Stephen F.

Austin State University

SFA ScholarWorks
Faculty Publications Spatial Science

2009

Integration of Modified Universal Soil Loss

Equation (MUSLE) into a GIS Framework to
Assess Soil Erosion Risk
Yanli Zhang
Arthur Temple College of Forestry and Agriculture, Stephen F. Austin State University, zhangy2@sfasu.edu

J. Degroote
University of Northern Iowa

Calvin Wolter
Iowa DNR, IGS

Ramanathan Sugumaran
University of Northern Iowa

Follow this and additional works at: http://scholarworks.sfasu.edu/spatialsci

Part of the Forest Management Commons
Tell us how this article helped you.

Recommended Citation
Zhang, Yanli; Degroote, J.; Wolter, Calvin; and Sugumaran, Ramanathan, "Integration of Modified Universal Soil Loss Equation
(MUSLE) into a GIS Framework to Assess Soil Erosion Risk" (2009). Faculty Publications. Paper 37.
http://scholarworks.sfasu.edu/spatialsci/37

This Article is brought to you for free and open access by the Spatial Science at SFA ScholarWorks. It has been accepted for inclusion in Faculty
Publications by an authorized administrator of SFA ScholarWorks. For more information, please contact cdsscholarworks@sfasu.edu.
 Land Degradation & Development

Integration of Modified Universal Soil Loss Equation

(MUSLE) into a GIS environment to assess soil erosion risk
Fo

Journal: Land Degradation & Development

Manuscript ID: LDD-08-0063.R1

r

Wiley - Manuscript type: Research Article

Pe

Date Submitted by the

n/a
Author:

Complete List of Authors: Zhang, Yanli; University of Northern Iowa, Geography

DeGroote, John; University of Northern Iowa, Geography
er

Wolter, Calvin; Iowa DNR, IGS

Sugumaran, Ramanathan; University of Northern Iowa, Geography

Keywords: MUSLE, Soil erosion, GIS , Watershed, Water

Re
vi
ew

http://mc.manuscriptcentral.com/ldd
Page 1 of 21 Land Degradation & Development

1
2
3
4
5 1 Integration of Modified Universal Soil Loss Equation
6
7 2 (MUSLE) into a GIS environment to assess soil erosion risk
8 3
9 4
10 5
11 6 Short title: Integration of MUSLE into GIS to assess soil erosion risk
12
13 7
14 8
15 9
16 10 Y. ZHANG 1*, J. DEGROOTE1, C. WOLTER2 AND R. SUGUMARAN1
17 11
18 1
12 Dept. of Geography, Univ. of Northern Iowa, Cedar Falls, IA 50614, USA
Fo
19 2
20 13 Iowa DNR/IGS, 109 Trowbridge Hall, Iowa City, IA 52242, USA
21 14
*
22 15 Correspondence to:
r
23 16 Y. Zhang, Dept. of Geography, Univ. of Northern Iowa, Cedar Falls, IA 50614, USA.
24
17 Tel: 1-319-273-3877
Pe

25
26 18 Fax: 1-319-273-7103
27 19 Email: zhangy@uni.edu
28 20
er

29 21
30 22 ABSTRACT
31
32
23 Soil erosion is an important economic and environmental concern throughout the world.
Re

33
34
35 24 In order to assess soil erosion risk and conserve water and soil resources, soil erosion
36
vi

37
38
25 modeling at the watershed scale is urgently needed. This study integrated the Modified
39
ew

40 26 Universal Soil Loss Equation (MUSLE) in a Geographic Information System (GIS)

41
42 27 framework in the form of a tool called ArcMUSLE, an extension of ArcGIS® software, to
43
44
45 28 assist soil and water conservation agencies in soil erosion risk assessment and
46
47 29 prioritization of critical areas for soil erosion control practices. With widely available
48
49 30 spatial data, this tool can be applied to determine curve numbers, to estimate runoff,
50
51
52 31 peakflow, and soil loss for a rainfall event within a watershed. An application example
53
54 32 for a watershed in Black Hawk County, Iowa, USA, is presented.
55
56 33
57
58
59
60
1
http://mc.manuscriptcentral.com/ldd
 Land Degradation & Development Page 2 of 21

