EVALUATION OF SLOPE EFFECTS ON SOIL EROSION OF OFF-ROAD VEHCILE 
TRAILS USING WEPP 
 
Except where reference is made to the work of others, the work described in this thesis is 
my own or was done in collaboration with my advisory committee.  This thesis does not 
include proprietary or classified information.   
 
_______________________________ 
Jonathan Donald Melton 
 
Certificate of Approval: 
 
___________________________   ___________________________ 
Timothy P. McDonald, Co-Chair   Wesley C. Zech, Co-Chair 
Associate Professor     Assistant Professor 
Biosystems Engineering    Civil Engineering 
 
___________________________   ___________________________ 
Puneet Srivastava     Dan A. Brown 
Assistant Professor     Gottlieb Associate Professor 
Biosystems Engineering    Civil Engineering 
 
___________________________   ___________________________ 
John M. Grace      George T. Flowers 
Research Engineer     Interim Dean 
US Forest Service     Graduate School 
 
 
EVALUATION OF SLOPE EFFECTS ON SOIL EROSION OF OFF-ROAD VEHICLE 
TRAILS USING WEPP  
 
 
Jonathan Donald Melton 
 
 
A Thesis 
Submitted to  
the Graduate Faculty of 
Auburn University 
in Partial Fulfillment of the 
Requirements for the 
Degree of 
Master of Science 
 
 
 
Auburn, Alabama 
May 10, 2008
 
iii 
 
EVALUATION OF SLOPE EFFECTS ON SOIL EROSION OF OFF-ROAD VEHICLE 
TRAILS USING WEPP 
 
Jonathan Donald Melton 
 
 
Permission is granted to Auburn University to make copies of this thesis at its discretion, 
upon request of individuals or institutions and at their expense.  The author reserves all 
publication rights. 
 
________________________ 
Signature of Author  
________________________ 
Date of Graduation 
 
 
iv 
 
VITA 
 Jonathan Donald Melton, son of Donald and Patricia Melton, was born on March 
11, 1982 in Mobile, Alabama.  He grew up in Stapleton, Alabama, with his older sister 
Cindy.  Upon graduating from Baldwin County High School in May of 2000, he entered 
Faulkner State Community College in Bay Minnette, Alabama.  In August of 2001 he 
enrolled in the College of Engineering at Auburn University in Auburn, Alabama with a 
Major in Biosystems Engineering.  From August 2002 to August 2003, he participated in 
Auburn University?s Cooperative Education Program working with the USDA?s US 
Forest Service Southern Research Station in Auburn, Alabama. Graduating from Auburn 
University in May of 2005 with a Bachelor?s of Biosystems Engineering Degree, he 
entered Auburn University?s graduate program in Civil Engineering.  He worked as a 
graduate research assistant in Biosystems Engineering Department.  He is currently 
employed by Vision Engineering and Development Services in Sevierville, TN. 
 
v 
 
THESIS ABSTRACT 
EVALUATION OF SLOPE EFFECTS ON SOIL EROSION OF OFF-ROAD VEHICLE 
TRAILS USING WEPP 
Jonathan Donald Melton 
Master of Science, May 10, 2008 
(B.B.S.E., Auburn University, 2005) 
 
143 Typed Pages 
Directed by Timothy P. McDonald and Wesley C. Zech 
 
Erosion and sediment loss from the use of off-road vehicles (ORV) is an 
escalating problem on national forest lands.  The increasing number of ORV riders 
coupled with decreasing riding areas has caused the United States Forest Service and 
other agencies to begin development of best management practices (BMPs) that will 
reduce the amount of erosion and sediment loss occurring on trails.  
In order to quantify the sediment loss from ORV trails, two trail sections of the 
Kentuck ORV trail system in the Talladega National Forest were equipped with sediment 
sampling devices and rainfall data collectors.  These trail sections were located on a mild 
slope and a steep slope on the existing trail.  Storm data were collected and analyzed for 
five storms during a nine month period.  The storm producing the highest amount of soil 
erosion produced approximately 0.78 tonnes per hectare and 1.46 tonnes per hectare of 
sediment from mild and steep slopes, respectfully.  These data were used to evaluate an 
 
vi 
 
erosion model that would allow land managers to predict sediment loss on general trail 
sections throughout the trail system.  The Water Erosion Prediction Project (WEPP) 
model was used because of its ability to model forest roads.  The instrumented trail 
sections were modeled using WEPP and compared to sediment and runoff  data from 
field experiments  to calibrate the model.  Upon calibration, a Nash-Sutcliffe efficiency 
(NSE) of 0.45 was achieved for runoff while a R
2
 value of 0.43 and NSE of -0.36 was 
established for soil erosion.  In general, the calibrated model predicted lower runoff, but 
much higher sediment production, than observed. A second model was calibrated in order 
to calibrate for sediment production only, which produced a NSE of 0.24.  The mild and 
steep slope data was separated and modeled for each individual slopes.  An attempt to 
validate these models was performed by using the opposite slope for validation.  The 
results from the validation showed that the model was not as good a predictor of expected 
sediment production for either slope. In addition to the slope comparison, the data was 
divided into trafficked and non-trafficked data which produced a NSE of -0.11 and 0.76, 
respectfully.  The moderate NSE for the non-trafficked data was expected due to the lack 
of the additional soil disturbance from the ATV traffic.   
The calibrated model for sediment only was used to simulate erosion on varying 
slopes.  From these simulations, recommendations for water diversion structure spacing 
were developed.  Water diversion structures carry runoff from the trail into the 
surrounding forested areas, reducing the velocity and sediment concentration of water 
flow over the trail surface.  The spacing recommendations developed from this study can 
be used as BMPs to aid land managers in designing systems to reduce the impacts on 
ORV trails. 
 
vii 
 
ACKNOWLEDGEMENTS 
 
 The author would like to thank Auburn University, Timothy McDonald, and his 
Graduate Committee for the opportunity to perform the research associated with this 
thesis.  The author would like to thank his parents, Donald and Patricia Melton, for the 
financial and emotional support throughout his educational experience.  The author 
would also like to thank his fianc?e, Jennifer Lofty, for her support during the research 
process.  The author would like to thank Corey Kichler and Gavin Jackson for assistance 
with the installation of research equipment and processing of data in the laboratory.   
 
 
 
viii 
 
Style manual used Auburn University Graduate School Guide to Preparation and 
Submission of Theses and Dissertations 
 
Computer software used Microsoft Office 2003 --- Microsoft Word 2003, Microsoft 
Excel 2003 --- Water Erosion Prediction Project (WEPP) model --- Hoboware 
 
 
 
 
 
ix 
 
TABLE OF CONTENTS 
 
LIST OF FIGURES .......................................................................................................... xii 
LIST OF TABLES........................................................................................................... xvi 
CHAPTER 1: INTRODUCTION....................................................................................... 1 
1.1 Problem Statement.................................................................................................... 1 
1.2 Objectives ................................................................................................................. 4 
CHAPTER 2:  LITERATURE REVIEW........................................................................... 5 
2.1 Introduction............................................................................................................... 5 
2.2 Management Practices .............................................................................................. 7 
2.3 Prediction of Soil Loss.............................................................................................. 9 
2.4 Summary................................................................................................................. 13 
CHAPTER 3:  RESEARCH PROCEDURES .................................................................. 14 
3.1 Introduction............................................................................................................. 14 
3.2 Site Location........................................................................................................... 14 
3.3 Sampling Description........................................................................................ 18 
3.3.1 Sediment Sampling .......................................................................................... 18 
3.3.2 Rainfall Data .................................................................................................... 20 
3.3.3 Sediment Analysis ........................................................................................... 20 
3.4 Erosion Prediction Modeling with WEPP .............................................................. 21 
 
x 
 
3.4.1 Weather Generator ........................................................................................... 22 
3.4.2 Hillslope Designation and Model Calibration ................................................. 22 
3.4.3 Statistical Analysis........................................................................................... 27 
3.4.4 Model Validation ............................................................................................. 28 
CHAPTER 4:  RESULTS................................................................................................. 29 
4.1 Runoff Sampling..................................................................................................... 29 
4.2 WEPP Model Calibration and Validation............................................................... 32 
4.2.1 Runoff Calibration ........................................................................................... 32 
4.2.2 Sediment Calibration ....................................................................................... 35 
4.2.3 Sediment ? Only Calibration ........................................................................... 37 
4.2.4 WEPP Model Validation.................................................................................. 39 
4.2.5 Effect of Traffic ............................................................................................... 42 
4.3 BMP Simulation ..................................................................................................... 46 
CHAPTER 5:  CONCLUSION ........................................................................................ 49 
5.1 Objective 1.............................................................................................................. 49 
5.2 Objective 2.............................................................................................................. 50 
5.3 Objective 3.............................................................................................................. 50 
5.4 General Conclusion................................................................................................. 51 
5.5 Future Research ...................................................................................................... 51 
REFERENCES ................................................................................................................. 54 
 
xi 
 
APPENDIX A:  Rainfall and Sediment Sampling............................................................ 59 
APPENDIX B:  Rainfall and Runoff Data ....................................................................... 70 
APPENDIX C:  Parameter Calibration for Runoff and Sediment Production ................. 73 
APPENDIX D:  Parameter Calibration for Sediment Production .................................... 78 
APPENDIX E:  Calibration of Mild Slope....................................................................... 88 
APPENDIX F:  Calibration of Steep Slope ...................................................................... 94 
APPENDIX G:  Calibration of Trafficked Period .......................................................... 100 
APPENDIX H:  Calibration of Non-Trafficked Period.................................................. 105 
APPENDIX I:  Tipping Bucket Design.......................................................................... 112 
APPENDIX J:  Water Diversion Spacing....................................................................... 122 
 
 
xii 
 
LIST OF FIGURES 
 
Figure 3.1 -- Location of research site.............................................................................. 15 
Figure 3.2 -- Trail map...................................................................................................... 16 
Figure 3.3 -- US Forest Service bull dozer performing maintenance............................... 17 
Figure 3.4 -- Tipping bucket sampler during installation on the mild slope..................... 20 
Figure 3.5 -- Mild slope with insloped unrutted forest road-bare ditch management layer.
................................................................................................................................... 23 
Figure 3.6 -- Steep slope with insloped unrutted forest road-bare ditch management layer.
................................................................................................................................... 24 
Figure 4.1 -- Cumulative runoff for 12/22/2006............................................................... 31 
Figure 4.3 -- Actual runoff compared to predicted runoff after calibration using WEPP. 35 
Figure 4.4 -- Actual sediment yield compared to predicted sediment yield after calibration 
using WEPP. ............................................................................................................. 37 
Figure 4.5 ? Model calibrated for sediment, without previous calibration for runoff...... 38 
Figure 4.6 ? Validation using model parameters estimated from mild slope data for the 
steep slope................................................................................................................. 40 
Figure 4.8 ? Traffic vs. Non-Traffic period...................................................................... 42 
Figure 4.9 ? Calibrated data for trafficked period ............................................................ 44 
 
xiii 
 
Figure 4.10 ? Calibration of non-trafficked period........................................................... 44 
Figure A.1 ? Cumulative rainfall for 11/15/2006. ............................................................ 60 
Figure A.2 ? Rainfall hydrograph for 11/15/2006............................................................ 60 
Figure A.3 ? Cumulative runoff for mild and steep slope plots for 11/15/2006. ............. 61 
Figure A.4 -- Runoff intensity for 11/15/06...................................................................... 61 
Figure A.5 ? Cumulative rainfall for 12/22/2006. ............................................................ 62 
Figure A.6 --Rainfall hydrograph for 12/22/2006. ........................................................... 62 
Figure A.7 ? Cumulative runoff for mild and steep slope plots for 12/22/2006. ............. 63 
Figure A.8 -- Runoff intensity for 12/22/06...................................................................... 63 
Figure A.10 ? Rainfall hydrograph for 2/1/2007.............................................................. 64 
Figure A.11? Cumulative runoff for mild and steep plots for 2/1/2007........................... 65 
Figure 12 -- Runoff intensity for 2/1/07. .......................................................................... 65 
Figure A.13 ? Cumulative rainfall for 3/1/2007. .............................................................. 66 
Figure A.14 ? Rainfall hydrograph for 3/1/2007.............................................................. 66 
Figure A.15 ? Cumulative runoff for mild and steep slope plots for 3/1/2007. ............... 67 
Figure A.16 -- Runoff intensity for 3/1/07........................................................................ 67 
Figure A.17 ? Cumulative rainfall for 3/15/2007. ............................................................ 68 
Figure A.18 ? Rainfall hydrograph for 3/15/2007............................................................ 68 
Figure A.19 ? Cumulative runoff for mild slope and steep slope for 3/15/2007.............. 69 
Figure A.20 -- Runoff intensity for 3/15/2007.................................................................. 69 
Figure C.1 ? Calibration of effective hydraulic conductivity for runoff and sediment .... 75 
Figure C.2 ? Calibration of critical shear for runoff and sediment................................... 76 
Figure C.3 ? Calibration of rill erodibility for runoff and sediment................................. 77 
 
xiv 
 
Figure C.4 ? Calibration of interrill erodibility for runoff and sediment.......................... 78 
Figure D.1 ? Calibration of rill erodibility........................................................................ 80 
Figure D.2 ? Calibration of critical shear for sediment production.................................. 82 
Figure D.3 ? Calibration of interrill erodibility for sediment production......................... 85 
Figure D.4 ? Calibration of effective hydraulic conductivity for sediment production. .. 87 
Figure E.2 ? Calibration of rill erodibility on mild slope. ................................................ 91 
Figure E.3 ? Calibration of critical shear on mild slope................................................... 92 
Figure E.4 ? Calibration of effective hydraulic conductivity on mild slope. ................... 93 
Figure F.1 ? Calibration of interrill erodibility on steep slope. ........................................ 96 
Figure F.2 ? Calibration of rill erodibility on steep slope................................................. 97 
Figure F.3 ? Calibration of critical shear on steep slope. ................................................. 98 
Figure F.4 ? Calibration of effective hydraulic conductivity on steep slope.................... 99 
Figure G.1 ? Calibration of interrill erodibility for trafficked period............................. 101 
Figure G.2 ? Calibration of rill erodibility for trafficked period. ................................... 102 
Figure G.3 ? Calibration of critical shear for trafficked period...................................... 103 
Figure G.4 ? Calibration of effective hydraulic conductivity for trafficked period. ...... 104 
Figure H.1 ? Calibration of interrill erodibility for non-trafficked period. .................... 107 
Figure H.2 ? Calibration of rill erodibility for non-trafficked period............................. 108 
Figure H.3 ? Calibration of critical shear for non-trafficked period............................... 109 
Figure H.4 ? Calibration of effective hydraulic conductivity for non-trafficked period.111 
Figure I.1 -- Tipping bucket design. ............................................................................... 113 
Figure I.2 ? Tipping bucket base. ................................................................................... 113 
Figure I.3 ? Middle and vertical member drawings........................................................ 115 
 
xv 
 
Figure I.4 ? Horizontal member drawing. ...................................................................... 116 
Figure I.5 ? Bottom mount drawing. .............................................................................. 117 
Figure I.6 ? Tipping bucket axle mount drawing. .......................................................... 118 
Figure I.7 ? Rainfall chamber drawing........................................................................... 119 
Figure I.8 ? Rainfall chamber divider drawing............................................................... 120 
Figure I.9 ? Tipping bucket design and parts location.................................................... 121 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
xvi 
 
