User`s Manual for the University of British Columbia Regime Model
Contents
1. STUDY AREA STUDY AREA DATE DATE CROSS SECTION LOCATION CROSS SECTION LOCATION STADIA READINGS ELEV DIST STADIA READINGS ELEV DIST STATION UPPER MIDDLE LOWER m STATION UPPER MIDDLE LOWER m m BENCH MARK BENCH MARK RBT RBT LBT LBT BENCH MARK BENCH MARK L RBT left right bank top L RBB left right bank bottom L REV left right edge of L RBT left right bank top L RBB left right bank bottom L REV left right edge of vegetation L RBF left right bank full elevation BR bar surface L RWE left right vegetation L RBF left right bank full elevation BR bar surface L RWE left right water s edge WC wetted channel T thalweg water s edge WC wetted channel T thalweg2. 28 5 1 Response to Changes in Formative FlOW cccsccesseeceeseeceeeeeceeeeeceeeeeeeneeeeaees 28 5 2 Response to Riparian Disturbance 30 6 33 UBCRM Manual 11 6 5 08 1 INTRODUCTION The regime model can be run from two different platforms The most widely accessible platform is Excel where the program relies on the Solver Add in component to seek numerical solutions The most flexible and useful platform is the numerical analysis program MATLAB http www mathworks com This user s manual presents the instructions for installing and running both versions of the program and provides some examples of how the model can be used practically 2 INSTALLATION INSTRUCTIONS 2 1 EXCEL VERSION To install the UBCRM program simply copy the Excel file UBCRM xls to your hard drive and then open the file with Excel The file contains four worksheets entitled notes CRB Eg UBCRM and UBCRM_H respectively The notes page is intended to be used to temporarily if not permanently store the results of various analyses run by the model and is no different than any other blank worksheet in Excel Additional sheets for storing simulation results can be created by clicking on the menu then on Worksheet from the list of options in that menu The CRB Eg sheet contains a set of equations that
3. 44 11 7 0 07 0 87 45 64 444 c HH HL 1 21 0 20 0 67 32 45 111 111 HY HH 1 21 0 20 0 47 227 32 HA HE HH 15 0 14 0 33 16 22 7 444 HH 1111 14 0 13 0 20 11 16 H4 111 8 0 07 0 13 8 11 11 2 0 02 0 11 5 6 8 11 2 0 02 0 09 lt 5 6 14 10 0 09 0 00 Total 107 P is the number in each class divided by the total CF is the total proportion of the distribution finer than the given class the Cumulative Frequency distribution For a field sampling template see UBCRM_Forms pdf Figure 9 Sample Wolman data from Fishtrap Creek UBCRM Manual 17 6 5 08 Based on the sample data in Figure 9 the Dos falls in the size class ranging from 128 mm to 181 mm the Dg falls between 64 mm and 90 5 mm and falls between 45 mm and 64 mm By plotting the cumulative frequency values against the lower bound for the size class we can generate grain size distribution curve Figure 10 Based on the curve the estimated Dos is 130 mm Ds4 is 85 mm is 47 mm Grain size mm 3 5 BANK STRENGTH INDEX Figure 10 Grain size distribution curve Unfortunately there is only limited work that we can draw on to accurately estimate the bank strength as a function of characteristics that can be measured in the field This is perhaps the primarily impediment to accurately modeling stream channel morphology because bank stren
4. apparent friction angles of 40 45 49 and 55 for types I through IV respectively However both zz and vary with the size of the channel under consideration even when the riparian vegetation characteristic remain the same for example see the data analysis conducted by Eaton and Church 2007 Eaton and Millar 2004 conducted an analysis which suggests that the effect of vegetation on bank strength disappears for channels with formative discharges greater than about 400 m s for gt 400 streams will characterized by w 1 0 and 40 unless of course cohesive sediment in the floodplain or other factors significantly affect relative bank strength 80 t Q A Relative bank strength d Relative bank strength NO 10 10 10 10 Formative Discharge m3 s Formative Discharge m gt s Figure 12 Variation in with formative discharge for alluvial channels in the Salmon River area data from Emmett 1975 and from streams in Colorado Andrews 1984 for channels with thin vegetation circles and thick vegetation UBCRM Manual 19 6 5 08 The UBCRM model employs Millar and Quick s 1993 original bank strength formulation while the CRB eg tab in the Excel version of the program uses the simplified relative bank strength Eaton 2006 developed and tested an alternate bank strength
5. 112 8 671 689 Kirkby M J 1977 Maximum sediment efficiency as a criterion for alluvial channels In K J Gregory Editor River Channel Changes John Wiley amp Sons Ltd Chichester pp 429 442 Marcus W A Roberts K Harvey L and Tackman G 1992 An evaluation of methods for estimating Manning s n in small mountain streams Mountain Research and Development 12 3 227 239 Millar R G 2005 Theoretical regime equations for mobile gravel bed rivers with stable banks Geomorphology 64 207 220 Millar R G and Quick M C 1993 Effect of bank stability on geometry of gravel rivers Journal of Hydraulic Engineering Asce 119 12 1343 1363 Millar R G and Quick M C 1998 Stable width and depth of gravel bed rivers with cohesive banks Journal of Hydraulic Engineering Asce 124 10 1005 1013 Moore R D 1996 Snowpack and runoff response to climatic variability Southern Coast Mountains British Columbia Northwest Science 70 4 321 333 Parker G 1976 On the cause and characteristic scales of meandering and braiding in rivers Journal of Fluid Mechanics 76 3 457 480 Parker G 1990 Surface based bedload transport relation for gravel rivers Journal of Hydraulic Research 28 4 417 436 Phillips J C 2007 Post fire dynamics of a gravel bed stream Fishtrap Creek British Columbia MSc Thesis The University of British Columbia Vancouver 120 pp White W R Bettess R and Paris E 1982 Analytic
6. a reach average value for H from them However this approach applies only to gravel bed streams having a coarse gravelly lower stratum in UBCRM Manual 20 6 5 08 the channel bank associated with deposition of bed material as generally speaking channel bars and an upper stratum of finer overbank deposits re enforced by riparian vegetation root systems Where this very general sedimentological model does not apply H should not be used to parameterize bank strength 4 MODEL CALIBRATION Once the input parameters have been determined the model should be run and the results compared with the known channel dimensions If there are significant deviations between the model predictions and the observed channel geometry then the input parameters should be re evaluated and adjusted where appropriate based on consideration of the field observations Even if the model predictions agree well with the observed conditions it is worthwhile varying the inputs to determine how sensitive the model predictions are to the inputs After the model has been successfully calibrated it can be used to evaluate channel response to changes in the governing conditions due for example to land use changes natural disturbance in the watershed or direct human modification of the stream channel 4 1 RUNNING THE EXCEL MODEL The simplified regime equations presented by Millar 2005 are presented on the CRB_Eq tab of the QuickTime and a ER LAN Gra Exc