1
2
3
4
5
34 KEY WORDS: MUSLE; soil erosion; GIS; watershed
6
7 35 INTRODUCTION
8
9 36 Soil erosion is a serious world wide problem due to its adverse economic and
10
11 37 environmental impacts. Soil loss in excess of soil production results in soil deterioration
12
13
14 38 and reduced productivity. Sediment and attached pollutants such as nutrients, pesticides,
15
16 39 and toxic metals can negatively impact water quality, aquatic habitat, and hydrologic
17
18
40 system. As much as 90 percent of the total loss of nitrogen and phosphorus from
Fo
19
20
21 41 agricultural watersheds is associated with sediment (Miller et al., 1988). Jones et al.
22
r
23 42 (1997) estimated that because of soil erosion the total productivity loss in United States is
24
Pe

25
26 43 more than $25 billion (US $), and off site loss is more than $17 billion, which totals
27
28 44 about $100 ha-1 per year. Pimentel et al. (1995) estimated that 90% of US cropland is
er

29
30 45 losing soil above the sustainable rate which generally is between 5-12 tons ha-1 per year
31
32
46 (metric ton). In Iowa, one of the main agricultural states in the US, with 89% of its land
Re

33
34
35 47 in cropland or pasture, half of the fertile topsoil has been lost during the last 150 years of
36
vi

37 48 cultivation and loss of topsoil continues at a rate of about 30 tons ha-1 per year (Pimentel
38
39
ew

40 49 et al., 1995). It is critical to manage soil erosion to protect water resources and maintain
41
42 50 land productivity, and ideally, areas most prone to severe erosion should be prioritized
43
44
45
51 for conservation practices or projects.
46
47 52 Soil erosion is a complex process that is related to soil properties, topography, land
48
49 53 cover, and human activities. In order to estimate soil erosion and optimize soil
50
51
52 54 conservation management, many soil erosion models have been developed. Lal (2001)
53
54 55 and Merritt et al. (2003) summarized major soil erosion models such as the Universal
55
56 56 Soil Loss Equation (USLE) and the Watershed Erosion Prediction Project (WEPP), and
57
58
59
60
2
http://mc.manuscriptcentral.com/ldd
Page 3 of 21 Land Degradation & Development

1
2
3
4
5 57 the erosion models in the Agricultural Non-Point Source model (AGNPS), the Areal
6
7 58 Non-point Source Watershed Environment Response Simulation (ANSWERS) model,
8
9 59 and the Chemical Runoff and Erosion from Agricultural Management System model
10
11
12 60 (CREAMS). Within these models, USLE and its derivatives, Revised USLE (RUSLE)
13
14 61 and Modified USLE (MUSLE), are the most widely used empirical models because of
15
16
17
62 their minimal data and computation requirements (Lal, 2001; Merritt et al., 2003; Lim et
18
Fo
19 63 al., 2005; Xu et al., 2008). The USLE and RUSLE models estimate average annual gross
20
21 64 erosion as a function of rainfall energy. In MUSLE, the rainfall energy factor is replaced
22
r
23
24 65 with a runoff factor. This improves the sediment yield prediction, eliminates the need for
Pe

25
26 66 delivery ratios, and allows the equation to be applied to individual storm events.
27
28 67 Sediment yield prediction is improved because runoff is a function of antecedent
er

29
30
31 68 moisture condition as well as rainfall energy (Williams, 1975a and 1975b; Williams and
32
Re

33 69 Berndt, 1977; Neitsch et al., 2005; Kinnell, 2005). However, the application of MUSLE
34
35
70 at a watershed scale is a sophisticated data processing procedure and requires
36
vi

37
38 71 professional GIS knowledge and technology. Therefore, we developed a user friendly
39
ew

40 72 ArcGIS® tool, named ArcMUSLE, to facilitate soil erosion modeling and soil erosion
41
42
43
73 risk assessment work at watershed scale.
44
45 74 This paper documents the development of the ArcMUSLE tool and its application to
46
47 75 a watershed for the determination or prediction of curve number, topographic factor
48
49
50 76 runoff, peak flow, sediment, and sediment delivery based on the input spatial information
51
52 77 of elevation, soil, land cover, and rainfall from a given rainfall event. The following
53
54
55
56
57
58
59
60
3
http://mc.manuscriptcentral.com/ldd
 Land Degradation & Development Page 4 of 21