LIST OF TABLES 
 
Table 4.1 -- Rainfall Data. ................................................................................................ 30 
Table 4.2 -- Sediment yield............................................................................................... 31 
Table 4.3 -- Soil Test Results............................................................................................ 32 
Table 4.4 -- Calculated Soil Parameter Values................................................................. 33 
Table 4.5 -- Values used in to Calibrate WEPP................................................................ 36 
Table 4.6 -- Model Calculated for sediment only............................................................. 38 
Table 4.7 -- Mild Slope Calibration.................................................................................. 40 
Table 4.8 -- Steep Slope Calibration................................................................................. 41 
Table 4.11 ? Summary of Calibration Data...................................................................... 45 
Table 4.12 -- Recommended spacing for water diversion structures................................ 47 
Table 4.13 -- Water Diversion Spacing developed by Brodbeck ..................................... 48 
Table 4.14 -- Recommended Distances between Water Bars on Skid Trails and Logging 
Roads after Logging is Complete (Brinker, 1995).................................................... 48 
Table B.1 -- Rainfall and Runoff Data collected from sites. ............................................ 71 
Table B.2 -- Sediment Data from research sites. .............................................................. 72 
Table C.1 -- Calibration of Effective Hydraulic Conductivity ......................................... 74 
Table C.2 -- Calibration of Critical Shear......................................................................... 75 
 
xvii 
 
Table C.3 -- Calibration of Rill Erodibility ...................................................................... 76 
Table C.4 --  Calibration of Interrill Erodibility ............................................................... 77 
Table D.1 -- Calibration for Rill Erodibility..................................................................... 79 
Table D.2 -- Calibration for Critical Shear ....................................................................... 81 
Table D.3 -- Calibration for Interrill Erodibility............................................................... 84 
Table D.4 -- Calibration for Effective Hydraulic Conductivity........................................ 86 
Table E.1 -- Calibration of Interrill Erodibility ................................................................ 89 
Table E.2 -- Calibration of Rill Erodibility....................................................................... 90 
Table E.3 -- Calibration of Critical Shear......................................................................... 91 
Table E.4 -- Calibration of Effective Hydraulic Conductivity ......................................... 92 
Table F.1 -- Interrill Erodibility Calibration..................................................................... 95 
Table F.2 -- Rill Erodibility Calibration ........................................................................... 96 
Table F.3 -- Critical Shear Calibration ............................................................................. 97 
Table F.4 -- Effective Hydraulic Conductivity Calibration.............................................. 98 
Table G.1 -- Interrill Erodibility Calibration .................................................................. 101 
Table G.2 -- Rill Erodibility Calibration......................................................................... 102 
Table G.3 -- Critical Shear Calibration........................................................................... 103 
Table G.4 -- Effective Hydraulic Conductivity .............................................................. 104 
Table H.1 -- Interrill Erodibility Calibration .................................................................. 106 
Table H.2 -- Rill Erodibility Calibration......................................................................... 107 
Table H.3 -- Critical Shear Calibration........................................................................... 108 
Table H.4 -- Effective Hydraulic Conductivity .............................................................. 110 
Table I.1 ? Tipping Bucket Parts List............................................................................. 114 
 
xviii 
 
Table J.1 -- Recommended spacing for water diversion structures................................ 123 
Table J.2 -- Water Diversion Spacing developed by Brodbeck...................................... 123 
Table J.3 -- Recommended Distances between Water Bars on Skid Trails and Logging 
Roads after Logging is Complete (Brinker, 1995).................................................. 124 
Table J.4 ? Water Bar and Broad-based Dip spacing as recommended by the Alabama 
BMP Manual for Forest Operations........................................................................ 125 
 
1 
 
CHAPTER 1: INTRODUCTION 
 
1.1 Problem Statement 
Over recent decades, the use of off-road vehicles (ORV) for recreation has 
increased rapidly.  Outdoor enthusiasts often use all-terrain vehicles (ATV) to traverse 
trails on which ordinary vehicles are unable to operate.  The thrill and excitement of 
going to places that were inaccessible in the past has been a major contributor to the 
growth of ATV use.  According to the United States Department of Agriculture (USDA) 
Forest Service, over 36 million people rode ATVs in 2000 in the United States (USDA, 
Office of Communication).  In 2003, five percent of the visitors to national forests and 
grasslands consisted of ORV users (Brodbeck, 2005).  Because of potential 
environmental problems associated with their use, this increase in demand for ATV 
recreation creates a need to protect our public land and its natural resources for future 
generations. 
 In 2003, the Chief of the US Forest Service, Dale Bosworth, identified four major 
threats to our national forest lands.  The four major threats were fire and fuels, invasive 
species, fragmentation, and unmanaged recreation (USDA, Office of Communications, 
2004).  The use of ORV on national forest lands falls under the category of unmanaged 
recreation.  Unmanaged ORV use, which includes ATVs, is creating a number of 
 
2 
 
undesirable impacts on national forest lands including (1) user-created, unplanned roads 
and trails, (2) severely eroded soils, (3) damaged wetlands and harm to wetland species, 
and (4) habitat destruction and degraded water quality due to trail-generated sediment 
(Foltz, 2006).  This major threat to our national forest systems has caused significant 
changes in management strategies for the Forest Service to combat this and other major 
threats, but their mission remains the same, ?to sustain the health, diversity, productivity 
of the Nation?s forests and grasslands to meet the needs of present and future 
generations? (USDA, Office of Communications, 2004).   
 ORV trails, when heavily used and left unmanaged, can cause significant 
environmental impacts.  The erodibility of soil, terrain, and type of vegetation are some 
primary factors affecting degradation of trail surfaces.  Soil erosion is one of the main 
adverse impacts associated with ORV trails.  Erosion is a concern because it tends to 
degrade water quality and potentially can adversely impact plant and wildlife habitat 
(Novotny, 2003).  Erosion, for example, can lead to sedimentation within water bodies, 
causing loss of spawning habitat for species that require gravel-bottom streams for 
reproduction (Novotny, 2003). Sedimentation can also kill some specialized types of 
vegetation that require undisturbed stream habitat for survival (Novotny, 2003). 
The use of site-specific management is a key tool in decreasing the impacts 
caused by ORV on the trails.  Site-specific management involves selecting the best 
management practice (BMP) for a given area within an entire trail system, then adapting 
the practice to make it correspond to each specific circumstance. A broad based dip, for 
example, is used to divert runoff from bare soil surfaces and may be an appropriate 
erosion control tool for a general location, but the design for a given spot in a trail system 
 
3 
 
may require alteration from standard practice in order to effectively collect water, divert 
it off the trail, and reduce its velocity by spreading the flow over a large area.  These site-
specific changes to common practices are necessary to manage trails when they are 
located on sensitive soils, steep slopes, or in lowland areas. 
The use of site-specific BMPs is always the optimal approach when trying to 
minimize impacts from man-made sources, but information is not always available to  
base decisions about particular situations.  Managers of the Kentuck ORV area of the 
Talladega National Forest, for example, currently use several site-specific management 
practices to maintain the trail system.  Broad based dips, water-bars, and wing ditch 
turnouts are applied as is felt appropriate to control erosion, yet problems with the trail 
system persist.  The implemented structures are intended to shed water from the trail to 
eliminate puddling of water, rutting, and channeling of flow into waterways.  Due to 
heavy trail traffic in both wet and dry periods, high levels of wheel slip associated with 
ATVs, and steep terrain in which the trail is located, these maintenance techniques tend 
to be effective short periods of time.  Once the trail structure has been compromised, the 
trail tends to become a source of increased erosion.  The economic cost of replacing these 
erosion control structures is high, as is the environmental impact associated with not 
keeping them functioning at peak effectiveness.  Information is needed by trail managers 
on which to base decisions concerning frequency of maintenance practices, as well as 
improved guidelines for application of erosion control structures in general. 
The purpose of this study was to investigate erosion processes on ORV trails; 
specifically, the effect of slope on sediment generation. In addition, the intent was to 
further develop analytical tools for making decisions regarding the use of sediment 
 
4 
 
control structures on ORV trails. Previous research has shown that the Water Erosion 
Prediction Program (WEPP) can be used to estimate the sediment load generated at a 
stream crossing on the Kentuck ORV Trail. The present study was designed to investigate 
primary production of sediment as a function of trail slope and, to the extent possible, 
calibrate the WEPP model with the erosion data for use as an analytical tool in making 
decisions regarding trail design, layout and maintenance. The outcomes of the study 
should be an understanding of the expected sediment generation of an ORV trail as a 
function of slope, and a database of parameters that serve as inputs to the WEPP program 
and are appropriate for analyzing alternative management strategies for the Kentuck trail 
system. The erosion data will supplement the fundamental understanding of soil 
detachment and transport from heavily disturbed and compacted trail surfaces. The 
modeling component of the research will provide an additional opportunity to verify the 
WEPP program as an analytical tool. Finally, the work will add to the general literature 
on erosion from ORV trails, a subject about which there is a relative lack of information. 
1.2 Objectives 
The objects of this research  are: 
1.) Outfit two sections of the Kentuck ORV trail, one each on a steep and mild 
slope with instrumentation to measure total sediment transport. 
2.) Calibrate WEPP for the Kentuck ORV trail for total sediment loading. 
3.) Demonstrate the use of WEPP as a tool for management planning on ORV 
trails. 
 
 
5 
 
CHAPTER 2:  LITERATURE REVIEW 
2.1 Introduction 
 Sediment has been identified as one of the most recognized non-point source 
(NPS) pollutants of streams, lakes, and estuaries (Novotny, 2003).  In forested areas, the 
predominant source of sediment is from the construction and maintenance of access 
roads.  Since ORV trails and forest roads share the same characteristics and theoretical 
principles in terms of location, use, and in some cases size, most ORV trails are 
considered as forest roads. Extensive research has been performed on erosion of forest 
roads but very little research has been performed on ORV trails.   Research has shown 
that up to 90% of sediment producing erosion in forested areas can be attributed to access 
roads, while the erosion from the undisturbed areas are negligible (Appelboom, 2001; 
Van Lear, 1995).  The erosion from forest roads is a major component in quality 
degradation of forest streams and wildlife (Novotny, 2003). The sediment can clog 
spawning beds, shorten the life of reservoirs, and degrade drinking water.  Due to these 
factors, road management is an important component in preserving and maintaining 
healthy forest throughout the nation (Grace, 2006).  As demand increases for timber 
products and recreational use increases, roads must be constructed and maintained for 
access to these areas (Appelboom, 2001).  This demand will dramatically increase the 
potential for sediment production (Appelboom, 2001).   
 
6 
 
Soil erosion from forest roads is primarily due to the disturbance of the roadbed.  
Most forest roads have no protection for the soil surface.  According to Brooks (2003), 
the dislodgement of soil particles on the soil surface by energy imparted to the surface by 
falling raindrops is a primary agent of erosion, particularly on soils with sparse vegetation 
cover.  For example, the energy released at the surface during a large storm is sufficient 
to splash greater than 200 tons of soil into the air on a single hectare of bare and loose 
soil (Brooks, 2003).  Once a soil particle has been loosened from its position, the runoff 
tends to flow towards the point of lowest resistance carrying soil particles as it flows.  As 
the runoff begins to collect in small channels, the turbulence of the water causes the 
particles suspended in the water to dislodge additional particles, which increases erosion.  
The erosion from these small channels is characterized as rill erosion.  The channels 
formed by rill erosion concentrate the water into a confined area causing the velocity to 
increase, thereby causing an increase in the turbulence of the flow (Brooks, 2003). This 
increase in turbulent flow increases the rate at which erosion occurs on bare surfaces such 
as ORV trails.  The volume of water moving down the slope also is an important factor 
on the soil erosion.  Moving water, such as water flowing downhill on a road surface, can 
be thought of as a source of kinetic energy. This energy is one of the primary physical 
causes of erosion often seen on forest roads (Brinker, 1995).  The velocity of the moving 
water is a function of the slope on which the road is built. The velocity contributes to the 
amount of kinetic energy exponentially; thus a small increase in slope steepness rapidly 
increases the kinetic energy and erosive capacity of moving water (Brinker, 1995).   
 The use of ORVs causes physical damage to trail soils that increases with each 
vehicle pass over the trail system.  Research on the environmental impact from ORVs has 
 
7 
 
been performed in desert climates, notably because these areas tend to receive a great 
deal of ORV traffic.  Adams et al. (1982) conducted a study in which a comparison of 
erosive potential was made between motorcycles and off-road trucks on soil compaction 
and subsequent vegetative growth in desert environments.  The study found that though 
the truck and the motorcycle both impacted soils, the truck produced the greatest impact.  
Eckert et al (1979) studied the traffic from both motorcycles and trucks on undisturbed 
areas in desert soils and their response to simulated rainfall.  Eckert concluded that the 
truck increased sediment by 3.5 to 20 times compared to undisturbed areas in just 20 
passes, while the motorcycle increased sediment by 2 to 4 times the undisturbed areas, 
depending on soil type.  Kutiel et al. (2000) studied the effects of motorcycles and foot 
traffic on vegetation on sand dunes.  Kutiel?s study showed that just one pass from a 
motorcycle reduced ground cover and plant diversity while it took more than 500 foot 
passes to decrease vegetation diversity. High sediment flux was found on ORV trails in 
Ohio by Sacks and da Luz (2003).  Their results showed that soil loss was associated with 
periods of heavy traffic and reached estimated erosion rates of approximately 2000 kg ha
-
1
yr
-1
.   
2.2 Management Practices 
In 1972 the Clean Water Act (CWA) was passed with the sole purpose to protect 
the chemical, biological, and physical integrity of our nation?s most valuable resource, 
water.  In 1977, the act was amended and the Environmental Protection Agency (EPA) 
was given the authority of permitting, enforcement, and administration of the laws set 
forth by the CWA.  Section 208 of the CWA called for area-wide water pollution control 
 
8 
 
planning in areas designated by the governor of each state that would include both point 
and nonpoint sources and pollution abatement programs (Novotny, 2003).  Section 208 of 
the amendments identified silviculture as a NPS of pollution and required that states set 
guidelines (i.e. BMPs) to reduce NPS pollution.  States were delegated the responsibility 
to define and develop either regulatory or voluntary BMPs for forestry practices 
(Novotny, 2003).  BMPs for forestry practices are recommended for all aspects of forest 
operations such as streamside management zones (SMZ), stream crossings, forest roads, 
timber harvesting, forested wetland management, and reforestation.  BMPs relating to 
forest roads are perhaps the most critical practices to influence the environmental impacts 
of forest operations (Grace, 2005b).  The applications of forest road BMPs to protect 
water quality have become a common practice in forest activities in the United States in 
response to the CWA (Grace, 2005b). BMP use on forest operations, while voluntary, is 
strongly recommended in order to protect our resources and also prevent strict regulation 
on the silvicultural operations.  The BMPs developed in Alabama?s Best Management 
Practices for Forestry Manual include BMPs for Streamside Management Zones (SMZ), 
Stream Crossings, Forest Roads, Timber Harvesting, Reforestation/Stand Management, 
Forested Wetland Management, and Revegetation/Stabilization.  The forest road BMPs 
include the proper planning, adequate drainage, crowned roads, turnout ditches, outsloped 
roads, insloped roads, excessive road steepness, water bars, broad-base dips, outfall 
protection, control of non-essential traffic, and construction and maintenance of 
permanent roads.  The Alabama?s Best Management Practices for Forestry Manual uses 
charts in order to show the recommended distances between structures such as water bars 
and broad-base dips for specific slope percentages (AFC, 1993).  The appropriate spacing 
 
9 
 
between water diversion devices depends on the steepness of the road grade and the 
erosivity of the soil.  The steeper the grade, the greater the velocity of the water body and 
the closer water diversion devices must be placed together (Brinker, 1995).  Brinker 
developed an equation for the calculation of water diversion spacing for forest roads 
applications.     
 