7. formulation which is used in the UBCRM_H model Figure 13 Bank strength analysis based on characteristic rooting depth H The alternate bank stability analysis invokes a characteristic riparian rooting depth H which produces a vertical upper bank section above a cohesionless gravel toe Figure 13 Each of Hey and Thorne s 1986 are associated with a typical rooting depth value type I 0 36 m type II H 0 53 m type H 0 89 m type IV 1 07 m This is effectively equivalent to invoking a root cohesion term which Eaton 2006 demonstrates are similar to the root cohesion values determined in studies of debris slides Buchanan and Savigny 1990 The equivalent root cohesion values are 1 5 kPa 2 2 kPa 3 7 kPa and 4 5 kPa for types I to IV respectively The primary advantage of this alternate bank stability analysis is that the effect of a given forest cover type does not depend on channel scale The same value of H applies to small headwater streams where roots dominate the bank stability and large mainstem channels where the effect of riparian vegetation is quite limited which is not the case for the approach Figure 12 There is also the possibility of taking measurements in the field of the average vertical bank height to estimate H in order to better constrain the bank stability analysis If the cross sectional surveys taken in the field are sufficiently detailed it should be possible to estimate
8. generate first approximations of the channel dimensions based on a reduced set of input parameters The other two sheets contain a series of cells into which the user must enter a complete set of input parameters characterizing the river system under consideration and other cells whose values are determined by the values the user enters The UBCRM model contains the regime model based on Millar and Quick s 1993 bank stability analysis The UBCRM_H model contains a regime model based on Eaton s 2006 bank stability analysis Both models function in essentially the same way To run the complete UBCRM and UBCRM_H models see Section 4 for details the user enters the required data values and then selects the Solver tool from the Tools menu The The freely available program Octave http www gnu org software octave octave html may also be capable of running the MATLAB scripts However this manual does not explicitly cover installation and running the program in Octave UBCRM Manual 3 6 5 08 user must be on the worksheet for the model that they wish to run in order to access the appropriately configured Solver Solver is accessible from any worksheet but the settings which have already been specified in the Excel program are customized for each worksheet Figure 1 shows the Solver window and Solver parameters for the UBCRM model Figure 2 shows the window and parameters for the UBCRM model QuickTime and a TIFF LZW
9. picture Figure 24 Results and analysis summary for the UBCRM_H model 5 EVALUATING CHANNEL RESPONSE SCENARIOS Once a calibrated regime model has been set up for a given field site various channel response scenarios can be evaluated The purpose of the investigation determines the form that the investigation will take Below several examples are presented that demonstrate some ways in which the model may be used 5 1 RESPONSE TO CHANGES IN FORMATIVE FLOW The long term average discharge carried by a stream can vary in concert with climate indicators such as the PDO index Moore 1996 or in response to long term climate change The analysis is relatively straightforward keeping all other input parameters the same vary the stream discharge and document the change in channel dimensions and sediment transport capacity UBCRM Manual 28 6 5 08 Sensitivity Analysis O W Wo 1 8 R Ro Qb Qbo Q Qo Figure 25 Sensitivity analysis using the UBCRM_H model Q 90 m s 40 mm Dos 120 mm S 0 0028 n 0 035 H 0 59 m Sensitivity plots where the proportional change in width hydraulic radius and transport rate are plotted against proportional change in discharge are a very useful means of interpreting the model predictions Figure 25 The analysis shows that the channel width is relatively sensitive to changes in the formative discharge while th
10. the bank strength following riparian disturbance This example is based on the Fishtrap Creek study area as described by Phillips 2007 The floodplain of Fishtrap Creek was burned in 2003 and began systematically widening in 2006 in response to flows close to the long term mean annual peak flow presumably as a result of the loss of root strength Based on Phillips analysis Q 7 35 m s S 0 019 m m n 0 06 45 mm and Dos 181 mm The initial width of the channel prior to disturbance is estimated to be about 9 m with a mean hydraulic radius of about 0 60 m Following a peak flow of about 7 5 m s the channel widened by as much as 2 at some cross sections Surveys in 2007 show even more dramatic channel widening reaching about 15 m in some places again in response to flows close to the mean annual peak flow The following analysis relates the channel dimensions transport capacity and morphology to bank strength The channel dimensions predicted by the UBCRM_H model assuming H 0 55 m or root cohesion C 2 33 kPa are consistent with the pre disturbance conditions The calibrated UBCRM_H model predicts W 8 8 m R 0 51 m and v 1 5 m s the mean velocity measured in the field at bankfull conditions in 2006 was 1 59 m s The estimated sediment transport capacity is 0 075 m s Using the UBCRM model we get a good agreement assuming 63 u 3 40 So while the pre disturbance H value is consistent
11. 08 A second analysis was conducted using the UBCRM model In this case the apparent friction angle was varied from the initial value of 63 to 35 representing a range of relative bank strengths from 3 4 to 1 2 representing a range of bank strengths similar to that investigated with the UBCRM_H model above The results are nearly identical as well with width more than doubling and the widest solutions potentially associated with a shift to a multiple thread channel pattern Sensitivity to changes in u Fishtrap Creek Oo W Wo R Ro 8 Ob Obo 2 5 1 5 1 e Se O SSE 0 5 0 1 0 9 0 8 0 7 0 6 0 5 u Figure 27 Sensitivity analysis using the UBCRM model Q 7 35 m s D5 40 mm Dos 181 mm S 0 019 n 0 06 u 3 4 The results of both analyses suggest that if a good vegetative cover can be established before the riparian root systems of the trees killed by the fire lose all of their strength a process that requires between 3 and 15 years approximately then degree of widening that occurs may be reduced and a change in channel pattern is unlikely to occur Since the transport capacity will drop as the channel widens there is the potential for local or systemic aggradation to occur depending on what happens to the sediment supply to the stream channel following the fire A dramatic increase in sediment supply is likely to UBCR
