerically and by the position of an asterisk. This is given in order of the value of LogQ, so the relation between K and LogQ might be viewed at this point. The user has the option of deleting segments if a value of K is
considered to be an outlier. To eliminate one outlier, enter “1” then enter the number that identifies the recession segment to be deleted.
The RECESS program is designed for determination of the recession index and the master recession curve. Depending on the needs of hydrologic studies, the program might be used for other purposes. For example non-quantitative description of recession curves, without editing, may be useful in some studies. (Fig. 6).
A graphic is generated to the monitor near the end of program execution, and is repeated in the “X-file,” showing the relation between K and LogQ. This could be used to
evaluate the assumed linear relation. (See General Description of the Method, and Figure 2.) Data pairs could be extracted from the X-file, then plotting using separate graphics software. If the assumed linear relation is not valid, the user might decide not to use equation 2 to describe the MRC.
17
REFERENCES
Anderson, M.G., and Burt, T.P., 1980, Interpretation of recession flow: Journal of Hydrology, v. 46, p. 89-101.
Bevans, H.E., 1986, Estimating stream-aquifer interactions in coal areas of eastern Kansas by using streamflow records, in Subitzky, Seymour, ed., Selected papers in the Hydrologic Sciences: U.S. Geological Survey Water-Supply Paper 2290, p. 51-64.
Daniel, J.F., 1976, Estimating groundwater evapotranspiration from streamflow records: Water Resources Research, v. 12, no. 3, p. 360-364.
Gerhart, J.M. and Lazorchick, G.J., 1988, Evaluation of the ground-water resources of the lower Susquehanna River basin, Pennsylvania and Maryland: U.S. Geological Survey Water-Supply Paper 2284, 128 p.
Hoos, A.B., 1990, Recharge rates and aquifer hydraulic characteristics for selected
drainage basins in middle and east Tennessee: U.S. Geological Survey Water-Resources Investigations Report 90-4015, 34 p.
Hornberger, G.M., Ebert, Janet, and Remson, Irwin, 1970, Numerical solution of the Boussinesq Equation for aquifer-stream interaction: Water Resources Research, v. 6, no. 2, p. 601-608.
Horton, R.E., 1933, The role of infiltration in the hydrologic cycle: Transactions of American Geophysical Union 14, p. 446-460.
Ibrahim, H.A. and Brutsaert, Wilfried, 1965, Inflow hydrographs from large unconfined aquifers: American Society of Civil Engineers, Journal of Irrigation and Drainage Division, v. 91, no. IR 2, p. 21-38.
Ineson, J. and Downing, R.A., 1964, The ground-water component of river discharge and its relationship to hydrogeology: Journal of the Institution of Water Engineers, v. 18, no. 7, p. 519-541.
Nutbrown, D.A., and Downing, R.A., 1976, Normal-mode analysis of the structure of baseflow-recession curves: Journal of Hydrology, v. 30, p. 327-340.
Petras, Ivan, 1986, An approach to the mathematical expression of recession curves: Pretoria, South Africa, Hydrologic Research Institute, Water S.A., v. 12, no. 3, p. 145-150.
Riggs, H.C., 1964, The base-flow recession curve as an indicator of ground water: International Association of Scientific Hydrology, Publiction 63, p. 352-363.
18
Rorabaugh, M.I., 1964, Estimating changes in bank storage and ground-water contribution to streamflow: International Association of Scientific Hydrology, Publication 63, p. 432-441.
Rutledge, A.T., 1998, Computer programs for describing the recession of ground-water discharge and for estimating mean ground-water recharge and discharge from streamflow data – update: U.S. Geological Survey Water-Resources Investigations Report 98-4148, 43 p.
Rutledge, A.T., 2002, User guide for the PULSE program: U.S. Geological Survey Open-File Report 02-455, p. 34.
Rutledge, A.T., 2003, Use of RORA for complex ground-water flow conditions: U.S. Geological Survey Water-Resources Investigation Report 03-4304, 5 p.
Rutledge, A.T. and Mesko, T.O., 1996, Estimated hydrologic characteristics of shallow aquifer systems in the Valley and Ridge, the Blue Ridge, and the Piedmont physiographic provinces based on analysis of streamflow recession and base flow: U.S. Geological Survey Professional Paper 1422-B, 58 p.
Singh, K.P., 1969, Theoretical baseflow curves: American Society of Civil Engineers, Journal of Hydraulics, v. 95, no. HY 6, p. 2,029-2,048.
Singh, K.P. and Stall, J.B., 1971, Derivation of base flow recession curves and parameters: Water Resources Research, v. 7, no. 2, p. 292-303.
Szilagyi, Jozsef, 1999, On the use of semi-logarithmic plots for baseflow separation: Ground Water, v. 37, no. 5, p. 660-662.
Toebes, C., and Strang, D.D., 1964, On recession curves: 1, Recession equations: Journal of Hydrology (New Zealand), v. 3, no. 2, p. 2-14.
Trainer, F.W., and Watkins, F.W., 1975, Geohydrologic reconnaissance of the upper Potomac River basin: U.S. Geological Survey Water Supply Paper 2035, 68 p.
Werner, P.W., and Sundquist, K.J., 1951, On the groundwater recession curve for large water-sheds: International Association of Scientific Hydrology Publication 33, v. 2, p. 202-212.
Wilder, H.B, and Simmons, C.E., 1978, Program for evaluating stream quality in North Carolina: U.S. Geological Survey Circular 764, 16 p.
Wood, C.R., 1980, Groundwater resources of the Gettysburg and Hammer Creek formations, southeastern Pennsylvania Geological Survey, 4th ser., Water Resources Report 49, 87 p.
19
因篇幅问题不能全部显示，请点此查看更多更全内容