2.3 Prediction of Soil Loss 
Since the early 1950?s, there has been a need to develop a tool to aid forest 
managers in planning or evaluating forest road operations.  In recent years, several 
models have emerged that can be applied in land management.  These models include the 
Universal Soil Loss Equation (USLE), the Modified Universal Soil Loss Equation 
(MUSLE), the Revised Universal Soil Loss Equation version 2 (RUSLE2), the 
WATSHED model and the Water Erosion Prediction Project (WEPP).   
The Universal Soil Loss Equation (USLE), developed in the 1960s, was primarily 
for agricultural use but was updated for use in non-agricultural conditions such as 
construction sites.  USLE predicts long term average annual erosion on field slopes from 
rainfall patterns, soil types, topography, crop system and management practices.  Soil loss 
that might occur from gully, wind, or tillage erosion cannot be predicted due to the fact 
that USLE only predicts the amount of soil loss from sheet or rill erosion (USDA ARS, 
2007).  In order to predict soil loss for a given area, five major factors are used.  The 
factors are (1) rainfall, (2) soil erodibility, (3) slope length and gradient, (4) crop 
vegetation and management practices, and (5) support practices.  These factors were 
 
10 
 
developed from over 11,000 plot-years of data collected in 1930s to 1950s (USDA ARS, 
2007).   
The Modified Universal Soil Loss Equation (MUSLE) is an updated version of 
USLE, which has been updated in order to estimate soil loss from individual storm 
events.  This model is an improvement on USLE but should be used with caution due to 
the fact that it was developed on limited data from Texas and the Southwestern United 
States.   
The Revised Universal Soil Loss Equation Version 2, (RUSLE2), is the latest 
addition of erosion prediction models based on the original USLE.  RUSLE2 estimates 
soil loss, sediment yield, and sediment characteristics from rill and interrill (sheet and 
rill) erosion caused by rainfall and its associated overland flow.  RUSLE2 uses factors 
that represent the effects of climatic erosivity, soil erodibility, topography, cover 
management and support practices to compute erosion (RUSLE2, 2003).  RUSLE2 is not 
a simulation model that attempts to mathematically replicate field processes.  RUSLE2 is 
used to guide conservation planning, inventory erosion rates over large areas, and 
estimate sediment production on upland areas that might become sediment yield in 
watersheds (RUSLE2, 2003).  An advantage to RUSLE2 is that the model has the 
capability to estimate the sediment deposition that occurs on a landscape, compared to 
RUSLE1 and USLE that only estimated the total sediment at the output of the landscape.  
One drawback to the model is that it does not have the capacity to estimate erosion from 
concentrated flows such as gullies or stream channels.  According to the USDA?s 
RUSLE2 User Guide, the four major factors affecting interrill and rill erosion are climate, 
soil, topography, and land use.  The North Carolina Department of Transportation 
 
11 
 
(NCDOT) has introduced the use of RUSLE2 into their erosion and sediment control 
planning.  According to a case study by Ted Sherrod, the NCDOT has used RUSLE2 in 
order to estimate erosion rates on paving projects for secondary roads.  RUSLE2 was 
used to simulate the erosion control structures on the narrow right of ways in order to 
determine the most applicable solution for low budget projects.  Through the use of 
RUSLE2 the NCDOT has been able to demonstrate that throughout most of the state, 
sediment delivery rates are far less than the required 3600 ft
3
/acre/year (Sherrod, 2004).  
Most are in the 35 ? 1550 ft
3
/acre/year range when ditchline grades are less than 2.0% 
(Sherrod, 2004). 
The Water Erosion Prediction Project (WEPP) is possibly the most robust model 
available for road erosion prediction currently available (Grace, 2005).  WEPP, which 
was developed by the USDA, represents a new erosion prediction technology based on 
the fundamentals of stochastic weather generation, infiltration theory, hydrology, soil 
physics, plant science, hydraulics, and erosion mechanics.  The most notable advantages 
include capabilities for estimating spatial and temporal distributions of soil loss, and 
since the model is process-based it can be extrapolated to a broad range of conditions that 
could be more practical or more economical than field tests.  In a small watershed, 
sediment yield from entire fields can be estimated.  Processes considered in hillslope 
profile model applications include rill and interrill erosion, sediment transport and 
deposition, infiltration, soil consolidation, residue and canopy effects on soil detachment 
and infiltration, surface sealing, rill hydraulics, surface runoff, plant growth, residue 
decomposition, percolation, evaporation, transpiration, snow melt, frozen soil effects on 
infiltration and erodibility, climate, tillage effects on soil properties, effects of soil 
 
12 
 
random roughness, and contour effects including potential overtopping of contour ridges 
(NSERL, 1995).  The model also has the ability to model spatial and temporal variability 
in topography, surface roughness, soil properties, crops, and land use conditions on 
hillslopes (NSERL, 1995).  WEPP allows a linkage between hillslope profiles to channels 
and impoundments for watershed applications.  Water and sediment from one or more 
hillslopes can be routed through a small field scale watershed.  Almost all of the 
parameters updating for hillslopes is duplicated for channels.  Channel detachment, 
sediment transport and deposition are all simulated in the model.  Simulations can be 
made for the removal of sediment from the flow from impoundments such as farm ponds, 
terraces, culverts, filter fences and check dams (NSERL, 1995).  
The use of models such as WEPP is becoming a key tool for the planning and 
development of BMPs.  Researchers are now experimenting and calibrating WEPP to 
verify its accuracy for use in prediction of erosion from roads.  From September 1995 to 
December 2003 a study was performed on a forest road located in the Talladega National 
Forest to compare actual erosion and predicted erosion using WEPP.  WEPP predictions 
of sediment yield from cut- and fill slopes with two vegetation treatments and an 
untreated condition were compared to yields observed from the replicated erosion control 
plots over an 8-year period (Grace, 2005).  The model resulted in model efficiencies 
ranging from 0.51 to 0.92 for the cutslope and 0.53 to 0.99 for the fillslope.  These 
relatively high model efficiencies indicate that the model adequately describes sediment 
yields observed in the field experiment (Grace, 2005).  Brodbeck also performed a study 
on ORV trails comparing field data with predicted data through WEPP.  The data 
consisted of total suspended solids (TSS) data from the up- and down-stream sides of a 
 
13 
 
stream crossing which was used to calculate the sediment loading.  The values were 
determined using the Hillslope version of the WEPP model.  The watershed version of 
the WEPP model was determined not to be suitable for modeling erosion from ORV trails 
in an effort to develop BMPs (Brodbeck, unpublished thesis 2005).  According to Ayala 
(2005), the study suggests that ORV trail stream crossings have potential to contribute 
large sediment loads from storm events that have a return interval of 0.5 year or higher.   
In 2001, Klik and Zartl (2001) showed that WEPP simulated storm runoff with 
acceptable accuracy when effective hydraulic conductivity was calibrated with measured 
runoff.  They showed that the WEPP model was able to simulate runoff amounts as well 
as variability between events.  The soil erodibilities for this research had to be multiplied 
by factors of 9-10 for Interrill Erodibility (K
i
) and for 1-3 Rill Erodibility (K
r
) to match 
observed soil losses.   
 
2.4 Summary 
 Literature on the use of erosion prediction models for application on ORV trails in 
the Southeast is very limited.  Currently there are few studies that have been performed 
using WEPP as a simulation method to predict erosion rates for ORV trails.  These 
mainly are associated with stream crossings that occur on the trail systems.  Due to this 
lack of information, research is needed in these areas in order to quantify erosion rates, 
test current management practices, and develop additional management practices for 
ORV applications.  
 
 
14 
 
CHAPTER 3:  RESEARCH PROCEDURES 
3.1 Introduction 
This project was funded by the USDA Forest Service, Southern Research Station, 
Forest Operations Research Unit.  It was a joint effort between the USDA Forest 
Service?s Talladega Ranger District located in Talladega, Alabama, and Auburn 
University?s Department of Biosystems Engineering.  The overall goal of this project was 
to investigate erosion processes on ORV trails of varying gradients and to further develop 
the WEPP erosion model for use as a tool in evaluating trail design and maintenance 
alternatives, in particular for the Kentuck ORV trail system.   
3.2 Site Location 
 The research site was located on the Kentuck ORV trail system, 33? 31? 49? N, 
85? 51?50? W , part of the Talladega National Forest in the Talladega Ranger District 
located in east central Alabama (see Figure 3.1).  The area is generally quite hilly with 
terrain dominated by high ridges separated by steep-sided draws.  The Kentuck ORV trail 
system is a managed recreational use area providing about 48 km of trails motorized 
ATVs, primarily four-wheelers and off-road motorcycles.  The trails are unpaved and 
have an average width of 3 meters.  Current trail maintenance procedures consist 
principally of broad-based dips, water-bars, and wing ditch turnouts.  Water crossings are 
 
15 
 
typically small timber bridges and culverts to reduce in-stream incursions by ATVs.  The 
trail system is divided into four loops, with three longer trails connected by one interior 
loop.  This interior loop, known as the blue trail, is about 3.3 km long (see Figure 3.2).  
Primarily because of accessibility, the experimental plots established for this study were 
located on the blue trail.  This particular trail was also chosen because previous erosion 
work had been carried out on it (Brodbeck, 2007), providing a reference point from 
which to interpret results.  
Figure 3.1 -- Location of research site. 
 
 
 
 
 
16 
 
Figure 3.2 -- Trail map. 
 
 
 A survey of the blue trail in the vicinity of the previous study revealed some 
limitations on the potential range of trail gradients that could be studied.  Any plot chosen 
for this study had to meet several criteria for acceptability, including being of sufficient 
length and relatively straight, have a defined turnout at the bottom for installation of the 
experimental equipment, and not be located such that rider safety would be compromised 
any modifications done to the trail.  Only two sites having the proper characteristics were 
identified, one each with a mild and steep gradient.  The first was on a mild slope of 
approximately 3.5% and 16.2 meters.  The turnout for this site was situated at the bottom 
 
17 
 
of the slope and was approximately 30 feet long and on a 7% gradient.  The second site 
was located on a steeper section of trail with slope of about 20% and length of 36 meters.  
The turnout for the steeper plot was approximately 35 feet long on a 4.5% slope.  For 
both plots, a waterbar was used to divert flow off the trail into the turnout. 
Prior to installation of the experimental plots, the trail sections were shaped using 
a small bulldozer (see Figure 3.3) to a condition that was considered to be optimal 
according to Forest Service guidelines.  The reshaping eliminated ruts that would have 
channeled flow down the trail, and ensure that waterbars and turnouts would function 
correctly. 
Figure 3.3 -- US Forest Service bull dozer performing maintenance. 
 
 
18 
 
3.3 Sampling Description 
Both plots were instrumented to measure runoff and sediment production.  The 
equipment used to measure both variables was installed at the end of the turnouts 
described above.  The basic procedure used was to measure complete runoff volume and 
divert a fixed, small percentages of the runoff for later analysis of sediment 
concentration. 
3.3.1 Sediment Sampling 
 Runoff volume on each plot was measured using a tipping bucket sampler (Khan 
and Ong, 1997), a device that alternately accumulates a fixed quantity of water on each 
side of a collection tub.  As one side becomes filled, the tub tips on a rocker such that the 
collected sample is dumped into a channel and the other side of the tub begins collection.  
The particular tipping bucket used was designed especially for erosion sampling, 
principally by sizing the tub to give accurate measurements over the range of expected 
runoff volume.  The tipping bucket was designed for storms ranging from 0.058 to 110 
millimeters per hour and soil losses of 8.08 to 89.32 metric tons per hectare.  The tipping 
rate of the bucket was not too slow to cause sedimentation between tipping cycles, but 
was also not so fast to change volume between tips.  The samplers were constructed in 
the Biosystems Engineering Department mechanical shop.   
 The tipping volume of the samplers was calibrated before field-testing.  The 
calibration procedure involved using a hose to direct water onto the sampling bucket and 
counting the number of tips required to accumulate a know quantity of water typically 10 
liters.  This experiment was repeated until the standard error of the mean of the estimated 
 
19 
 
collector volume was less than 0.005 liters.   
As runoff occurred during a storm, the flow entered the sampler, causing it to tip 
once the required volume was reached.  As the sampler tipped, the water was poured over 
a collection pipe with multiple holes to allow a portion of the water to flow through.  The 
sample collected in these holes was then routed into a collection bucket through a series 
of pipes.  The number and diameter of holes in the pipe was fixed to divert about 1% of 
total runoff into the collection bucket.  Figure 3.4 is a picture of the sampler 
configuration as installed on the milder trail section.  Of note in the photo is the sheet-
metal apron used to collect water from the turnout ditch into the tipping bucket.   
 For this type of sampler, measurement of runoff volume was equivalent to 
counting the number of tips.  A HOBO
?
 Pendant Event Logger was used in this study to 
record the time, date and number of tips during field data collection.  The event logger 
counts were triggered using a mechanical switch that was attached to the tipping bucket.  
The Pendant
?
 Event Logger was also used to record ambient temperature at a 
predetermined time interval.   
 The water collected in the sampling buckets after each storm was emptied into 
sealable buckets for transport back to campus.  Data from the event logger were 
downloaded using a device from HOBO
?
 designed for that purpose, and the loggers and 
samplers were reset for the next storm.  Once back on campus, the tip data was loaded 
into a custom spreadsheet and a hydrograph was generated.  Sediment samples were 
placed in the Biosystems Engineering Wet Lab cooler for later analysis. 
 
 
 
20 
 
Figure 3.4 -- Tipping bucket sampler during installation on the mild 
slope.
 
 
3.3.2 Rainfall Data 
 Rainfall data was collected using an ISCO
?
 674 Tipping Bucket rain gauge 
located approximately 300 meters from the experimental plots.  The ISCO
?
 674 rain 
gauge measured both the intensity and the duration of each rainfall event.  The rainfall 
data was stored directly on an ISCO
?
 6700 water sampler and was downloaded after each 
storm event using hardware from the manufacturer.  
3.3.3 Sediment Analysis 
 Total sediment from each plot was estimated from the runoff samples collected 
 
21 
 
using the above equipment.  The entire sample was poured into an oven-safe container, 
weighed, and placed into an oven held at a constant 103?C.  The containers were weighed 
every 24 hours and removed when two consecutive weights were found to be the same.  
Tare weight of the container was subtracted and the difference was the total weight of 
sediment in the sample.  The total sediment loading was then calculated using the 
equation. 
S =
A? B
C
 
where: 
 S = total sediment from the experimental plot, kg 
 A = dry weight of sample and pan, kg 
 B = tare weight of the pan, kg 
 C = sample percentage of total runoff, = 1%. 
 The results from the lab analysis were compiled in combination with the 
information from the HOBO pendant data logger in order to calculate the sediment 
loading for each storm event for both research sites. 
 
3.4 Erosion Prediction Modeling with WEPP 
 The erosion modeling for this project was performed with the USDA?s Water 
Erosion Prediction Project (WEPP).  Two of WEPP?s main interfaces options are the 
hillslope and watershed interface.  They differ mainly in the extent of the area modeled.  
For the purpose of this project, the hillslope interface provided the ability to model the 
 
22 
 
details of a single slope section, much like the trail itself.  The watershed interface is 
normally used to model large areas with less specific soil and management information.   
The hillslope interface uses file builder programs to generate and format the 
significant amount of input data that are required to run the WEPP model.  The builder 
programs allow the user to specify management practices and input data for climate, 
slope, soil, cover management, and, if needed, irrigation.  The hillslope interface also 
provides a user the option of simulating a single storm or an extended period of time 
using computer-generated weather inputs.   
 