12. M Manual 32 6 5 08 overwhelm the system driving it out of regime at least temporarily and producing an unstable multi thread channel 6 REFERENCES Andrews E D 1984 Bed material entrainment and the hydraulic geometry of gravel bed rivers in Colorado Geological Society of America Bulletin 95 3 371 378 Baker V R 1977 Stream channel response to floods with examples from central Texas Geological Society of America Bulletin 88 1057 1071 Bathurst J C 2002 At a site variation and minimum flow resistance for mountain rivers Journal of Hydrology 269 11 26 Buchanan P and Savigny K W 1990 Factors Controlling Debris Avalanche Initiation Canadian Geotechnical Journal 27 5 659 675 Chang H H 1979 Minimum stream power and river channel patterns Journal of Hydrology 41 303 327 Chow V T 1959 Open channel Hydraulics Mcgraw Hill New York 680 pp Church M 1992 Channel morphology and typology In P Callow and G E Petts Editors The Rivers Handbook Blackwell Science pp 126 143 Cowan W I 1956 Estimating hydraulic roughness coefficients Agricultural Engineering 37 473 475 Davies T R H and Sutherland A J 1983 Extremal Hypotheses for River Behavior Water Resources Research 19 1 141 148 Eaton B Church M and Ham D 2002 Scaling and regionalization of flood flows in British Columbia Canada Hydrological Processes 16 16 3245 3263 Eaton B C 2006 Bank stability ana
13. THE UNIVERSITY OF BRITISH COLUMBIA REGIME MODEL UBCRM USER S MANUAL DEPARTMENTS OF CIVIL ENGINEERING amp GEOGRAPHY THE UNIVERSITY OF BRITISH COLUMBIA October 2007 Brett Eaton 2007 OVERVIEW OF THE UBC REGIME MODEL The UBC Regime model has been developed over a number of years in collaboration between researchers in the Department of Civil Engineering and the Department of Geography at the University of BC The model is based on the understanding that a simple model with modest data requirements is more likely to be useful than a data intensive numerically demanding one especially for environmental practitioners While simple the model does consider the relevant controlling factors the most important of which is the nature and erodibility of the channel banks The goal of our research on this topic is to determine which simplifying assumptions about river channel behaviour are reasonable to make and to identify the underlying physical processes The version of the model presented here can be calibrated for a field site and then used to evaluate how the field site could potentially respond to changes in the formative discharge bank strength or any of the other governing conditions related to channel morphology in the model Rational regime theories relating stream channel conditions to the external driving forces have a long history Chang 1979 Davies and Sutherland 1983 Ferguson 1986 Kirkby 1977 White et al 1982
14. Yang 1976 There are two main impediments to the general acceptance of rational regime models the first is the development of a scientifically reasonable understanding of the extremal hypotheses used in the models and the second is the incorporation of a bank stability analysis in the model Researchers at UBC including M Church B Eaton R Millar M Quick have made significant progress on these two issues We have been able to re formulate the extremal hypotheses in such a way as to make the underlying principle more easily understood Eaton et al 2004 Millar 2005 We have tested this principle against observed channel adjustments in the laboratory and in the field where we have observed behaviour that is consistent with our generalized extremal hypothesis Eaton and Church 2004 Eaton and Church 2007 Eaton and Millar 2004 We have also incorporated various bank strength formulations into the regime model which results in a general agreement between model predictions and observed channel dimensions overcoming the long standing criticism that regime models consistently under predict channel width Eaton 2006 Millar 2005 Millar and Quick 1993 Millar and Quick 1998 The regime model is gaining recognition and was awarded the Wiley Award by the British Geomorphological Research Group for best paper in ESPL for 2004 05 It is now being tested by various researchers and environmental consultants in BC who are looking for practical to
15. al approach to river regime Journal of the Hydraulics Division ASCE 108 HY 10 1179 1193 Yang C T 1976 Minimum unit stream power and fluvial hydraulics Journal of the Hydraulics Division Asce 102 HY7 769 784 UBCRM Manual 34 6 5 08 FIELD FORMS SURFACE GRAIN SIZE ANALYSIS SURFACE GRAIN SIZE ANALYSIS STUDY AREA STUDY AREA DATE DATE SAMPLE LOCATION SAMPLE LOCATION SIZE NUMBER OF STONES SIZE NUMBER OF STONES C F CLASS CLASS mm mm 512 362 512 362 362 265 362 265 256 181 256 181 128 181 128 181 90 5 128 90 5 128 64 90 5 64 90 5 45 64 45 64 32 45 32 45 22 7 32 22 7 32 16 22 7 16 22 7 11 16 11 16 8 11 8 11 5 6 8 5 6 8 lt 5 6 lt 5 6 Total Total P is the number in each class divided by the total CF is the total proportion of the distribution P is the number in each class divided by the total CF is the total proportion of the distribution finer than the given class the Cumulative Frequency distribution finer than the given class the Cumulative Frequency distribution NOTES NOTES CROSS SECTION SURVEY FOR AUTOMATIC LEVEL AND STADIA ROD CROSS SECTION SURVEY FOR AUTOMATIC LEVEL AND STADIA ROD
16. aled or exceeded only once in any 10 year period of record When there are no stream discharge measurements on the stream of interest it is sometimes possible to estimate the formative flows using regional hydrology analyses For example Eaton et al 2002 present a regional analysis of the peak flows for British Columbia for the mean annual peak flow as well as some less frequent peak flows The present a general equation that relates discharge Q to drainage area of the form Eq 1 Q where is the discharge for a drainage area of 1 km and A is the drainage area in km Eaton et al 2002 present a map of k factors for British Columbia which is reproduced here in color Figure 8 Eaton and Moore 2007 also discuss the variation of peak flows as well as the peak flow generating mechanisms in BC Eg 1 really amounts to a scaling equation that accounts for the fact that peak flows are not linearly related to drainage area It can be used in BC at least to take data for one basin and use it to characterize another ungauged basin The first step is to obtain a k value for your drainage basin In BC k values can be read from the map in Figure 8 which presents Eaton et al s analysis One can also estimate k values for an ungauged 2 A geo registered TIF file of this map can be downloaded from Brett Eaton s web page http www geog ubc ca beaton downloads html UBCRM Manual 10 6 5 08 study site usin