1
2
3
4
5 78 sections are organized as methodology (calculation algorithms), ArcMUSLE
6
7 79 development and interfaces introduction, and application example.
8
9
80 METHODOLOGY
10
11
12 81 Soil erosion is a hydrologically driven process and it depends on sediment being
13
14 82 discharged with runoff (Kinnell, 2005). By including the runoff as an independent factor
15
16
17
83 in modeling erosion, MUSLE has an improved accuracy of soil erosion prediction over
18
Fo
19 84 USLE and RUSLE (Willams, 1975; Williams and Berndt, 1977; Erskine et al. 2002;
20
21 85 Neitsch et al., 2005; Sadeghi et al., 2007). The MUSLE equation is applicable to the
22
r
23
24 86 points where overland flow enters the streams and then all those points are summed up to
Pe

25
26 87 give the total amount of sediment delivered to the stream network within a watershed. In
27
28 88 general, MUSLE can be expressed as follows (Williams, 1975a and 1975b):
er

29
30
31 89 Y = 11.8 × (Q × qp)0.56 × K × LS × C × P (1)
32
Re

33 90 where Y is the sediment yield to the stream network in metric tons, Q is the runoff
34
35
91 volume from a given rainfall event in m3, qp is the peak flow rate in m3 s-1, K is the soil
36
vi

37
38 92 erodibility factor which is a soil property available from the Soil Survey Geographic
39
ew

40 93 (SSURGO) data, LS is the slope length and gradient factor, C is the cover management
41
42
43
94 factor and can be derived from land cover data, and P is the erosion control practice
44
45 95 factor which is a field specific value. The Q, qp, and LS parameters can be derived from
46
47 96 Digital Elevation Model (DEM), land cover, soil, and rainfall data. A detailed
48
49
50 97 explanation of the calculation algorithms for Q, qp, and LS follows.
51
52
53
54
55
56
57
58
59
60
4
http://mc.manuscriptcentral.com/ldd
Page 5 of 21 Land Degradation & Development

1
2
3
4
5 98 Calculation of Runoff Q
6
7 99 The runoff is calculated based on the widely used Soil Conservation Service (SCS, now
8
9
100 Natural Resources Conservation Service, NRCS) curve number method. This method was
10
11
12 101 developed and published in 1954 in the National Engineering Handbook Section 4:
13
14 102 Hydrology (NEH-4). Revisions of the book have been made in 1956, 1964, 1965, 1971,
15
16
17
103 1972, 1985, 1993, and 2004. As Ponce and Hawkins (1996) pointed out, the curve
18
Fo
19 104 number method owes its popularity among hydrology practitioners to its simplicity,
20
21 105 predictability, and stability. The theory of the curve number method is that for a given
22
r
23
24 106 rainfall event, the ratio of runoff depth to rainfall is equal to the ratio of actual retention
Pe

25
26 107 (the rain not converted to runoff) after runoff begins to the potential maximum retention
27
28 108 after runoff begins (NRCS, 2004). It is shown with equation 2.
er

29
30
31 Qd F
32
109 = (2)
R S
Re

33
34
35 110 where Qd is runoff depth in mm, R is the event rainfall in mm, F is the actual retention
36
vi

37 111 after runoff begins (mm), and S is the potential retention parameter which is related to the
38
39 112 soil and land cover conditions of the watershed. When initial abstraction (Ia) is
ew

40
41
42 113 considered, the amount of rainfall available for runoff is (R - Ia). Actual retention is the
43
44 114 difference between rainfall and runoff depth (F = R - Qd). An empirical relationship
45
46
115 between Ia and S was expressed as Ia = 0.2S. After substituting these three equations to
47
48
49 116 equation 2, runoff can be calculated from rainfall and potential maximum retention with
50
51 117 the equations:
52
53
54 ( R / 25.4 − 0.2S ) 2
118 Qd = 25.4 (if R ≥ 0.2S)
55 ( R / 25.4 + 0.8S )
56
57
58
59
60
5
http://mc.manuscriptcentral.com/ldd
 Land Degradation & Development Page 6 of 21

1
2
3
4
5 119 Qd = 0 (if R ≤ 0.2S) (3)
6
7 120 S is calculated from curve number (CN): S= 1000/CN – 10 in inches. In GIS, CN can be
8
9
10 121 derived from soil HSG (hydrologic soil group) classification and land cover type (NRCS,
11
12 122 1986). In other words, equation 3 means that if rainfall is less than initial abstraction,
13
14
15
123 there is no runoff, if rainfall is larger than initial abstraction, runoff depth can be
16
17 124 calculated from the rainfall amount and the curve number. After the user specifies the
18
Fo
19 125 rainfall amount, Qd will be calculated, then multiplied by the area to get runoff volume Q
20
21
22 126 for each cell.
r
23
24
Pe