3.4.1 Weather Generator 
 WEPP uses a weather generator to drive long-term simulations that calculate 
average annual sediment yield.  This weather generator program is known as CLIGEN v. 
4.3 and requires the user to specify a weather station location from which realistic rainfall 
information can be acquired.  The Anniston, Alabama rain gauge station was chosen for 
the long-term simulation due to its proximity to the research site (located approximately 
5.5 km from the trail). 
3.4.2 Hillslope Designation and Model Calibration 
 Hillslopes in WEPP are defined by many parameters collected into three 
functional groups, known as layers.  The first layer, defined as a management layer, 
allows the user to establish specific management practices for a given hillslope. 
Management information was available in the database for agricultural and forest road 
applications.  The specialized forest road applications management database was used for 
the modeling performed in this study.  The specific management practice chosen was an 
 
23 
 
insloped, unrutted forest road with a bare ditch, which was felt to most closely 
approximate how the ORV trails were managed.  
The second input layer defined the trail extent and profile.  This information was 
measured using a clinometer and meter tape for each trail and turnout section.  A 
hillslope definition for input to WEPP consisted of data describing the specific trail and 
turnout as contiguous sections of a single slope. A graphical representation of the two 
hillslopes defined for this study is shown in figures 3.5 and 3.6. 
 
Figure 3.5 -- Mild slope with insloped unrutted forest road-bare ditch management layer. 
 
 
 
 
 
 
 
24 
 
Figure 3.6 -- Steep slope with insloped unrutted forest road-bare ditch management layer. 
 
The third input layer was used to describe soil information for the specific 
location and site conditions.  Through literature review on WEPP it was determined that 
the most sensitive soil parameters for WEPP are:  the interrill erodibility, rill erodibility, 
critical shear, and effective hydraulic conductivity (Brodbeck, 2005).  These parameters 
have a physical interpretation and can be measured, but are difficult to determine 
experimentally. Moreover, the relationship between these parameters as defined 
physically and as used in the WEPP simulation is not always one and the same. For 
example, the hydraulic conductivity parameter is defined physically as a measure of how 
quickly water moves through a given soil under the force of gravity and is indicative of 
the its relative porosity.  As defined in WEPP, however, the effective hydraulic 
conductivity is related to the actual hydraulic conductivity, though not equal. 
WEPP provides an alternative to physically measuring the parameters important 
to the simulations in which reasonable estimates can be calculated using a set of 
 
25 
 
equations. The resulting estimates are appropriate for how WEPP itself interprets them. 
The equations use as input soil physical properties that are relatively easy to measure, 
including textural characteristics, organic matter content, and cation exchange capacity. 
The equations for critical soil parameters used by WEPP are: 
K
i
= 2,728,000+192,100S
vf
K
r
= 0.00197+ 0.0003S
vf
+ 0.03863e
?1.84 O
K
b
=?0.265+ 0.0086S
1.8
+11.46E
C
?0.75
T
c
= 2.67+ 0.65C ?0.058S
vf
   [3.2]
 
where K
i
 is interrill erodibility (kg s m
-4
) which is a measure of sediment delivery rate to 
rills, S
vf
 is percent very fine sand in the soil, K
r
 is rill erodibility (s m
-1
) and a measure of 
soil susceptibility to detachment by concentrated flow, O percent organic matter in the 
soil, K
b
 is effective hydraulic conductivity (mm hr
-1
), S is percent sand in soil, E
C
 is soil 
cation exchange capacity (meq 100g
-1
), T
c
 is soil critical shear (N m
-2
), and C is percent 
clay in the soil.   
The interrill erodibility, rill erodibility, critical shear, and effective hydraulic 
conductivity calculated were used as initial input parameters in the WEPP soil component 
layer. Soils were considered uniform throughout the hillslope, the same whether on the 
trail or on turnout.  This was, by observation, not the case in fact. The turnouts tended to 
have a layer of loose deposited sandy material on the surface, whereas the trails tended to 
be uniformly densely compacted with little to no loose material on the surface. Soils 
beneath the surface of the turnouts, however, were compacted to a similar level as the 
trails.  It was decided, however, that using two soil types in the WEPP soil layer 
definition (one each for trails and turnouts) would make the calibration process more 
 
26 
 
difficult. Calibrating WEPP to some ?average? level of soil parameters was felt to be 
more appropriate than introducing an additional level of complexity when tuning the 
model. 
With the WEPP input layers initialized, multiple event-by-event simulations were 
conducted and the results compared to measured values of runoff and sediment 
production.  Soil parameter definitions were then altered in what was felt an appropriate 
manner and the simulations run again. This process was repeated until no further 
improvements were seen in the correlation between simulated and measured runoff and 
sediment. The measure of correlation used between simulated and measured variables 
was the Nash-Sutcliffe efficiency. When this value reached a maximum, the model was 
considered calibrated. 
There were two alternatives when calibrating the WEPP model: a) step-wise 
calibration of runoff followed by sediment and b) ignoring runoff and calibrate directly to 
sediment. Both alternatives were attempted, but the later of the two methods tended to 
produce higher Nash-Sutcliffe correlations.   
Because of the limited storm data available, it was not possible to divide the 
available data into unique calibration and validation sets. As an alternative, data were 
divided by slope and calibration was performed on one class (mild or steep) and 
validation was performed using the other. This process was repeated in both directions, 
i.e. the calibration and validation roles were reversed. 
The storm data collected occurred over a period of time in which the trail was 
both open to traffic, and closed. This fact allowed an additional comparison of model and 
measured results between trafficked and non-trafficked conditions.  
 
27 
 
 
3.4.3 Statistical Analysis  
 In order to analyze for correlation between the actual field data and the predicted 
data from the model, the Nash-Sutcliffe efficiency (NSE) was used for the hydrology 
modeling and R
2
 was used for the sediment loading.   
According to Moriasi et al. (2007) the Nash-Sutcliffe efficiency is a normalized 
statistic that determines the relative magnitude of the residual variance (?noise?) 
compared to the measured data variance (?information?).  The NSE evaluates how field 
data compared to the predicted data fits a 1:1 line.  The equation for the NSE is as 
follows: 
NSE =1?
Q
m
?Q
p
()
2
?
?
Q
m
?Q
a
()
2
      [3.3] 
Where: 
 Q
m
 = the measured sediment or runoff, 
 Q
p
 = the model-predicted sediment or runoff, 
 Q
a
 = the average measured sediment or runoff. 
The range for the NSE values are between -? and 1, with 1 being the best fit.  The model 
is considered to have acceptable performance if the NSE is between 0.0 and 1.0.   While 
values greater than 0 are acceptable, values 0.0 to 0.5 are consider poor with 0.5 to .75 
are moderate and .75 to 1.0 being good.  Negative values are usually considered 
unacceptable performance. According to Moriasi et. al. (2007), the Nash-Sutcliffe is 
recommended by the American Society of Civil Engineers (ASCE) for hydrologic 
modeling. 
 
28 
 
 The R
2
 value is an index of the degree of linear relationship between observed and 
simulated data (Moriasi et al., 2007).  The R
2
 value ranges from 0 to 1, with zero having 
no linear relationship.  The R
2
 value represents the proportion of variance with the higher 
the number the less variance, therefore values greater than 0.5 are usually considered 
acceptable for model calibration (Moriasi et al., 2007).   
 
3.4.4 Model Validation 
 Model validation consists of running simulations with the calibrated model for 
storm events where field data has been collected.  The predicted values from the 
calibrated model could then be compared to those from the actual field data collected.  
The validity of the calibrated model was judged by comparing the NSE of the validation 
set of data to that of the calibration set.  If the NSE for the validated model was similar to 
that of the calibrated model, the validation was considered acceptable.  
 
29 
 
CHAPTER 4:  RESULTS 
4.1 Runoff Sampling 
 Runoff sampling was performed for approximately nine months, from October 
2006 through June 2007.  During that time period, eight storm events occurred on the 
trail sites.  Due to a malfunction in the sampling equipment, complete data was recorded 
for five storms.  For the storms with complete data sets, the cumulative rainfall, rainfall 
intensity, rainfall duration, total runoff, and total sediment yield were recorded.  Data are 
summarized in Table 4.1 for rainfall and Table 4.2 for runoff and sediment loading. 
The storm producing the largest sediment loading occurred on December 22, 
2006.  Rainfall for the event totaled 30.7 centimeters over 6.75 hours with a maximum 
intensity of approximately 0.71 centimeters per hour.  The sampling system recorded a 
volume of 2190 liters from the mild slope and 2670 liters of runoff from the steep slope 
(Figure 4.1).  Total sediment loss resulting from this storm for the mild and steep slopes 
was 0.71 tonnes per hectare and 2.5 tonnes per hectare, respectively. 
 
 
 
 
 
 
30 
 
Table 4.1 -- Rainfall data. 
Storm 
Date 
Max 
Rainfall 
Intensity 
(cm/hr) 
Rainfall 
Duration 
(hr) 
Total 
Cumulative 
Rainfall 
(cm) 
11/15/2006 1.22 9.5 5.79
12/22/2006 0.71 6.75 3.07
2/1/2007 0.91 3.75 2.67
3/1/2007 1.42 0.5 0.36
3/15/2007 2.82 4.25 5.08
 
The storm on December 12, 2006 produced the largest sediment loading, although it did 
not have the highest runoff volume or largest rainfall amount during the sample period.  
The storm that had the highest runoff volume occurred on November 15, 2006 with the 
mild slope producing 6110 liters, and the steep slope 7750 liters of runoff.  This storm 
also produced the highest cumulative rainfall during the study period.  The highest 
rainfall intensity occurred in March with a maximum intensity reaching 2.82 centimeters 
per hour and 5.08 centimeters of cumulative rainfall.  For data on all storm events, refer 
to Appendix A. 
 
 
 
 
 
 
 
 
31 
 
Table 4.2 -- Sediment yield. 
Storm Date Plot 
Total 
Runoff (L) 
Total 
Sediment 
Loading (Kg) 
Total 
Sediment 
Loading 
(T/Ha) 
11/15/2006 
Mild 6112 11.01 1.76 
  
Steep 7746 53.17 4.62 
12/22/2006 
Mild 2186 28.81 4.61 
  
Steep 2672 101.17 8.78 
2/1/2007 
Mild 2551 5.31 0.85 
  
Steep 2180 11.52 1.00 
3/1/2007 
Mild 567 1.7 0.27 
  
Steep 1908 7.91 0.69 
3/15/2007 
Mild 2095 16.57 2.65 
  
Steep 5489 45.32 3.93 
 
Figure 4.1 -- Cumulative runoff for 12/22/2006. 
Cumulative Runoff 12/22/06
0
500
1000
1500
2000
2500
3000
12/21/2006
21:36
12/22/2006 0:00 12/22/2006 2:24 12/22/2006 4:48 12/22/2006 7:12 12/22/2006 9:36 12/22/2006
12:00
Time
Li
te
r
s
 p
e
r
 m
i
nute
Steep Slope
Mild Slope
 
 
32 
 
 
4.2 WEPP Model Calibration and Validation 
4.2.1 Runoff Calibration 
 The WEPP model requires several parameters be provided, the most important of 
which are interrill erodibility, rill erobility, critical shear, and effective hydraulic 
conductivity.  These parameters were estimated from soil samples taken from the trail 
running surface and analyzed at the Auburn Soil Testing Laboratory.  The soil analysis 
indicated trail soils were a silty clay loam composed of 23% clay, 19% sand, 59% silt, 
and 0% organic matter, refer to Table 4.3.  Using the equations discussed in the research 
procedures, the required WEPP input parameters were calculated and are listed in Table 
4.4. 
Table 4.3 -- Soil test results 
Soil Parameter Value 
% Clay 23 
%Silt 59 
% Sand 19 
% Very Fine Sands 27 
% Rock 0 
% Organic Matter 0 
Cation Exchange Capacity 12 
 
 
 
 
 
 
33 
 
Table 4.4 -- Calculated soil parameter values. 
Parameter Value 
Interrill Erodibility 
(kg*s/m4) 
7.91E+06 
Rill Erodibility 
(s/m) 
0.010 
Critical Shear 
(N/m2) 
2.57 
Effective Hydraulic 
Conductivity 
(mm/h) 
3.19 
  
The calculated values were input into WEPP along with the slope lengths, 
steepness, and observed weather information.  Simulations were performed and compared 
to the actual data collected from the plots.  The simulation with the calculated soil 
parameters values (Table 4.4) resulted in no predicted runoff or sediment yield.  This was 
a significant departure from the observed sediment production behavior of the trails and 
was felt to be most likely a result of the high estimated effective hydraulic conductivity 
(EHC), which represented an undisturbed soil condition.  The trail receives a high 
volume of ORV traffic and its surface has become densely compacted over time. The 
calculated hydraulic conductivity value for an undisturbed soil might, therefore, not 
apply.    In addition to the traffic effect, the trail has been repeatedly reshaped using a 
bulldozer during maintenance intervals. As a consequence, the trail surface has been cut 
into the landscape to a depth of about 2 feet on average, or deeply into the subsurface soil 
horizon. This layer was a dense clayey material having low natural permeability. 
The EHC value was therefore systematically changed first in order to achieve the 
maximum Nash Sutcliffe Efficiency (NSE), refer to Appendix C for values tested and 
 
34 
 
sensitivity plots.  A maximum in NSE occurred at EHC of 0.13 mm hr
-1
 with the other 
parameters held fixed at their calculated values (from Eq 3.2) - see Figure 4.2.  The 
maximum NSE was 0.45, a value considered acceptable for simulations, but not above 
0.5, the threshold generally accepted as representing a ?good fitting? model.  
 
Figure 4.2 -- Effective hydraulic conductivity calibration for runoff. 
Effective Hydraulic Conductivity Calibration
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0 0.1 0.2 0.3 0.4 0.5 0.6
K
b
NS
E
 
 
Having fixed an EHC value, the WEPP model was run for the measured storms 
using calculated values (Eq 3.2) for the erodibility and critical shear factors. Figure 4.3 is 
a graph showing measured and predicted runoff for the model as specified using these 
parameters along with a 1:1 line as a reference. The model consistently under predicted 
runoff for the conditions tested. 
 
 
 
35 
 
Figure 4.3 -- Actual runoff compared to predicted runoff after calibration using WEPP. 
Actual Runoff vs Predicted Runoff
y = 0.6677x - 22.562
R
2
 = 0.7367
NSE = 0.45
y = x
0
1000
2000
3000
4000
5000
6000
7000
0 1000 2000 3000 4000 5000 6000 7000 8000 9000
actual (Liter)
Predicted 
(Liters)
 
 
4.2.2 Sediment Calibration 
The gross calibration of the model with measured runoff was achieved through 
changes in hydraulic conductivity.  Further adjustments were subsequently made to the 
other parameters (interrill erodibility, rill erodibility, and critical shear) to fine-tune the 
predictions of sediment yield.  Once again, the objective in changing the parameters was 
to produce a better linear relationship between the predicted and measured data.  The 
values of these parameters were adjusted and simulated in WEPP to produce the highest 
coefficient of determination (R
2
), see Appendix C for all values and related R
2
 values.  
The R
2
 measure was used to assess quality of the predictions for sediment since NSE 
values were typically below 0, and therefore unacceptable. Parameter values were fixed 
by first changing critical shear until the maximal R
2
 was observed. The process was 
repeated for the two erodibility factors. This sequential approach to WEPP model 
Predicted vs. Measured ? 1:1 
NSE = 0.45 
2
 
36 
 
calibration seemed to work very well, but was not guaranteed to produce the globally 
maximal NSE or R
2
. Final parameter values are summarized in Table 4.5. 
 