17. ath has been set to the folder containing the m files the UBCRM model is run by simply typing UBCRM on the command line then pressing enter A window prompting the user for input parameters pops up automatically Figure 20 Once the data has been entered press OK to continue The program evaluates a wide range of potential and stable channel geometries ranging from wide and shallow to narrow and deep then selects the channel associated with the maximum sediment transport rate The solution can be saved QuickTime and as matrix file The save dialog box appears TIFF LZW decompressor picture automatically Figure 21 if the user wishes to save the results select the desired folder location and file name The file will contain a matrix called State that contains in order i the bed width Poea 1 the trapezoid depth Y iii side slope Fi 20 Data input wind ae the di angle iv the shear stress acting on the bed Thea v the shear stress acting on the bank Thank Vi the transport rate m s the precision with which the solution width is known and viii the mean velocity v Once the user has either saved the results or pressed QuickTime and a Cancel the model analysis TIFF LZW decompressor are needed to see this picture and results are presented as shown in Figure 22 The results are best viewed by enlarging the figure panel to the maximum possible size for y
18. be surveyed as accurately as possible The best method for estimating the channel slope accurately is to construct a fairly detailed longitudinal profile along the channel thalweg recording the elevation of the thalweg as well as the water depth and plotting those data against distance along the thalweg Any bedrock boulder of LWD steps should be carefully surveyed and the locations of the surveyed cross sections should be noted on the longitudinal profile The survey should extend of multiple sequences of pools and riffles in order to determine a suitable reach UBCRM Manual 15 6 5 08 average slope Once the survey data has been reduced and the longitudinal profile has been plotted a linear regression should be fit to both the bed elevation and the water surface elevation data While there will be trends in the data produced by sequences of riffles and pools there should not be any systematic changes in slope throughout the study reach such changes will show up at systematic deviations in the data from the linear fit over distances greater than the riffle pool spacing which is about 3 to 7 channel widths If there are systematic changes in slope evident on the longitudinal profile consider subdividing the reach and fitting the linear regressions and the regime model to the upper and lower parts of the reach If significant trends still persist consider further subdividing the reach In cases where there are no discharge measurements f
19. ch average channel slope It also requires knowledge of the range of grain sizes found on the bed surface of the channel and of the relative erodiblility of the channel banks The following sub sections define the input variables that the model requires and describes how suitable estimates can be determined 3 1 FORMATIVE DISCHARGE The regime model is predicated upon the idea that channel morphology is related to the flows carried by the stream averaged over some suitably long timescale see Eaton et al 2004 Gravel bed streams only ever mobilize their bed material during periods of relatively high flow The bankfull flow which is the discharge that just fills the channel up to the level of the floodplain surface is the best representative of the formative discharge since flows less than bankfull are not capable of doing much geomorphic work and since flows greater than bankfull spill out onto the floodplain contributing little to the flow acting directly upon the stream channel For streams that are gauged estimating the formative discharge is relatively straightforward In Canada the Water Survey of Canada WSC collects and archives data on streamflow for selected streams The available data can be downloaded from the WSC website http www wsc ec gc ca index_e cfm cname main_e cfm An example of a typical data report from WSC is shown in Figure 7 The WSC provides estimates of the maximum instantaneous peak flow in each year of re
20. cord as well as the highest average daily discharge for each year referred to as the maximum daily discharge Since channel morphology changes over relatively long time periods it is probably most appropriate to use an average of the maximum daily discharge values to estimate the formative discharge Given that the climate variability produces shifts in stream flow UBCRM Manual 8 6 5 08 regimes e g Moore 1996 it is probably best to represent the formative discharge using a 5 year to 10 year average immediately prior to the period of interest QuickTime anda TIFF Uncompressed decompressor are needed to see this picture Figure 7 Sample WSC data for Fishtrap Creek UBCRM Manual 9 6 5 08 In general the average mean annual peak flow estimated from stream flow records is similar to the bankfull discharge which is determined by the average channel morphology However if one is considering streams in arid regions the formative discharge may be much less frequent and the evidence of large infrequent floods may persist over decades or even centuries Baker 1977 The range in peak flow magnitudes also tends to be higher in more arid regions which can also affect the frequency of the formative discharge For example analysis of the streams in Colorado described by Andrews 1984 by Eaton and Church 2007 seem to be associated with formative discharges having return periods of about 10 years i e the flows likely to be equ
21. decompressor are needed to see this picture Figure 1 Solver parameters for UBCRM QuickTime and a TIFF LZW decompressor are needed to see this picture Figure 2 Solver parameters for UBCRM_H If you do not see the Solver tool in the Tools menu then it has not been installed on your machine It is an add in program as described in Figure 3 This is a two step process UBCRM Manual 4 6 5 08 involving first installing the add in on your computer then loading the add in into Excel The procedures for doing this vary with the version of Excel being used so please consult the Excel Help menu for details QuickTime a TIFF LZW decompressor are needed to see this picture Figure 3 Excel help for add in programs It is recommended that you keep a backup copy of the original UBCRM xls file on your hard drive and create working copies in which to conduct analyses saving the program under a different name has no effect on how the program works Should the formulae in the working spreadsheet or the settings in the Solver program become irrevocably changed then reverting to the backup file should fix the problem 2 2 MATLAB VERSION First you need to have a working copy of MATLAB The programs were written with MATLAB version 7 4 0 R2007a There are two versions of the MATLAB code located in two separate folders The folder UBCRM_files contains the programs necessary to run a regime model using Millar and Quic