Table 4.5 ? Parameter values used to calibrate WEPP. 
Parameter Calibrated Value 
Interrill Erodibility 
(kg*s/m4) 
3.00E+06 
Rill Erodibility 
(s/m) 
0.0005 
Critical Shear 
(N/m2) 
12.5 
Effective Hydraulic 
Conductivity 
(mm/h) 
0.130 
 
The final regression of the predicted versus actual sediment production is shown 
in Figure 4.4.  The model had a slope of approximately 2 to 1, consistently over 
predicting sediment, with R
2
 of 0.43, and NSE of -0.36.  In general, the calibrated WEPP 
model therefore predicted lower runoff, but much higher sediment production, than 
observed. This implied that the model estimated much higher concentrations of sediment 
in a smaller volume of runoff. 
 
 
 
 
 
 
 
 
37 
 
Figure 4.4 -- Actual sediment yield compared to predicted sediment yield after calibration 
using WEPP. 
Actual vs Predicted
y = 1.9455x + 25.11
R
2
 = 0.4266
NSE = -0.36
y = x
0
50
100
150
200
250
300
0 20 40 60 80 100 120
actual
predicted
 
 
4.2.3 Sediment ? Only Calibration 
The original model was calibrated to runoff and sediment production, but the 
calibrated model performance was not close to the observed erosion behavior. As an 
alternate approach, a calibration directly to sediment, bypassing runoff, was attempted. 
The same calibration process was performed as in the previous approach.  For the 
recalibration of the model, the NSE was used to determine the goodness of fit since this 
statistic represents a 1:1 slope instead of a standard linear relationship.  The soil 
parameters were again adjusted and simulated in order to determine values producing the 
best model fit, see Appendix D for detailed parameter and NSE values.  Table 4.6, shows 
the final values obtained from this calibration.  The final model had NSE of 0.24, still 
less than 0.5 but much higher than the ?0.36 for the model calibrated first to runoff. 
 
38 
 
 
Table 4.6 -- Model calibrated for sediment only 
Recalibration of Sediment using the NSE for 
goodness of fit 
Parameters Value 
Interrill Erodibility 
(kg*s/m
4
) 
3.30E+06 
Rill Erodibility 
(s/m) 
0.0001 
Critical Shear 
(N/m
2
) 
0.006 
Effective Hydraulic 
Conductivity 
(mm/h) 
0.08 
 
 
Figure 4.5 ? Model calibrated for sediment, without previous calibration for runoff. 
Model Calibrated for Sediment
y = 0.5208x + 9.904
R
2
 = 0.3418
NSE = 0.24
y = x
0
10
20
30
40
50
60
70
80
90
0 20 40 60 80 100 120
Measured Data
Predicted
 
 
39 
 
 
4.2.4 WEPP Model Validation 
 The measured data were divided into mild and steep slope categories to perform a 
limited validation of the calibrated model. The process was to alternately calibrate the 
model on one slope class and validate on the other. Given the calibrated model could be 
confirmed as valid, the model itself could be used to evaluate the effect of slope on 
erosion rates. 
 A similar procedure as in the above analysis was used to calibrate the model 
independently for the mild and steep slope data sets. Refer to Appendices E and F for a 
summary of the parameter values used and resulting NSE for the mild and steep slope 
calibrations, respectively. The mild slope calibration produced a model with maximum 
NSE of -0.18 with the parameter values in Table 4.7. Using the mild slope parameters to 
estimate sediment production on the steep slope resulted in NSE of -1.12, see Figure 4.6. 
The calibration on the steep slope had NSE of 0.11, somewhat higher than for the mild 
slope. Table 4.8 contains final parameter estimates for the steep slope calibration. The 
comparison of measured and calibrated values for the mild slope resulted in a NSE of -
1.77, refer to Figure 4.7.  In both cases, the calibrated model was not as good a predictor 
of expected sediment production for the other slope. 
 
 
 
 
 
 
40 
 
Table 4.7 -- Mild slope calibration 
Calibration of Mild Slope 
Parameters Value 
Interrill Erodibility 
(kg*s/m
4
) 
5.20E+06 
Rill Erodibility 
(s/m) 
0.00016 
Critical Shear 
(N/m
2
) 
0.01 
Effective Hydraulic 
Conductivity 
(mm/h) 
0.09 
 
Figure 4.6 ? Validation using model parameters estimated from mild slope data for the 
steep slope. 
Mild Slope Calibration Validated using Steep Slope Data
y = 0.4602x + 45.742
R
2
 = 0.1313
NSE = -1.12
y = x
0
20
40
60
80
100
120
140
2040608010120
Measured
Predicted
 
 
 
NSE = -1.12 
 
41 
 
Table 4.8 -- Steep slope calibration 
Calibration of Steep Slope 
Parameters Value 
Interrill Erodibility 
(kg*s/m
4
) 
1.10E+06 
Rill Erodibility 
(s/m) 
0.00012 
Critical Shear 
(N/m
2
) 
0.009 
Effective Hydraulic 
Conductivity 
(mm/h) 
0.085 
 
Figure 4.7 ? Validation using model parameters estimated from the steep slope data for 
the mild slope. 
Steep Slope Calibration Validated using Mild Slope Data
y = 0.0001x + 0.0032
R
2
 = 0.1689
NSE = -1.77
0
0.001
0.002
0.003
0.004
0.005
0.006
0.007
0.008
0.009
0.01
0 5 10 15 20 25 30 35
Measured
Predicted
 
 
 
NSE = -1.77 
 
42 
 
Figure 4.8 ? Traffic vs. Non-Traffic period. 
Traffic vs. Non-Traffic
0
10
20
30
40
50
60
70
80
2040608010120
Measured
Predicted
Traffic
Non-Traffic
 
4.2.5 Effect of Traffic 
 Figure 4.8 is a plot of predicted and measured sediment from the combined 
calibration for sediment only with the points classified by time period. During the periods 
October through December, 2006 and April through June, 2007 the trail system was open 
for use. From January to April, 2007 the trail was closed and no traffic was permitted. It 
was clear from the data that the model sediment predictions were closer to measured 
values for the non-trafficked period. For the storms during the trafficked period, the 
WEPP model significantly under predicted the amount of sediment relative to the 
observed. Two calibrations of the WEPP model were therefore made to determine the 
effect of the traffic. 
The calibration for the traffic period resulted in NSE of -0.11 while the non-
trafficked period produced a NSE value of 0.76; see Appendix G and H for parameter and 
NSE values.  The calibrated values for each of the processes are shown in Tables 4.8 and 
 
43 
 
4.9. Refer to Figures 4.9 and 4.10 for plots of measured and predicted sediment data for 
trafficked and non-trafficked conditions, respectively. The results from the calibration 
process for the two models indicated WEPP was more accurate for the non-trafficked 
time period.  The erosion from the non-trafficked period resulted from weather 
occurrences only.  The WEPP model was designed to simulate soil erosion from 
precipitation only and does not account for traffic on the ORV trails. Anthropogenic 
factors are much less predictable and have been researched to a much lesser extent than 
precipitation. From this research (Tables 4.8 and 4.9), it was clear that traffic increased 
WEPP erosion parameters by a factor of about 2. It was also clear that simply changing 
the model input parameters did not account fully for the observed changes in erosion, as 
was seen in the much lower NSE for the trafficked condition. Some other type of erosion 
mechanism must not be accounted for in WEPP, probably related to soil loosening from 
the action of the tires on the trail surface. Further research will be needed to both confirm 
this result and to propose a means of accounting for this traffic factor in the model itself. 
 
 
 
  
 
 
 
 
 
Table 4.9 -- Calibrated values 
for trafficked period 
 
Parameters Value 
Interrill 
Erodibility 
(kg*s/m
4
) 
4.26E+06 
Rill Erodibility 
(s/m) 
0.00012 
Critical Shear 
(N/m
2
) 
0.008 
Effective 
Hydraulic 
Conductivity 
(mm/h) 
0.071 
Table 4.10 -- Calibrated 
values for non-trafficked 
period 
Parameters Value 
Interrill 
Erodibility 
(kg*s/m
4
) 
4.10E+05 
Rill Erodibility 
(s/m) 
9.00E-05 
Critical Shear 
(N/m
2
) 
0.0058 
Effective 
Hydraulic 
Conductivity 
(mm/h) 
0.152 
 
44 
 
Figure 4.9 ? Calibrated data for trafficked period 
Measured vs. Predicted -- Traffic Period
y = 0.4313x + 17.188
R
2
 = 0.2114
NSE = -0.11
y = x
0
20
40
60
80
100
120
2040608010120
Measured
Predicted
 
Figure 4.10 ? Calibration of non-trafficked period. 
Measured vs. Predicted -- Non-Traffic Period
y = 0.8744x - 1.9628
R
2
 = 0.83
NSE = 0.76
y = x
0
5
10
15
20
25
30
35
40
45
50
1020304050
Measured
Predicted
 
 
The results of the six calibration processes (refer to table 4.11) showed that the 
non-trafficked data was the most accurately predicted from the model simulations while 
the trafficked period produced an unacceptable goodness of fit (NSE) value. Apparently, 
NSE = -0.11 
NSE = 0.76 
 
45 
 
the runoff model of WEPP was also less accurate than the sedimentation prediction 
portion since bypassing the runoff calibration step increased the accuracy of predictions, 
although neither one was accurate enough to call it a ?good? fit. The effect of slope was 
also not large. Calibrated model parameters were similar between the mild and steep 
slopes, differing by greater than 25% only in the case of the inter-rill erodibility, which 
went down quite a bit between the mild and steep slope conditions. The drop in inter-rill 
erodibility could be associated with the type of soil.  The steep slope was more deeply 
dug into the landscape, into a different soil horizon than the mild slope. In general, the 
WEPP model performed fairly well in predicting erosion, but further research is needed if 
high correlation with measured erosion is desired. Most important will be developing a 
means of accounting for surface soil disturbance, principally wheel traffic, into the 
model. 
Table 4.11 ? Summary of calibration data. 
Calibration 
Slope 
used for 
Calibration 
Calibrated 
NSE 
Interrill 
Erodibility
Rill 
Erodibility
Critical 
Shear 
Effective 
Hydraulic 
Conductivity 
Slope 
used for 
Validation
Validated 
NSE 
      
Ki Kr Tc Kb 
    
Original 
Calibration 
Both -0.36 3.00E+06 0.00050 12.5 0.130 NA NA 
Calibration 
for 
Sediment 
only 
Both 0.24 3.30E+06 0.00010 0.0060 0.080 NA NA 
Mild Slope Mild -0.19 5.20E+06 0.00016 0.0100 0.090 Steep -1.12 
Steep 
Slope 
Steep 0.10 1.10E+06 0.00012 0.0090 0.085 Mild -1.77 
Trafficked Both -0.11 4.26E+06 0.00012 0.0080 0.071 NA NA 
Non-
Trafficked 
Both 0.76 4.10E+05 0.00009 0.0058 0.152 NA NA 
 
46 
 
4.3 BMP Simulation 
Current practice is to use BMPs developed for forest roads for ORV trails as well, 
primarily because there has been little research directed toward developing BMPs for 
ORV trails. The WEPP model calibrated for this project offered an opportunity to 
evaluate WEPP for the use of developing management alternatives for trails.  The WEPP 
model was used to simulate trail erosion for a 30-year time period in order to estimate 
average annual sediment loss.  The simulations were performed on multiple slopes to 
determine the maximum distance between water diversion structures.  In all simulations, 
it was assumed that all water was diverted off the trail at each water diversion structure. 
In order to run the simulations, WEPP requires a maximum sediment loading 
value.  According to Ayala (2005), the National Forest in Alabama considers more than 
44.5 tons/ha/year (18 tons/acre/year) of sediment yield from temporary roads and 
disturbed areas as unacceptable and this value was used in all simulations. 
Given the maximum allowable sediment export level, simulations were run at a 
given slope to calculate the length of trail at which the threshold erosion level would be 
exceeded.  The model calibrated for sediment production only was used in order to 
estimate the recommended water bar spacing.  Table 4.12 lists the maximum slope 
lengths between water diversion structures, produced by WEPP, to ensure compliance 
with the erosion limit. This table shows the results from the WEPP model, these values 
were simulated and are estimates of structure spacing.  Due to the limited data used to 
calibrate and validate the model these results should not be used in practice until the 
model has been validated and the results have been verified.  The spacing was calculated 
 
47 
 
such that the hillslope would not exceed the maximum loading at the point at which water 
leaves the trail and enters the diversion. 
Table 4.12 -- Water diversion structures spacing from WEPP results. 
Recommended Water Diversion Structure Spacing 
Slope ( % ) Spacing ( ft ) 
2 265 
4 100 
6 55 
8 30 
10 18 
12 14 
14 8 
 
 Once the structure results were calculated, we compared these spacing values to 
diversion spacing information established by Brodbeck (2005) at Auburn University, see 
Table 4.12.  Brodbeck suggested spacing for the milder slopes were much less than that 
calculated from our simulations.  Brodbeck recommended 46 feet for a 4% slope while 
our simulations showed that a spacing of 100 feet was appropriate.  Brodbeck suggests a 
22 feet spacing on a 14 % hillslope, whereas our model suggests that a 8 ft maximum 
spacing should be installed.  The recommendations from both models are designed for 
trails that are not in proximity with a stream channel or other water body.  The difference 
in spacing  could be associated with the difference in field collection.  The data collected 
by Brodbeck occurred at a stream crossing in which TSS and stream flow data were 
collected and analyzed.  The data collected for this project consisted of runoff samples 
from two hillslopes with different steepness which was then used to calculate total 
sediment loading.  The different data collection locations could be the reason for the 
differences in the spacing results. 
 
48 
 
Table 4.13 -- Water diversion spacing developed by Brodbeck 
Recommended Diversion Structure Spacing developed by 
Brodbeck 
Slope ( % ) Spacing ( ft ) 
4 46 
6 36 
8 33 
10 29 
12 27 
14 22 
18 19 
 
 In the past, the BMPs used on ORV trails were adopted from forest roads.  We 
compared the calculated water bar spacing from WEPP to the recommended water bar 
spacing for forest roads.  In 1995 Brinker developed tables to assist managers in water 
bar spacing. The spacing recommended for water bar spacing for skid trails and logging 
roads is in Table 4.14 taken from (Brinker, 1995).  This table shows that the 
recommended spacing for a road with a 2 % grade is 250 ft.  This spacing is similar on 
milder slopes, but the slope length  resulting from this research for steeper slopes is much 
shorter.   
Table 4.14 -- Recommended distances between water bars on skid trails and logging 
roads after logging is complete (Brinker, 1995). 
Recommended Distances Between Water Bars On Skid 
Trails and Logging Roads After Logging Is Complete. 
Road Grade Spacing (feet) 
2 250 
5 135 
10 80 
15 60 
20 45 
25 40 
30 35 
40 30 
 
49 
 
CHAPTER 5:  CONCLUSION 
 
 The goal of his project was to study the sediment loading from ORV trails and to 
use WEPP to calculated maximum slope lengths for varying slopes using field data..  
Two sites were studied on the Kentuck ORV trail system for approximately 9 months.  
During this time period the trail received 8 storms, three of which were not sampled due 
to equipment malfunction.  A runoff sample was taken from the remaining five storms in 
order to determine the amount of sediment being eroded from the trail.  In addition to the 
runoff sample, the total runoff, runoff intensity, rainfall, and rainfall intensity were also 
recorded for use in the analysis.  Data analysis was performed to calculate the total 
amount of sediment being eroded from the trail.  This data was used to simulate the 
erosion from the trail system through the Water Erosion Prediction Project model, WEPP.  
The trail sections were simulated and compared to the actual field data collected from the 
site.  Using this process, WEPP was calibrated in order to increase the accuracy of the 
simulations.  With the model calibrated, it was used to develop BMPs for the trail system.   
 