22. e hydraulic radius remains nearly constant In contrast the transport capacity varies by over an order of magnitude An important check is to verify that the predicted channel shape is in fact stable Parker 1976 expressed the threshold for channel braiding as a function of channel shape W d and flow conditions Fr In a similar analysis Fredsoe 1978 associated the onset of braiding with a W d of approximately 50 For example setting 250 m s produces a W d ratio of just over 50 suggesting at least the possibility that such an increase could produce a change in channel pattern If a model run predicts a channel that appears to be above the threshold for braiding based on Parker s criterion Fredsoe s simplified threshold or some other braiding threshold then assume the channel is divided into two UBCRM Manual 29 6 5 08 stable channels divide the discharge in two and re compute the solution geometry If the resulting channel geometry is still unstable divide by 3 and re compute the solution If a stable configuration can be reached without dividing the channel into an unreasonable number of anabranches the predicted channel state is best interpreted as a wandering channel If dividing does not bring about a stable geometry then the predicted state is fundamentally unstable i e braided 5 2 RESPONSE TO RIPARIAN DISTURBANCE Another potential use of the model is to predict the morphologic response to changes in
23. e study reach the intent is to characterize the bed UBCRM Manual 16 6 5 08 texture along the thalweg of the stream channel Ideally several samples will be taken within the reach of interest Wolman samples are conducted by establishing regular grid using a flexible tape and measuring the intermediate or b axis diameter of every sediment grain at the nodes of the grid The grid spacing i e the distance between nodes should be at least 4 times the diameter of the largest stone to ensure that there is no spatial autocorrelation between subsequent stones on the grid The grains are classified into one of several size classes and the frequency data is used to construct a cumulative frequency distribution see Figure 9 Typically at least 100 grains must be measured to accurately characterize the surface Estimates of the Dos the grain for which 95 of the distribution is finer Dg4 84 of the distribution is finer and Ds the median grain size can be extracted from the cumulative frequency distribution The Dos is used for the bank stability analysis Dg4 is used to estimate n in Eq 4 and Dso is used to estimate the typical sediment transport rate STUDY AREA FISHTRAP CREEK DATE __ SEPT 29 2007 SAMPLE LOCATION HEAD OF BAR 2 MID REACH SIZE NUMBER OF STONES P C F CLASS mm 256 181 0 1 00 128 181 1 1 0 01 0 99 90 5 128 211 1 6 0 06 0 93 64 905
24. eaton ubc ca UBCRM Manual 26 6 5 08 To calibrate the model to your field site try adjusting the input parameters starting with the most uncertain ones usually bank strength but possibly also Manning s n and or Q When you re run the UBCRM model the input values last provided are automatically written to the input dialog box Figure 20 in order to facilitate rapid calibration and sensitivity testing of the model It is recommended that the width and hydraulic radius of the field site be compared against the width and hydraulic radius predicted by the model In order to run the UBCRM_H model set the path and type UBCRM_H on the command line The data requirements are similar Figure 23 The data output is nearly identical too There is one important QuickTime and a feature that requires some comment at large bed TIFF LZW decompressor are needed to see this picture widths 10 m there is a discontinuity in the estimated bank stress This continuity corresponds to the width for which the depth is equal to the specified rooting depth H For d lt H the model Figure 23 Data input window for the UBCRM_H model does not assess bank stability As a result the model is capable of modeling channels for which the solution d is less than H but it assumes that the bank stability does not limit the optimization UBCRM Manual 27 6 5 08 QuickTime anda TIFF LZW decompressor are needed to see this
25. el version These are needed to see this picture equations require only that the relative bank strength formative discharge median grain size and reach average Figure 14 Data requirements and predicted dimensions for the CRB Eg model Millar 2005 slope be known The model is run by simply entering Q S and in the required cells on the worksheet the estimated channel dimensions are re UBCRM Manual 21 6 5 08 calculated automatically Figure 14 This tab also includes cell that calculate the mean velocity shear stress and sediment transport rate The transport rate is calculated using the Parker 1990 equation Running the UBCRM model involves two steps 1 entry of the input parameters and 2 numerical solution of the regime equations The model requires estimates of Q 5 0 Dos for bank stability for sediment transport the data are entered in the column of cells under the heading of Field Characteristics As a point of reference for the user the entered value of is automatically translated into a value of relative bank strength on the worksheet The user must also enter initial values for the channel dimensions under the heading Model Variables These parameters will be adjusted by the numerical di ta sban a TIFF LZW decompressor model as 1t seeks a are needed to see this picture solution subject to the constraints of continuity and bank stability A good startin
26. elatively poorly known in most cases this is really up to the discretion of the user and the calibration should be based on physical Figure 18 Hydraulics section ous of UBCRM model reasoning and an understanding of which input parameters are least well known rather than arbitrary tuning of one parameter or another to get a good fit To run the UBCRM_H model the user follows the same procedure described above The parameters in Field Characteristics are similar but require the user to specify a friction angle for the cohesionless gravel toe typically assumed to be 30 and an effective root cohesion Figure 19 The equivalent vertical bank height equivalent to the riparian vegetation rooting depth is calculated automatically Also the initial values of the Model Variables are slightly different in that the total depth which is calculated automatically see Figure 19 is the sum of the vertical bank height H and the specified trapezoid depth YI The QuickTime and a Excel version of the UBCRM_H model TIFF LZW decompressor are needed to see this picture cannot accurately deal with situations where the channel depth is less than the specified rooting depth so it is recommended that the user verify that Figure 19 Data requirements and initial values for the UBCRM_H model the trapezoid depth Y is greater than Zero UBCRM Manual 24 6 5 08 4 2 RUNNING THE MATLAB CODE Once the p
27. ett 1984 presents an equation for mountain streams 0 02 lt 5 lt 0 04 that relates n to channel slope S and the hydraulic radius approximated here using the mean hydraulic depth d Eq 3 0 325 016 However tests of this equation indicate that it tends to over predict Manning s by on average about 30 Marcus et al 1992 Bathurst 2002 presents a formula for estimating the minimum flow resistance in streams with gradients less than 0 08 m m His formula was originally used to estimate the Darcy Wiesbach friction factor Rearranging to give estimates of yield the equation 1 6 0 547 Eq 4 n z s 24 3 84 where Dg is grain size for which 84 of the surface grains are finer see section on grain size distribution for a description of how to collect and analyze the relevant data and d is the mean hydraulic depth The parameter g is the acceleration of gravity 9 8 m s R is the hydraulic radius of the channel given by the ratio of the cross sectional area for flow A and the wetted perimeter of the channel P R A P If we interpret Jarrett s equation as an upper bound and Bathurst s equation as a lower bound we should expect the In wide shallow channels the hydraulic radius R is nearly identical to the mean hydraulic depth 4 UBCRM Manual 13 6 5 08 average of the two to produce a reasonable estimate of the likely Manning s n value We can also use the upper and l
28. g a nearby gauged catchment since both and A are known k Q A The estimated value can then be used to estimate the formative discharge for the study site though differences in the catchment physiography and climatology will produce large uncertainties in the estimate It is therefore recommended that any region peak flow estimate be checked against the observed channel dimensions in the field as described below QuickTime anda TIFF Uncompressed decompressor are needed to see this picture Figure 8 Map of k factors presented in Eaton and Moore Eaton and Moore 2007 3 2 MANNING S FLOW RESISTANCE PARAMETER This parameter has a significant effect on the model predictions so getting a reasonable value is critical There are a number of references that can be used to estimate flow resistance For example Cowan s 1956 method is a useful approach since it attempts to UBCRM Manual 11 6 5 08 attribute flow resistance to various components of the river form His equation takes the form Eq 2 n n n n n n ms The parameters and typical values are given in Table 1 This is a reasonable technique to apply to large rivers c f Church 1992 provided the channel gradient is not too high However since this scheme relies heavily on the qualitative judgment of the individual assessing the stream channel the predicted value of n is strongly dependent on the individual making the assessment in t
29. g point is to use the Figure 15 Data reguirements and initial values for the values of W and d UBCRM model estimated using the CRB Eg model and to set the side slope angle to the one degree less than the selected angle As the numerical model searches for a solution it will crash if the side slope angle exceeds d it may also crash if the initial values of W and d are very different than the solution values Once the Field Characteristics and the Model Variables have been entered the numerical model is started by selecting Solver from the Tools menu The Solver should appear with all settings completed as shown in Figure 1 Click Solve and the numerical model starts Once the model has reached a solution the Solver Results message box appears Figure 16 there are a number of options allowing the user to accept the results keep Solver Solution or reject them Restore Original Values and to generate reports the most UBCRM Manual 22 6 5 08 useful of which is the Answer report that summarizes the initial and final values and the status of the constraints applied to the solution Figure 17 If the Solver is unable to reach a solution the user will see a slightly different Solver Results message Figure 16 but the same options will appear in this case it is generally best to Restore Original Values rather than Keep Solver Solution If the Solver fails to find a solution the user should 1 verify that the Field Character
30. gth is probably one of the most important factors controlling channel dynamics The best that we can do at this point is to use analyses of previously published datasets to estimate of bank general categories strength indices The most straightforward index of bank strength is that proposed by Millar 2005 He uses the ratio between the bank erosion threshold and the bed erosion threshold as a bank stability index 4 His analysis of Hey and Thorne s 1986 dataset shows that increases systematically with the Bank strength Bed strength 50 Apparent friction angle Figure 11 Relation between relative bank strength and apparent friction angle 4 density of vegetation on the floodplain For type I channels in their dataset grass no trees or shrubs 0 98 effectively 1 For type II channels 1 to 5 percent shrub UBCRM Manual 18 6 5 08 or tree cover 1 17 for type channels 5 to 50 percent shrub or tree cover u 1 41 and for type IV channels gt 50 percent tree or shrub cover 1 92 This index is based on Millar and Quick s 1993 original use of an apparent friction angle to characterize relative bank strength The two indices are very closely related as shown graphically in Figure 11 There have been several analyses relating ww to Hey and Thorne s stream channel classes For example Eaton 2006 calculated representative
31. he field It is recommended that the field crew take numerous photographs to document the channel conditions in order to allow the various judgments to be re evaluated if necessary Table 1 Cowan s 1956 method for estimating Manning s n Parameter Values Sediment Type n Earth 0 020 Rock cut 0 025 Fine gravel 0 024 Coarse gravel 0 028 Degree of cross section irregularity n Smooth 0 000 Minor 0 005 Moderate 0 010 Severe 0 020 Downstream variations in cross section shape nz e g thalweg shifts from side to side Gradual 0 000 Alternating occasionally 0 005 Alternating frequently 0 010 to 0 015 Relative effect of obstructions e g logs boulders Negligible 0 000 Minor 0 010 to 0 015 Appreciable 0 020 to 0 050 Severe 0 040 to 0 060 Vegetation n4 Low 0 005 to 0 010 Medium 0 010 to 0 025 High 0 025 to 0 050 Very high 0 050 to 0 100 Degree of meandering ms Minor sinuosity 1 0 to 1 2 1 00 Appreciable sinuosity 1 2 to 1 5 1 15 Severe sinuosity greater than 1 5 1 30 UBCRM Manual 12 6 5 08 The Cowan method and most other commonly used methods perform poorly in steep mountain streams Marcus et al 1992 According to Chow 1959 mountain streams with gravel cobble boundaries have Manning s n values that average 0 040 ranging from 0 030 to 0 050 Mountain streams with cobble boulder boundaries have higher Manning s values mean of 0 050 ranging from 0 040 to 0 070 Jarr