5.1 Objective 1 
 The 5 storms sampled produced approximately 169 millimeters of rainfall, in a 
region that typically receives approximately 1300 millimeters per year.  This produced a 
 
50 
 
total of 63.4 kg and 219.09 kg of sediment from the mild and steep slopes, respectfully.  
Due to unusual weather conditions the location only received one storm event with one 
inch cumulative rainfall.  During this time, a storm occurred producing 3.07 centimeters 
of rainfall on December 22, 2006.  The storm peaked at 0.71 centimeters per hour with 
duration of 6.75 hours.   This storm caused approximately 28.81 kg and 101.17 kg from 
the mild and steep slope, respectively.   
 
5.2 Objective 2 
 Using the data collected from the storms, the WEPP Hillslope Interface was 
calibrated to predict sediment loading.  Using information taken from soil samples on the 
trails, required soil parameters were determine and input into the WEPP software.  Using 
the Nash-Sutcliffe efficiency, the field data were compared to predicted data from the 
model.  WEPP was calibrated for multiple scenarios in order to see the effects of slope 
and traffic on the trail.  The model that was calibrated for the non-trafficked period 
produced the highest NSE of 0.76.  This model was expected to produce the highest 
values since it only accounted for the weather related erosion on the trail and not the 
added variable of ATV traffic.  The calibration for sediment production which included 
both slopes produced the second highest NSE of 0.24.  Due to the lack of field data, this 
was the best calibration that could be obtained for ORV use on the trail 
5.3 Objective 3 
 The calibrated WEPP model was used calculate slope lengths for ORV trails.  
This was needed because of the lack of data available on ORV trails.  The results 
 
51 
 
indicated that the maximum spacing for water diversion structures is approximately the 
same as was published by Brinker (1995) for mild slopes.  Simulations showed that while 
on mild slopes, the results are approximately the same as forest roads while spacing on 
steep slopes are much less than on forest roads.  The spacing resulting from the data 
collected were used for the purpose for modeling varying slopes.  The calibration model 
did not produce an acceptable model to accurately predict the max loading.  Therefore, 
these slope lengths should not be used in practice until more research can conform there 
accuracy. 
5.4 General Conclusion 
 It is concluded that ORV trail systems such as the Kentuck ORV area can be a 
serious threat to the water resources on our forest landscapes.  Through the use of 
specialized BMPs, along with additional research and BMP development the threat from 
sediment pollution from ORV trails can be lowered.  This is essential for the protection of 
our forested lands and the recreational use of those lands. 
 
5.5 Future Research 
 In the future, additional research should be performed on the ORV trail to further 
reduce the environment impacts of the trail systems.   
? Research of erosion from the use of ORVs would add significant information to 
the research performed in this study.  The type of ORV, type of tires, and the 
aggressiveness of the driver all have an effect on the sediment production.  Due to 
 
52 
 
these factors, an in-depth study should be performed in order to determine the 
type and speed of the ORVs operating on the trails.  This could be done through 
motion sensing cameras at each location.  These cameras would show the exact 
number of riders and speed at which the riders crossed through the study area.  A 
controlled study of soil damage from ORVs would also generate information that 
could be used in this study.   
? Research on the erosion from the turnouts located on the trail could be used in the 
model from this project to increase the model?s efficiency.  For this project, the 
soil was considered uniform throughout the slope.  The soil in the trail and turnout 
are of similar type, but the compaction is much lower in the turnout.  Sediment 
loading from the turnout could be measured and calibrated in WEPP, then used in 
the model along with data from the trail itself.  This additional data for the turnout 
could increase the ability of the model to accurately predict the trail erosion.   
? Additional field data would greatly improve the results from this study.  The 
additional data could be generated through an extended study at the two locations 
from this research, or by the addition of sampler locations.  More field data would 
allow for an improved model calibration and more complete validation 
procedures.  The data collected for this project occurred during an extremely dry 
period.  Further data collection could sample data through normal weather 
conditions, and also wet conditions, which would add to more accurate average 
annual erosion data.  The average annual erosion data could be used to simulate 
the trail system in RUSLE2.  RUSLE2 was designed more for planning and uses 
only average annual sediment production.     
 
53 
 
? Other factors could affect the prediction ability of the model also.  These factors 
should be addressed to determine there affect on erosion from the trail in future 
research.  These factors include:  time of year, temperature affects on soil 
parameters, effect of maintenance practices on soil erosion, and traffic direction, 
especially related to uphill or downhill driving. 
 
 
54 
 
REFERENCES 
 
Adams, J.A., A.S. Endo, L.H. Stolzy, P.G. Rowlands, and H.B Johnson.  1982. 
Controlled experiments on soil compaction produced by off-road vehicles in the 
Mojave Desert, California.  Journal of Applied Ecology 19:167-175. 
 
Alabama Forestry Commission.  1993. Alabama?s Best Management  
Practices for Forestry.  Montgomery, Al. 
 
Appelboom, T.W., G.M. Chescheir, R.W. Skaggs, D.L. Hesterberg.  2001.   
Evaluating Forest Road Management Practices for Reducing Sediment Production 
and transport From Forested Watersheds.  ASAE 701P10007 
 
Ayala, R. D., P. Srivastava, C. J. Brodbeck, E. A. Carter, and T.P. McDonald.  Water 
Quality Impacts from an ORV Trail Stream Crossing in the Talladega National 
Forest, Alabama, USA.  Presented at the Annual COFE Meeting.  2005. 
 
Brinker, R.W.  1995.  Forest roads and construction of associated water diversion 
devices.  Alabama Cooperative Extension Service.  ANR-916.
 
 
55 
 
Brodbeck, C.J.,  2005.  Modeling Water Quality Impacts Off Road Vehicles in Forested 
Watersheds..  Auburn, AL.  MS Thesis, Auburn University. 
 
Brooks, K.N., P.F. Ffolliott, H.M. Gregersen, and L.F. Debano.  2003.  Hydrology and 
the Management of Watersheds.  Ames, Iowa.:  Blackwell Publishing. 
 
Eav, L.  Unmanaged recreation.  Key Messages and Talking Points from the Fourt 
Threats of Leadership Information Packer, USDA Forest Service, Office of 
Communication ? Washington D.C. 
 
Eckert, Jr., R.E., M.K. Wood, W.H. Blackburn, and F.F. Peterson. 1979.  Impacts of off-
road vehicles on infiltration and sediment production of two desert soils. Journal 
of Range Management 32(5):394-397. 
 
Foltz, R.B., 2006.  Erosion from all terrain vehicle (ATV) trails on National Forest lands.  
ASABE Paper Number 068012.  St. Joseph, MI:  ASABE.  Pranson, M.A., 
managing ed.  1985.  Standard methods for the examination of water and 
wastewater, 16
th
 ed. Baltimore, MD:  Port City Press. 
 
Grace, J.M.  2002.  Overview of Best Management Practices related to  
Forest Roads:  The Southern States.  ASAE Meeting Presentation.  Paper No: 
025013 
 
 
56 
 
Grace, J.M.  2005.  Application of WEPP to a Southern Appalachian Forest  
Road.  ASAE International Meeting Presentation Paper No: 052016 
 
Grace, J.M.  2005b.  Forest Operations and Water Quality in the South.  Trans. ASAE 
48(2):871-880 
 
Grace, J.M. and B.D. Clinton.  2006.  Forest Road Management to Protect Soil and 
Water.  ASABE International Meeting Presentation Paper No:  068010. 
 
Khan A.A., and C. K. Ong.  1997.  Design and calibration of tipping bucket system for 
field runoff and sediment quantification.  Journal of Soil and Water Conservation.  
52(6) 437-443. 
 
Klik, A, and A.S. Zartl.  2001. Comparison of Soil Erosion Simulations Using WEPP and 
RUSLE with Field Measurements.  Soil Erosion Research for the 21
st
 Century, 
Proc, Int. Conf.  ASABE Pub # 701P0007. 
 
Kutiel, P., E. Eden, and Y.Zhevelev.  2000.  Effect of experimental tramping and off-road 
motorcycle traffic on soil and vegetation of stabilized coastal dunes, Isreal. 
Environmental Conservation 27(1) : 14-23. 
 
 
 
 
57 
 
Moriasi, N., J.G. Arnold, M.W. Van Liew, R.L. Bingner, R.D. Harmel, and T. Veith.  
2007.  Model evaluation guidelines for systematic quantification of accuracy in 
watershed simulations.  Transactions of ASABE.  In press. 
 
Novotny, V.  2003.  Water Quality:  Diffuse Pollution and Watershed Management.  New 
York:  John Wiley & Sons. 
 
NSERL.  1995.  USDA-Water Erosion Prediction Project Hillslope Profile and  
Watershed Model Documentation. NERSL National Soil Erosion Research 
Laboratory 
 
Ohio State University.  Streams: Stream Restoration, Ecology, and Aquatic Management 
Solutions.  http://streams.osu.edu/book/equation_pdf/Ch9-Equations.pdf.  Last 
visited October 7, 2007. 
 
Sack, D., and S. da Luz.  2003.  Sediment flux and compaction trends on off-road vehicle 
(ORV) and other trails in an Appalachian forest setting.  Physical Geography 
24(6): 536-554. 
 
Sherrod, T., A. Blankenship.  RUSLE2 Analysis as an Erosion and Sedimentation 
Control Design Alternative for Linear Transportation Projects.  North Carolina 
department of Transportation Transaction.  2004. 
 
 
58 
 
USDA-ARS.  2003.  USDA- Revised Universal Soil Loss Equation, Version 2: User?s 
Guide.  USDA-NRCS, Columbus Ohio. 
 
USDA-ARS MWA.  Universal Soil Loss Database.  
http://topsoil.nserl.purdue.edu/usle/page7.html.  Last visited October 7, 2007. 
 
 
59 
 
APPENDIX A 
 
Rainfall and Sediment Sampling 
 
60 
 
Figure A.1 ? Cumulative rainfall for 11/15/2006. 
Cumulative Rainfall 11/15/2006
0
0.5
1
1.5
2
2.5
11/15/2006 4:48 11/15/2006 9:36 11/15/2006 14:24 11/15/2006 19:12 11/16/2006 0:00 11/16/2006 4:48 11/16/2006 9:36
time
in
c
h
e
s
 o
f
 r
a
in
fa
ll
 
Figure A.2 ? Rainfall hydrograph for 11/15/2006. 
Rainfall Intensity 11/15/06
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
11/15/2006 4:48 11/15/2006 9:36 11/15/2006 14:24 11/15/2006 19:12 11/16/2006 0:00 11/16/2006 4:48 11/16/2006 9:36
Time
i
n
ches
 pe
r
 
ho
ur
 
 
61 
 
Figure A.3 ? Cumulative runoff for mild and steep slope plots for 11/15/2006.  
Cumulative Runoff for 11/15/06
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
11/14/2006
19:12
11/15/2006 0:00 11/15/2006 4:48 11/15/2006 9:36 11/15/2006
14:24
11/15/2006
19:12
11/16/2006 0:00
Time
Li
t
e
r
s
Mild Slope
Steep Slope
 
Figure A.4 -- Runoff intensity for 11/15/06. 
Runoff Intensity
-10
0
10
20
30
40
50
60
70
11/14/2006
19:12
11/15/2006
0:00
11/15/2006
4:48
11/15/2006
9:36
11/15/2006
14:24
11/15/2006
19:12
11/16/2006
0:00
11/16/2006
4:48
Time
Li
t
e
r
s
 per
 hour
Steep Slope
mild slope
 
 
62 
 
Figure A.5 ? Cumulative rainfall for 12/22/2006. 
Cumulative Rainfall 12/22/2006
1
1.05
1.1
1.15
1.2
1.25
12/22/2006 0:00 12/22/2006 4:48 12/22/2006 9:36 12/22/2006
14:24
12/22/2006
19:12
12/23/2006 0:00 12/23/2006 4:48 12/23/2006 9:36
Time
I
n
ch
es
 
Figure A.6 --Rainfall hydrograph for 12/22/2006. 
Rainfall Intensity 12/22/2006
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
12/22/2006 0:00 12/22/2006 4:48 12/22/2006 9:36 12/22/2006
14:24
12/22/2006
19:12
12/23/2006 0:00 12/23/2006 4:48 12/23/2006 9:36
time
i
n
ches per
 hr
 
 
63 
 
Figure A.7 ? Cumulative runoff for mild and steep slope plots for 12/22/2006. 
Cumulative Runoff 12/22/06
0
500
1000
1500
2000
2500
3000
12/21/2006
21:36
12/22/2006 0:00 12/22/2006 2:24 12/22/2006 4:48 12/22/2006 7:12 12/22/2006 9:36 12/22/2006
12:00
Time
Li
t
e
r
s
 per
 m
i
nu
t
e
Steep Slope
Mild Slope
 
Figure A.8 -- Runoff intensity for 12/22/06. 
Runoff Intensity
-10
0
10
20
30
40
50
60
70
12/21/2006
21:36
12/22/2006 0:00 12/22/2006 2:24 12/22/2006 4:48 12/22/2006 7:12 12/22/2006 9:36 12/22/2006
12:00
Time
Li
t
e
r
s
 per
 hou
r
Steep Slope
Mild Slope
 
 
64 
 
Figure A.9 ? Cumulative rainfall for 2/1/2007. 
Cumulative Rainfall 2/1/2007
0
0.2
0.4
0.6
0.8
1
1.2
1/31/2007 4:48 1/31/2007 9:36 1/31/2007 14:24 1/31/2007 19:12 2/1/2007 0:00 2/1/2007 4:48 2/1/2007 9:36
Time
I
n
ch
es
 
Figure A.10 ? Rainfall hydrograph for 2/1/2007. 
Rainfall Intensity 2/1/2007
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
1/31/2007 4:48 1/31/2007 9:36 1/31/2007 14:24 1/31/2007 19:12 2/1/2007 0:00 2/1/2007 4:48 2/1/2007 9:36
Time
In
c
h
e
s
 p
e
r
 H
r
 
 
65 
 
Figure A.11? Cumulative runoff for mild and steep plots for 2/1/2007. 
Cumulative Runoff for 2/1/07
0
500
1000
1500
2000
2500
3000
1/31/07 9:36
PM
2/1/07 12:00
AM
2/1/07 2:24
AM
2/1/07 4:48
AM
2/1/07 7:12
AM
2/1/07 9:36
AM
2/1/07 12:00
PM
2/1/07 2:24
PM
Time
Li
t
e
r
s
Steep Slope
Mild Slope
 
Figure 12 -- Runoff intensity for 2/1/07. 
Runoff Intensity
-5
0
5
10
15
20
25
30
35
1/31/07 9:36
PM
2/1/07 12:00
AM
2/1/07 2:24
AM
2/1/07 4:48
AM
2/1/07 7:12
AM
2/1/07 9:36
AM
2/1/07 12:00
PM
2/1/07 2:24
PM
Time
Li
t
e
r
s
 per
 ho
ur
Steep Slope
Mild Slope
 
 
66 
 
Figure A.13 ? Cumulative rainfall for 3/1/2007. 
Cumulative Rainfall 3/1/2007
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
3/1/2007 14:24 3/1/2007 19:12 3/2/2007 0:00 3/2/2007 4:48 3/2/2007 9:36 3/2/2007 14:24 3/2/2007 19:12 3/3/2007 0:00
Time
I
n
ch
es
 
Figure A.14 ? Rainfall hydrograph for 3/1/2007. 
Rainfall Intensity
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
3/1/2007 14:24 3/1/2007 19:12 3/2/2007 0:00 3/2/2007 4:48 3/2/2007 9:36 3/2/2007 14:24 3/2/2007 19:12 3/3/2007 0:00
Time
I
n
ches per
 
h
our
Series1
 
 
67 
 
Figure A.15 ? Cumulative runoff for mild and steep slope plots for 3/1/2007. 
Cumulative Runoff for 3/1/2007
0
200
400
600
800
1000
1200
1400
1600
1800
3/1/07 12:00 PM 3/2/07 12:00 AM 3/2/07 12:00 PM 3/3/07 12:00 AM 3/3/07 12:00 PM 3/4/07 12:00 AM
Time
Li
t
e
r
s
Steep Slope
Mild Slope
 