32. istics have been entered properly with the correct units and 2 adjust the initial Model Values QuickTime and a QuickTime anda TIFF LZW decompressor TIFF LZW decompressor are needed to see this picture are needed to see this picture Figure 16 Solver Results dialog box for a successful run left hand panel and an unsuccessful run right hand panel Once a successful run has been completed the predicted channel dimensions should be compared QuickTime and a TIFF LZW decompressor are needed to see this picture with the observed dimensions It should be noted that the Model Variables represent neither the average width nor the mean hydraulic depth they Figure 17 Answer Report sheet variables that describing a trapezoid with the specified bed width side slope angle and trapezoid depth It is preferable to compare the measured channel dimensions against the variables in the column under the heading Hydraulics Figure 18 which includes the total width W cross sectional area for flow A hydraulic radius R and average velocity v UBCRM Manual 23 6 5 08 The primary means for calibrating the model is by changing the bank strength parameter since this is generally the least well known and since the model predictions are quite sensitive to the selected value However it may be reasonable to ap TIE LAI decompressor adjust the other input parameters in particular Q and n which are also r
33. k s 1993 bank stability analysis The folder UBCRM_H_files contains the files for a model using the stability analysis presented by UBCRM Manual 5 6 5 08 Eaton 2006 Each folder contains one main program file UBCRM m and UBCRM_H m respectively that runs the regime model and three sub programs also m files that are called by the main program There is also a file that keeps track of the last input values used in the model UBCRMdef mat and UBCRMHdef mat To install these programs copy the folders onto your hard drive Then in MATLAB add the folders to the path In the File menu select Set Path Then click on the Add Folder button in the Set Path window Figure 4 and navigate to the folder for the program you are installing UBCRM or UBCRM_JH and click Open This will add the folder to the list of directories in which MATLAB will look for programs and you will return to the Set Path window Then click on Close in the Set Path window MATLAB will then ask you if you want to save the changes to the path if you want to permanently add the folder to the path and the program to the set of programs that MATLAB can access every time it is opened click Yes if you prefer to set the path every time you want to use the program click No QuickTime a TIFF LZW decompressor are needed to see this picture Figure 4 Set Path window in MATLAB To run the main program type UBCRM or UBCRM_H on the command line then pres
34. lysis for regime models of vegetated gravel bed rivers Earth Surface Processes and Landforms DOI 10 1002 esp 1364 Eaton B C and Church M 2004 A graded stream response relation for bedload dominated streams Journal of Geophysical Research Earth Surface 109 F03011 doi 10 1029 2003JF000062 Eaton B C and Church M 2007 Predicting downstream hydraulic geometry A test of rational regime theory Journal of Geophysical Research Earth Surface 112 F03025 doi 10 1029 2006JF000734 Eaton B C Church M and Millar R G 2004 Rational regime model of alluvial channel morphology and response Earth Surface Processes and Landforms 29 4 511 529 Eaton B C and Millar R G 2004 Optimal alluvial channel width under a bank stability constraint Geomorphology 62 35 45 Eaton B C and Moore R D 2007 Chapter 4 Regional Hydrology In R Pike et al Editors Compendium of Forest Hydrology and Geomorphology in British Columbia BC Ministry of Forests Victoria BC Emmett W W 1975 The Channels and Waters of the Upper Salmon River Area Idaho U S Geol Surv Prof Pap 870 A Ferguson R I 1986 Hydraulics and Hydraulic Geometry Progress in Physical Geography 10 1 1 31 UBCRM Manual 33 6 5 08 Fredsoe J 1978 Meandering and braiding of rivers Journal of Fluid Mechanics 84 4 609 624 Hey R D and Thorne C R 1986 Stable channels with mobile gravel beds Journal of the Hydraulics Division ASCE
35. ols for making better decisions about stream channel management We believe that there are numerous potential applications for this model including the replacement of arbitrary and theoretically meaningless hydraulic geometry equations in numerical models of downstream sediment transfer and longitudinal profile evolution and landscape evolution UBCRM Manual 1 6 5 08 TABLE OF CONTENTS Overview of the UBC Regime i Table Of i e E a A A AEE adeasaesadoassancatess Soneeawenaceetes ii ORNS EEEE E EN EEEE 3 2 Installation siu tions ei ieee EEA 3 2 1 EXCEL VErsiOn A E T AS 3 2 21 Wiad cain ihe 5 y Data a E E amp 3 1 16 T amp 3 2 Manning s Flow Resistance 11 3 3 Bankfull Channel Dimensions and Channel Gradient 15 3 4 Surface Grain Size 1 1 16 sad sak baton 18 4 Model Calibration 21 47 Running the Excel Model sresti 21 4 2 Running Matlab Code ei he eve eed 25 5 Evaluating Channel Response
36. or the stream i e no WSC gauge then stream discharge and Manning s both need to be estimated from the bankfull dimensions and the channel slope Once bankfull estimates of W d and S are known from the channel survey a value of n should be selected in order to estimate using Eq 7 Wherever possible the resulting estimate of Q should be compared with estimates from regional hydrology analyses described above in order to verify that i bankfull stage was reasonably identified in the field and ii that the value of n selected was appropriate If necessary and justifiable the value of can be adjusted based on the results of this test The regime model simply uses n in conjunction with S and to estimate the channel depth and from that the shear stress acting on the bed so provided that the values of n S and Q associated with the surveyed bankfull channel are all entered into the regime model it should perform reasonably well 3 4 SURFACE GRAIN SIZE DISTRIBUTION Accurate information on the surface texture is also required to run the model The median grain size is used to estimate the sediment transport rate and grain size characteristic of the coarse tail of the distribution is used to assess channel stability and the flow resistance parameter The best means of determining the surface grain size distribution is by conducting a surface Wolman sample Samples should ideally be conducted on the coarsest part of the bar heads in th