Figure A.16 -- Runoff intensity for 3/1/07. 
Runoff Intensity
-5
0
5
10
15
20
25
30
35
40
45
3/1/07 12:00 PM 3/2/07 12:00 AM 3/2/07 12:00 PM 3/3/07 12:00 AM 3/3/07 12:00 PM 3/4/07 12:00 AM
Time
L
i
t
e
rs
 p
e
r m
i
n
steep slope
Mild Slope
 
 
68 
 
Figure A.17 ? Cumulative rainfall for 3/15/2007.  
Cumulative Rainfall for 3/15/07
0
0.5
1
1.5
2
2.5
3/15/2007 14:24 3/15/2007 19:12 3/16/2007 0:00 3/16/2007 4:48 3/16/2007 9:36 3/16/2007 14:24 3/16/2007 19:12
Time
In
c
h
e
s
 
Figure A.18 ? Rainfall hydrograph for 3/15/2007. 
Rainfall Intensity for 3/15/07
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
3/15/2007 14:24 3/15/2007 19:12 3/16/2007 0:00 3/16/2007 4:48 3/16/2007 9:36 3/16/2007 14:24 3/16/2007 19:12
Time
I
n
ch
es
 p
e
r ho
ur
 
 
69 
 
Figure A.19 ? Cumulative runoff for mild slope and steep slope for 3/15/2007.  
Cumulative Runoff for 3/15/2007
0
1000
2000
3000
4000
5000
6000
3/15/2007
14:24
3/15/2007
19:12
3/16/2007
0:00
3/16/2007
4:48
3/16/2007
9:36
3/16/2007
14:24
3/16/2007
19:12
3/17/2007
0:00
Time
Li
t
e
r
s
Steep Slope
Mild Slope
 
Figure A.20 -- Runoff intensity for 3/15/2007. 
Runoff Intensity
-20
0
20
40
60
80
100
120
140
3/15/2007
14:24
3/15/2007
19:12
3/16/2007
0:00
3/16/2007
4:48
3/16/2007
9:36
3/16/2007
14:24
3/16/2007
19:12
3/17/2007
0:00
Time
Li
t
e
r
s
 per
 m
i
n
Steep Slope
Mild Slope
 
70 
 
APPENDIX B 
 
Rainfall and Runoff Data 
 
71 
 
Table B.1 -- Rainfall and runoff data collected from sites. 
 
Mild Slope 
Plot Size trail length = 54 ft Plot area 672 ft
2
  
  turnout length= 30 ft  0.015 Acre 
  trail width = 8 ft       
Storm Date 
Storm Duration 
(hr) 
Max 
Storm 
Intensity 
(in/hr) 
Storm 
Rainfall 
(in) 
Duration 
to Peak 
intensity 
Actual 
Runoff 
(L) 
Actual 
Sediment 
Yield 
(Kg) 
11/15/2006 9.5 0.48 2.28 0.13 6112 11.012
12/22/2006 6.75 0.28 1.21 0.67 2186 28.809
2/1/2007 3.75 0.36 1.05 0.60 2551 5.310
3/1/2007 0.5 0.56 0.14 0.50 567 1.697
3/15/2007 4.25 1.11 2 0.24 2095 16.570
              
              
Steep Slope 
Plot Size trail length = 120 ft Plot area 1240 ft
2
 
  turnout length= 35 ft  0.028 Acre 
  trail width = 8 ft     
              
Storm Date 
Storm Duration 
(hr) 
Max 
Storm 
Intensity 
(in/hr) 
Storm 
Rainfall 
(in) 
Duration 
to Peak 
intensity 
Actual 
Runoff 
(L) 
Actual 
Sediment 
Yield 
(Kg) 
11/15/2006 9.5 0.48 2.28 0.13 3873.00 53.168
12/22/2006 6.75 0.28 1.21 0.67 1336.00 101.172
2/1/2007 3.75 0.36 1.05 0.60 1090.00 11.518
3/1/2007 0.5 0.56 0.14 0.50 1907.86 7.914
3/15/2007 4.25 1.11 2 0.24 5488.76 45.319
 
 
 
 
72 
 
 
 
Table B.2 -- Sediment data from research sites. 
 
 
Date of Storm Slope 
Runoff 
Vol (L) 
Runoff 
Sample 
Size 
Pan 
Wt. (g) 
Dry Wt. 
(g) 
Sediment 
Wt. (g) 
Total 
Sediment 
in 
sample 
(g) 
Total 
Sediment 
Loading (g) 
Total 
Sediment 
Loading 
(kg) 
11/15/2006 Mild 6112 0.01 764.54 764.9 0.36 63.96 11011.93 11.01 
        590.2 601 10.8       
        584.7 637.5 52.8       
11/15/2006 Steep 7746 0.01 572.8 593.6 20.8 123.55 53167.68 53.17 
        728.85 831.6 102.75       
12/22/2006 Mild 2186 0.01 572.9 823.3 250.4 250.4 28809.18 28.81 
12/22/2006 Steep 2672 0.01 584.9 1417.9 833 833 101171.64 101.17 
2/1/2007 Mild 2551 0.01 729.8 781.84 52.04 52.04 5310.16 5.31 
2/1/2007 Steep 2180 0.01 765 860.1 95.1 95.1 11517.67 11.52 
3/1/2007 Mild 567 0.01 584.93 589.57 4.64 4.64 1697.34 1.70 
3/1/2007 Steep 1907.86 0.01 590.15 906.7 316.55 316.55 7913.75 7.91 
3/15/2007 Mild 2095 0.01 730.95 874.5 143.55 143.55 16569.55 16.57 
3/15/2007 Steep 5488.76 0.01 766.8 921.2 154.4 154.4 45318.99 45.32 
 
 
73 
 
APPENDIX C 
 
Parameter Calibration for Runoff and Sediment Production 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
74 
 
Table C.1 -- Calibration of effective hydraulic 
conductivity for runoff and sediment 
Effective Hydraulic Conductivity 
  (Kb) NSE value 
Ki=3e005 0.5 -1.01281 
Kr=0.0005 0.4 -0.92853 
Tc=0.05 0.35 -0.9605 
  0.3 -0.89901 
  0.25 -0.8474 
  0.2 -0.79231 
  0.19 -0.73417 
  0.18 -0.47872 
  0.17 -0.25759 
  0.16 -0.06488 
  0.15 0.372289 
  0.14 0.410045 
  0.13 0.45408 
  0.125 0.407816 
  0.12 0.3588 
      
  Max NSE= 0.45408 
 
 
 
 
 
 
 
 
 
 
75 
 
Figure C.1 ? Calibration of effective hydraulic conductivity for runoff and sediment 
Effective Hydraulic Conductivity Calibration
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0 0.1 0.2 0.3 0.4 0.5 0.6
K
b
NS
E
 
 
Table C.2 -- Calibration of critical shear 
for runoff and sediment 
Critical Shear 
  (Tc) R
2
 value 
Ki=3.0e006 0.1 0.4182
Kr=0.0005 0.2 0.4163
Kb=0.13 0.05 0.4191
  0.04 0.4192
  0.03 0.4194
  0.02 0.4196
  0.01 0.4198
  0.005 0.4199
  0.004 0.4199
      
  Max R
2
 = 0.4199
 
76 
 
Figure C.2 ? Calibration of critical shear for runoff and sediment 
Critical Shear Calibration
0.416
0.4165
0.417
0.4175
0.418
0.4185
0.419
0.4195
0.42
0.4205
0 0.05 0.1 0.15 0.2 0.25
T
c
R
 s
qua
r
e
 
 
Table C.3 -- Calibration of rill erodibility 
for runoff and sediment 
Rill Erodibility 
  (Kr) R
2
 value 
Ki=3.0e006 0.0008 0.4192
Tc=0.005 0.0007 0.4199
Kb=0.13 0.0006 0.4204
  0.0005 0.4206
  0.0004 0.4202
      
  Max R
2
 = 0.4206
 
 
 
77 
 
Figure C.3 ? Calibration of rill erodibility for runoff and sediment. 
Rill Erodibility Calibration
0.419
0.4192
0.4194
0.4196
0.4198
0.42
0.4202
0.4204
0.4206
0.4208
0 0.0002 0.0004 0.0006 0.0008 0.001
K
r
R
 squar
e
 
 
Table C.4 --  Calibration of interrill 
erodibility for runoff and sediment 
Interrill Erodibility  
  (Ki) R
2
 value 
Kr=0.0005 1.10E+06 0.4199
Tc=0.005 9.00E+05 0.4212
Kb=0.13 8.00E+05 0.4219
  7.00E+05 0.4226
  6.00E+05 0.4232
  5.00E+05 0.4239
  4.00E+05 0.4246
  3.00E+05 0.4253
  2.00E+05 0.426
  1.00E+05 0.4266
  9.00E+04 0.4212
      
  Max R
2
 = 0.4266
 
78 
 
Figure C.4 ? Calibration of interrill erodibility for runoff and sediment 
Interrill Erodibility Calibration
0.419
0.42
0.421
0.422
0.423
0.424
0.425
0.426
0.427
0.00E+00 2.00E+05 4.00E+05 6.00E+05 8.00E+05 1.00E+06 1.20E+06
K
i
R s
q
u
a
r
e
 
 
 
 
 
 
 
 
 
APPENDIX D 
 
 
79 
 
Parameter Calibration for Sediment Production 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Table D.1 -- Calibration for rill erodibility for 
sediment production 
Rill Erodibility 
  (Kr) NSE value 
K
i
=6.0e006 0.05 -918.2707 
 
80 
 
T
c
=2.57 0.01 -750.0188 
K
b
=.013 0.0075 -665.8518 
  0.005 -443.1096 
  0.0025 -143.3983 
  0.001 -25.15484 
  0.0009 -20.29555 
  0.0008 -15.90674 
  0.0007 -12.00845 
  0.0006 -8.620284 
  0.0005 -5.762646 
  0.0004 -3.453229 
  0.0003 -1.717375 
  0.0002 -0.573163 
  0.0001 -0.044787 
  0.00009 -0.026533 
  0.00008 -0.015051 
  0.00007 -0.009664 
  0.00006 -0.010932 
      
  Max NSE= -0.009664 
 
 
 
 
 
 
 
 
 
Figure D.1 ? Calibration of rill erodibility. 
 
81 
 
Rill Erodibility(Ki=6.0e006, Tc=2.57, Kb=0.13)
-1000
-800
-600
-400
-200
0
200
0 0.01 0.02 0.03 0.04 0.05 0.06
Kr
NS
E
 
 
 
 
 
 
 
 
 
Table D.2 -- Calibration for critical 
shear for sediment production 
Critical Shear 
  (Tc) NSE value 
 
82 
 
K
i
=6.0e006 2.5 -0.007631
K
r
=0.00007 2.25 -0.001204
K
b
=.013 2 0.005517
  1.75 0.011844
  1.5 0.018285
  1.25 0.024687
  1 0.034254
  0.75 0.039267
  0.5 0.042696
  0.25 0.047449
  0.1 0.050166
  0.01 0.052061
  0.001 0.052171
  0.009 0.052061
  0.008 0.051954
  0.006 0.052401
  0.005 0.052401
  0.004 0.052171
  0.0001 0.052171
      
  
Max 
NSE= 0.052401
 
 
 
 
 
 
 
Figure D.2 ? Calibration of critical shear for sediment production. 
 
83 
 
Critical Shear (Ki=6.0e06, Kr=.00007, Kb=0.13)
-0.02
-0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
00.511.522.53
Tc
NS
E
 
 
 
 
 
 
 
 
 
 
 
84 
 
Table D.3 -- Calibration for interrill 
erodibility for sediment production 
Interrill Erodibility  
  (Ki) NSE value 
K
r
=0.00007 6.00E+06 0.052401
T
c
=0.006 6.10E+06 0.04359
K
b
=0.13 5.90E+06 0.067866
  5.80E+06 0.067866
  5.70E+06 0.075683
  5.60E+06 0.082601
  5.50E+06 0.089259
  5.00E+06 0.119105
  4.50E+06 0.141449
  4.00E+06 0.155622
  3.50E+06 0.162777
  3.40E+06 0.163035
  3.30E+06 0.16353
  3.20E+06 0.163197
  3.00E+06 0.161804
      
  
Max 
NSE= 0.16353
 
 
 
 
 
 
 
 
 
85 
 
 
Figure D.3 ? Calibration of interrill erodibility for sediment production. 
Interrill Erodibility (Kr=0.00007, Tc=.006, Kb=.013
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.00E+
00
1.00E+
06
2.00E+
06
3.00E+
06
4.00E+
06
5.00E+
06
6.00E+
06
7.00E+
06
Ki
NS
E
 
 
 
 
 
 
 
 
 
 
86 
 
 
Table D.4 -- Calibration for effective 
hydraulic conductivity for sediment 
production 
Effective Hydraulic Conductivity 
  (Kb) NSE value 
K
i
=3.2e006 1 -0.45666
K
r
=0.0001 0.6 -0.27898
Tc=0.006 0.52 -0.18386
  0.48 -0.16071
  0.15 0.201756
  0.13 0.214655
  0.12 0.219767
  0.1 0.228355
  0.09 0.232128
  0.08 0.235114
  0.07 -0.4151
  0.06 -0.4151
      
  Max NSE= 0.235114
 
 
 
 
 
 
 
 
 
87 
 
Figure D.4 ? Calibration of effective hydraulic conductivity for sediment production. 
Rill Erodibilty (Ki=3.2e06, Tc=0.006, Kb=0.13
-10
-8
-6
-4
-2
0
2
0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006
Kr
NS
E
 
 
88 
 
APPENDIX E 
 
Calibration of Mild Slope 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
89 
 
Table E.1 -- Calibration of interrill 
erodibility on mild slope 
Interrill Erodibility  
  (Ki) NSE value 
K
r
=0.0001 3.20E+06 -0.34185
T
c
=0.006 3.30E+06 -0.33529
K
b
=0.13 3.40E+06 -0.32748
  3.50E+06 -0.32008
  3.60E+06 -0.3131
  3.70E+06 -0.30816
  3.80E+06 -0.30202
  3.90E+06 -0.2963
  4.00E+06 -0.29101
  4.20E+06 -0.2833
  4.40E+06 -0.27569
  4.60E+06 -0.2712
  4.80E+06 -0.26697
  5.00E+06 -0.26577
  5.10E+06 -0.26512
  5.20E+06 -0.26491
  5.30E+06 -0.26638
  5.40E+06 -0.26718
      
  Max NSE= -0.26491
 
 
 
 
 
 
 
 
90 
 
Figure E.1 ? Calibration of interrill erodibility on mild 
slope.
Interrill Erodibilty (Kr=0.0001, Tc=0.006, Kb=0.13
-0.4
-0.35
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.00E+00 1.00E+06 2.00E+06 3.00E+06 4.00E+06 5.00E+06 6.00E+06
Ki
NS
E
 
 
Table E.2 -- Calibration of rill 
erodibility on mild slope 
Rill Erodibility 
  (Kr) NSE value 
K
i
=5.2e006 0.0005 -2.13696
0.006 0.0002 -0.24432
K
b
=.013 0.00017 -0.21404
  0.00016 -0.21119
  0.00015 -0.2115
  0.00011 -0.24736
  0.0001 -0.26491
      
      
  
Max 
NSE= -0.21119
 
91 
 
 
Figure E.2 ? Calibration of rill erodibility on mild slope. 
Rill Erodibility (Ki=5.2e06, Tc=0.006, Kb=0.13)
-2.5
-2
-1.5
-1
-0.5
0
0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006
Kr
NS
E
 
Table E.3 -- Calibration of 
critical shear on mild slope 
Critical Shear 
  (Tc) NSE value 
K
i
=5.2e006 0.01 -0.21174
K
r
=0.00016 0.009 -0.21079
K
b
=.013 0.008 -0.21079
  0.007 -0.21079
  0.006 -0.21119
      
  
Max 
NSE= -0.21079
 
 
92 
 
 
Figure E.3 ? Calibration of critical shear on mild slope 
Critical Shear (Ki=5.2e06, Kr=0.00016, Kb=0.13
-0.2118
-0.2116
-0.2114
-0.2112
-0.211
-0.2108
-0.2106
0 0.002 0.004 0.006 0.008 0.01 0.012
Tc
NS
E
 
Table E.4 -- Calibration of effective 
hydraulic conductivity on mild slope 
Effective Hydraulic Conductivity 
  (Kb) NSE value 
K
i
=5.2e006 0.13 -0.21174
K
r
=0.00016 0.12 -0.20403
Tc=0.009 0.1 -0.19414
  0.09 -0.19018
  0.08 -0.18489
  0.07 -0.98212
  0.06 -0.98212
      
  Max NSE= -0.18489
 
93 
 
 
Figure E.4 ? Calibration of effective hydraulic conductivity on mild slope. 
Effective Hydraulic Conductivity (Ki=5.2e06, 
Kr=0.00016, Tc=0.009)
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Kb
NS
E
 
94 
 
APPENDIX F 
 
Calibration of Steep Slope 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
95 
 
Table F.1 -- Interrill erodibility 
calibration on steep slope 
Interrill Erodibility  
  (Ki) 
NSE 
value 
K
r
=0.0001 3.20E+06 -0.05415
T
c
=0.006 3.00E+06 -0.03791
K
b
=0.13 2.80E+06 -0.02061
  2.60E+06 -0.00514
  2.40E+06 0.008256
  2.20E+06 0.019875
  2.00E+06 0.029664
  1.80E+06 0.037421
  1.60E+06 0.043375
  1.40E+06 0.047303
  1.20E+06 0.049561
  1.10E+06 0.050264
  1.00E+06 0.049848
  9.00E+05 0.049606
  8.00E+05 0.048239
      
  
Max 
NSE= 0.050264
 
 
 
 
 
 
 
 
 
96 
 
Figure F.1 ? Calibration of interrill erodibility on steep slope. 
Interrill Erodibility (Kr=0.0001, Tc=0.006, Kb=0.13)
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.00E+0
0
5.00E+0
5
1.00E+0
6
1.50E+0
6
2.00E+0
6
2.50E+0
6
3.00E+0
6
3.50E+0
6
Ki
NS
E
 
 
 
Table F.2 -- Rill erodibility 
calibration on steep slope 
Rill Erodibility 
  
(Kr) 
NSE 
value 
K
i
=1.1e006 0.0002 -0.51614
T
c
=0.006 0.00015 0.006348
K
b
=.013 0.00013 0.081673
  0.00012 0.090807
  0.00011 0.08065
  0.0001 0.050264
      
  
Max 
NSE= 0.090807
 
97 
 
Figure F.2 ? Calibration of rill erodibility on steep slope. 
Rill Erodibility (Ki=1.1e06, Tc=0.006, Kb=0.13)
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0 0.00005 0.0001 0.00015 0.0002 0.00025
Kr
NS
E
 
 
Table F.3 -- Critical shear 
calibration on steep slope 
Critical Shear 
  (Tc) 
NSE 
value 
K
i
=1.1e006 0.006 0.090807
K
r
=0.00012 0.007 0.090935
K
b
=.013 0.008 0.090935
  0.009 0.091074
  0.01 0.090478
      
  
Max 
NSE= 0.091074
 
 
98 
 
Figure F.3 ? Calibration of critical shear on steep slope. 
Critical Shear (Ki=1.1e06, Kr=0.00012, Kb=0.13)
0.0904
0.0905
0.0906
0.0907
0.0908
0.0909
0.091
0.0911
0.0912
0 0.002 0.004 0.006 0.008 0.01 0.012
Tc
NS
E
 
 
Table F.4 -- Effective hydraulic conductivity 
calibration on steep slope 
Effective Hydraulic Conductivity 
  (Kb) NSE value 
K
i
=1.1e006 0.085 0.095153 
K
r
=0.00012 0.076 0.105616 
Tc=0.009 0.075 0.106232 
  0.074 0.105616 
  0.07 -0.98167 
      
  Max NSE= 0.106232 
 
 
99 
 
 
 
 
Figure F.4 ? Calibration of effective hydraulic conductivity on steep slope. 
Effective Hydraulic Conductivity (Ki=1.1e06, 
Kr=0.00012, Tc=0.009)
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0 0.02 0.04 0.06 0.08 0.1
Kb
NS
E
 
 
100 
 
APPENDIX G 
 
Calibration of Trafficked Period 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
101 
 
Table G.1 -- Interrill erodibility 
calibration for trafficked period 
Interrill Erodibility  
  (Ki) NSE value 
K
r
=0.0001 3.30E+06 -0.16269
T
c
=0.006 3.50E+06 -0.15929
K
b
=0.08 3.75E+06 -0.15728
  4.00E+06 -0.15613
  4.25E+06 -0.15605
  4.26E+06 -0.15585
  4.27E+06 -0.15638
  4.50E+06 -0.15706
      
  Max NSE= -0.15585
 
 
Figure G.1 ? Calibration of interrill erodibility for trafficked period. 
Interrill Erodibility Calibration
-0.164
-0.163
-0.162
-0.161
-0.16
-0.159
-0.158
-0.157
-0.156
-0.155
0.00E+
00
5.00E+
05
1.00E+
06
1.50E+
06
2.00E+
06
2.50E+
06
3.00E+
06
3.50E+
06
4.00E+
06
4.50E+
06
5.00E+
06
Ki
NS
E
 
 
102 
 
Table G.2 -- Rill erodibility 
calibration for trafficked period 
Rill Erodibility 
  (Kr) 
NSE 
value 
K
i
=4.26e006 0.0001 -0.15585
T
c
=0.006 0.00011 -0.13375
K
b
=0.08 0.00012 -0.1268
  0.00013 -0.13557
  0.00015 -0.19968
      
  Max NSE= -0.1268
 
 
Figure G.2 ? Calibration of rill erodibility for trafficked period. 
Rill Erodibility Calibration
-0.25
-0.2
-0.15
-0.1
-0.05
0
0 0.00002 0.00004 0.00006 0.00008 0.0001 0.00012 0.00014 0.00016
Kr
NS
E
 
 
 
 
103 
 
Table G.3 -- Critical shear 
calibration for trafficked period 
Critical Shear 
  (Tc) NSE value 
K
i
=4.26e006 0.006 -0.1268
K
r
=0.00012 0.007 -0.1268
K
b
=0.08 0.008 -0.12656
  0.009 -0.12723
      
  
Max 
NSE= -0.12656
 
 
 
Figure G.3 ? Calibration of critical shear for trafficked period. 
Critical Shear Calibration
-0.1273
-0.1272
-0.1271
-0.127
-0.1269
-0.1268
-0.1267
-0.1266
-0.1265
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
Tc
NS
E
 
 
 
104 
 
Table G.4 -- Effective hydraulic 
conductivity for trafficked period 
Effective Hydraulic Conductivity
  (Kb) NSE value 
K
i
=4.26e006 0.09 -0.14049
K
r
=0.00012 0.085 -0.13309
Tc=0.008 0.081 -0.12812
  0.08 -0.12656
  0.079 -0.12204
  0.075 -0.11719
  0.074 -0.11573
  0.073 -0.11454
  0.072 -0.11311
  0.071 -0.11195
  0.07 -1.5348
      
  
Max 
NSE= -0.11195
 
Figure G.4 ? Calibration of effective hydraulic conductivity for trafficked period. 
Effective Hydraulic Conductivity Calibration
-1.8
-1.6
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Kb
NS
E
 
105 
 
APPENDIX H 
 
Calibration of Non-Trafficked Period 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
106 
 
Table H.1 -- Interrill erodibility calibration 
for non-trafficked period 
Interrill Erodibility  
  (Ki) NSE value 
K
r
=0.0001 3.40E+06 -0.10296
T
c
=0.006 3.30E+06 -0.04888
K
b
=0.08 3.20E+06 0.027744
  3.10E+06 0.053299
  3.00E+06 0.103018
  2.90E+06 0.150044
  2.80E+06 0.194245
  2.70E+06 0.238386
  2.60E+06 0.278986
  2.50E+06 0.318342
  2.40E+06 0.3561
  2.30E+06 0.392089
  2.20E+06 0.426221
  2.10E+06 0.458543
  2.00E+06 0.488975
  1.90E+06 0.516967
  1.80E+06 0.54444
  1.70E+06 0.568809
  1.60E+06 0.591904
  1.50E+06 0.636167
  1.40E+06 0.632752
  1.20E+06 0.665871
  1.00E+06 0.692136
  8.00E+05 0.711241
  6.00E+05 0.722283
  4.20E+05 0.726197
  4.10E+05 0.726253
  4.00E+05 0.726244
  3.90E+05 0.726261
  3.00E+05 0.725428
  2.00E+05 0.72326
      
      
  Max NSE= 0.726261
 
107 
 
 
Figure H.1 ? Calibration of interrill erodibility for non-trafficked period. 
Interrill Erodibility
-0.2
0
0.2
0.4
0.6
0.8
0.00E+
00
5.00E+
05
1.00E+
06
1.50E+
06
2.00E+
06
2.50E+
06
3.00E+
06
3.50E+
06
4.00E+
06
Ki
NS
E
 
 
 
Table H.2 -- Rill erodibility calibration 
for non-trafficked period 
Rill Erodibility 
  (Kr) NSE value 
K
i
=3.90e005 0.00011 0.671071
T
c
=0.006 0.0001 0.726253
K
b
=0.08 0.00009 0.746634
  0.00008 0.730471
      
  
Max  
NSE= 0.746634
 
 
108 
 
 
Figure H.2 ? Calibration of rill erodibility for non-trafficked period. 
Rill Erodibility
0.66
0.67
0.68
0.69
0.7
0.71
0.72
0.73
0.74
0.75
0.76
0 0.00002 0.00004 0.00006 0.00008 0.0001 0.00012
Kr
NS
E
 
 
Table H.3 -- Critical shear calibration 
for non-trafficked period. 
Critical Shear 
  (Tc) NSE value 
K
i
=3.90e005 0.005 0.746366
K
r
=0.00009 0.0055 0.746366
K
b
=0.08 0.0056 0.746812
  0.0057 0.746812
  0.0058 0.746812
  0.0059 0.746634
  0.006 0.746634
      
  Max  NSE= 0.746812
 
 
109 
 
 
 
 
Figure H.3 ? Calibration of critical shear for non-trafficked period. 
Critical Shear
0.7463
0.7464
0.7465
0.7466
0.7467
0.7468
0.7469
0.0048 0.005 0.0052 0.0054 0.0056 0.0058 0.006 0.0062
Tc
NS
E
 
 
 
 
 
 
 
 
 
 
110 
 
Table H.4 -- Effective hydraulic conductivity 
for non-trafficked period 
Effective Hydraulic Conductivity 
  (Kb) NSE value 
K
i
=3.90e005 0.08 0.746812 
K
r
=0.00009 0.082 0.746817 
Tc=0.0058 0.084 0.747365 
  0.086 0.74799 
  0.088 0.748073 
  0.09 0.748904 
  0.092 0.74934 
  0.094 0.749796 
  0.096 0.750377 
  0.098 0.750543 
  0.1 0.750842 
  0.102 0.751198 
  0.104 0.751454 
  0.106 0.751744 
  0.108 0.752596 
  0.11 0.752789 
  0.115 0.753645 
  0.12 0.754378 
  0.125 0.754989 
  0.13 0.755477 
  0.135 0.756415 
  0.14 0.757521 
  0.15 0.758116 
  0.151 0.758084 
  0.152 0.758137 
  0.153 0.757913 
  0.154 0.757872 
  0.16 0.756849 
      
  Max  NSE= 0.758137 
 
 
 
111 
 
 
 
 
Figure H.4 ? Calibration of effective hydraulic conductivity for non-trafficked period. 
Hydraulic Conductivity
0.746
0.748
0.75
0.752
0.754
0.756
0.758
0.76
0 0.05 0.1 0.15 0.2
Kb
NS
E
 
112 
 
APPENDIX I 
 
Tipping Bucket Design 
 
 
 
 
 
113 
 
Figure I.1 -- Tipping bucket design.  
 
Figure I.2 ? Tipping bucket base. 
          
 
 
114 
 
 
Table I.1 ? Tipping bucket parts list. 
 
Tipping Bucket Parts Sheet 
Component Material Quantity
vertical structural member 2" x 0.125" x 7" square tubing 4
middle structural member 2" x 0.125" x 8" square tubing 2
horizontal structural member 2" x 0.125" x 8.25" square tubing 2
bottom mount 0.125" x 2" x 8.25" bar stock 2
axle mounts 0.125" x 2" x 7.625" bar stock 2
rainfall chamber 19.7" x 23.625" x1/16" steel sheet 1
rainfall chamber divider 5.9" x 17.8" x1/16" steel sheet 1
pivoting axle 3/8" rod 1
 
 
 
 
 
 
 
 
115 
 
 
 
 
Figure I.3 ? Middle and vertical member drawings. 
 
 
 
 
 
 
 
116 
 
 
 
 
Figure I.4 ? Horizontal member drawing. 
 
 
 
 
 
 
 
117 
 
 
 
Figure I.5 ? Bottom mount drawing. 
 
 
 
 
 
 
 
 
118 
 
 
 
Figure I.6 ? Tipping bucket axle mount drawing.  
 
 
 
 
 
 
 
 
119 
 
 
Figure I.7 ? Rainfall chamber drawing. 
 
 
 
 
 
 
 
 
 
120 
 
 
Figure I.8 ? Rainfall chamber divider drawing. 
 
 
 
 
 
 
 
 
 
121 
 
 
 
Figure I.9 ? Tipping bucket design and parts location. 
 
 
 
 
 
 
 
 
122 
 
APPENDIX J 
 
Water Diversion Spacing 
 
123 
 
Table J.1 -- Recommended spacing for water diversion structures. 
 
Recommended Water Diversion Structure Spacing 
Slope ( % ) Spacing ( ft ) 
2 85 
4 45 
6 30 
8 20 
10 15 
12 13 
14 10 
 
Table J.2 -- Water diversion spacing developed by Brodbeck. 
 
Recommended Diversion Structure Spacing developed by 
Brodbeck 
Slope ( % ) Spacing ( ft ) 
4 46 
6 36 
8 33 
10 29 
12 27 
14 22 
18 19 
 
124 
 
 
 
Table J.3 -- Recommended distances between water bars on skid trails and logging roads 
after logging is complete (Brinker, 1995). 
 
Recommended Distances Between Water Bars On Skid 
Trails and Logging Roads After Logging Is Complete. 
Road Grade Spacing (feet) 
2 250 
5 135 
10 80 
15 60 
20 45 
25 40 
30 35 
40 30 
 
 
 
 
 
 
 
125 
 
Table J.4 ? Water bar and broad-based dip spacing as recommended by the Alabama 
BMP Manual for Forest Operations. 
 
 
 
 
 
 
 
 
 
 
 
% 
Slope 
Distance 
Between 
Water Bars 
(feet) 
Distance 
Between Broad-
base Dips, 
Turnouts (feet) 
3% 200 235 
5% 135 180 
10% 80 140 
15% 60 125 
20% 45  NA 
30% 35  NA 
40% 30 NA