37. our monitor The upper Figure 21 Save dialog box for the UBCRM model left hand panel displays the UBCRM Manual 25 6 5 08 estimated shear stress acting on the channel banks in green and on the channel bed red for the range of channel widths evaluated The critical shear stress for bank erosion is presented as a dashed blue line The lower left hand panel shows the estimated sediment transport rate for the range of stable channel geometries evaluated The relation between sediment transport and bed width for the channel close to the peak sediment transport capacity are shown in the lower right hand panel The peak of that relation is the solution selected by the model The solution geometry width hydraulic radius mean velocity bed and bank shear stress and sediment transport capacity are printed in the upper right hand panel of the figure QuickTime and a TIFF LZW decompressor are needed to see this picture Figure 22 Results and analysis summary for the model If the model crashes fails to converge verify that the user inputs have all been entered and that they are in the correct units If there does not appear to be any obvious problem with the data inputs try closing MATLAB and then restarting it If there is still some sort of error try downloading the code files from the UBC Regime Model website and then replace the files on your hard drive If the error still persists report it to Dr Brett Eaton brett
38. ower bounds to test the sensitivity of the model to the choice of values that are selected The best and most difficult option is always to use field measurements to back calculate n values for a study site This requires that the stream flow at or near the bankfull discharge be directly measured If the cross sectional area of flow A in m the mean water surface width W in m and mean water surface gradient S are also measured n can be estimated from the following equations Eq 5 Q Av Wdv where is the discharge m s and v is the mean velocity Manning s equation is Eq 6 v Combining Eq 5 and Eq 6 and then isolating n we get Wd R S wd s Eq 7 Q n n WARS Wa Ss Ea 8 3 Q Q The values of W d R and S are those associated with the measured discharge not the bankfull channel dimensions However the bankfull channel dimensions as surveyed at low flow can be used to estimate both the discharge and the flow resistance parameter n UBCRM Manual 14 6 5 08 3 3 1 CHANNEL DIMENSIONS AND CHANNEL GRADIENT While it is usually quite difficult to gauge streamflow in a stream at near formative flows it is much simpler to estimate and survey the bankfull channel dimensions In the field the bankfull channel dimensions must be carefully identified and several cross sections should be surveyed in order to determine the average bankfull width and de
39. pth The field crew surveying the channel should take care to differentiate between low terraces generated by either systematic or local channel degradation since mistaking a low terrace for a floodplain surface will dramatically inflate the estimated bankfull discharge Ideally the field crew will be able to survey at least 10 channel cross sections with the reach of interest demarcating the top of bank bottom of bank edge of vegetation any indictors of the bankfull elevation e g LWD accumulations vegetated bar tops clearly active floodplain surfaces etc the water s edge at the time of survey and the thalweg location They should also collect enough data to faithfully reproduce the general cross section shape but there is little to be gained by a hyper detailed survey of the details of the cross sectional topography Cross sections should be regularly spaced about one to two channel widths apart from each other An automatic level and stadia rod is adequate for this sort of survey but care must be taken when reading and recording the upper middle and lower stadia values since reading and transcription errors are common The preferred methodology for surveying channel cross sections is to use a Total Station to survey the cross sections While these instruments are more expensive and more complicated to use they are far more accurate and there are few opportunities for mistakes to be make The average channel slope should also
40. s enter Assuming that you are using the UBCRM version if you get the message undefined UBCRM Manual 6 6 5 08 function or variable UBCRM then the path has not been set properly Other error messages are likely to indicate that the code has been changed in the m files that execute the program or that one or more of the sub programs are missing Note that the file names are case sensitive and file transfer between operating systems can result in changes to the file name cases The file names as they are meant to appear are shown in Figure 5 and Figure 6 if the file names have been changed to all upper case or all lower case letters you will have to change them back to their original case sensitive names as shown QuickTime and a TIFF LZW decompressor are needed to see this picture Figure 5 Case sensitive file names for UBCRM version of the regime model MATLAB code QuickTime and a TIFF LZW decompressor are needed to see this picture Figure 6 Case sensitive file names for UBCRM_H version of the regime model MATLAB code UBCRM Manual 7 6 5 08 3 DATA REQUIREMENTS In order to set up and run the model it is necessary to have information on the stream channel under consideration this usually requires measurements taken in the field The model requires an estimate of the formative discharge for the channel the characteristic Manning s flow resistance parameter at the formative discharge and the rea
41. with Hey and Thorne s 1986 type II channels the relative bank strength is more than 3 times the strength of the bed The UBCRM Manual 30 6 5 08 dimensions and transport capacity predicted by the UBCRM are not substantially different from the UBCRM_H predictions The first analysis was conducted by re computing the regime solution for H values ranging from 0 55 m down to 0 05 m The results are shown in Figure 26 The width responds by increasing by up to 150 above the initial value while the hydraulic radius and the transport capacity are reduced by close to 50 If the value of H falls below 30 of the original value i e to about 0 15 m then it is possible that the channel pattern could change to either a wandering or a braided one Since dividing the channel into two threads brings the predicted channels back into the stable single thread regime taken here to correspond to W d lt 50 it is more likely that the channel will adopt a wandering pattern Therefore the model predicts widening and a possible change in channel pattern in response to the loss of bank strength Sensitivity to changes in H Fishtrap Creek Oo W Wo R Ro Qb Qbo 2 5 2 1 5 1 gt 0 5 0 1 0 9 0 8 0 7 0 6 0 5 H Ho Figure 26 Sensitivity analysis using the UBCRM_H model Q 7 35 m s Dso 40 mm Dos 181 mm 5 0 019 n 0 06 H 0 55 m UBCRM Manual 31 6 5
Download Pdf Manuals
Related Search