25
26 127 Calculation of Peak Discharge qp
27
28 128 The graphical peak discharge method is used to compute peak discharge qp (m3 s-1)
er

29
30
31
129 (NRCS, 1986):
32
Re

33 130 q p = q u AQd F p (4)

34
35
36 131 where qu is the unit peak discharge in m3 s-1 km-2 mm-1, A is the drainage area in km2, Qd
vi

37
38 132 is the runoff depth in mm, and Fp is the pond and swamp adjustment factor which is the
39
ew

40
133 percentage of pond and swamp area over the watershed area. The time of concentration,
41
42
43 134 T, is required for the calculation of qu and is developed by calculating a travel time for
44
45 135 sheet flow and a travel time for shallow concentrated flow. Sheet flow is flow over plane
46
47
48
136 surfaces and usually occurs in the headwater of streams. The time of travel for sheet flow
49
50 137 (Ttsheet) of less than 91.4 m is calculated using Manning’s kinematic solution:
51
52
0.091(nL) 0.8
53 138 Tt sheet = (5)
54 J 0.5 s 0.4
55
56
57
58
59
60
6
http://mc.manuscriptcentral.com/ldd
Page 7 of 21 Land Degradation & Development

1
2
3
4
5 139 where the travel time is in hours, n is the Manning's roughness coefficient, L is the flow
6
7 140 path length in meters, s is the slope in percentage and J is the 2-yr, 24-hr rainfall (typical
8
9 141 24-hour duration precipitation with a 2-year return period) in mm which is published by
10
11
12 142 the National Weather Service (NWS). After a maximum of 91.4 m, sheet flow usually
13
14 143 becomes shallow concentrated flow and the time of travel for shallow concentrated flow
15
16
17
144 (Ttshallow) is calculated using the following equation for the remainder of the flow path
18
Fo
19 145 until it flows into a defined channel:
20
21 3.281 × L
22 146 Tt shallow = (6)
3600 × 16.1345 × s 0.5
r
23
24
Pe

25 147 where L is the flow path length in meter. This equation is based on Manning’s equation
26
27
148 and two assumptions for Manning’s roughness coefficient (0.05) and hydraulic radius
28
er

29
30 149 (0.4). Now the two times of travel calculations can be added together to come up with the
31
32 150 time of concentration (T) where each drainage way discharges into a stream.
Re

33
34
35 151 T = Tt sheet + Tt shallow (7)
36
vi

37 152 The unit peak discharge can then be calculated using the following equation:
38
39
ew

2
40 153 q u = 10 (C0 +C1 log T +C2 (log T ) )
(8)
41
42
43 154 Coefficients C0, C1, and C2 are available from the Urban Hydrology for Small
44
45 155 Watersheds manual (NRCS, 1986). They are determined by rainfall type and the ratio of
46
47
156 initial abstraction (Ia) and 2-yr 24-hr rainfall (P).
48
49
50 157 The peak discharge can then be calculated with eq. 4 as Qd is available from the
51
52 158 previous section and A can be derived directly from DEM data.
53
54
55
56
57
58
59
60
7
http://mc.manuscriptcentral.com/ldd
 Land Degradation & Development Page 8 of 21

1
2
3
4
5 159 Determining LS Factor
6
7 160 The LS factor reflects the effect of topography on soil erosion. Within it, L represents the
8
9
161 effect of slope length on erosion and S reflects the influence of slope gradient on erosion
10
11
12 162 (Williams, 1975a; Williams and Berndt, 1977; Lu et al., 2004). The equation (eq. 9)
13
14 163 developed by Moore and Burch (1986 a, b) is used to calculate the LS factor in
15
16
17
164 ArcMusle:
18
Fo
19 A 0.4 sin θ 1.3
20 165 LS = ( ) ×( ) (9)
22.13 0.0896
21
22
166 where A is the product of flow accumulation and cell size, θ is the slope in degrees, both
r
23
24
Pe

25 167 of which are derived from the DEM directly.

26
27
28
168 ArcMUSLE DEVELOPMENT AND APPLICATION
er

29
30 169 ArcMUSLE was developed with C# and ArcObjects® as a distributable ArcGIS®
31
32 170 extension. It is an integration of the MUSLE algorithm with up-to-date GIS technology.
Re

33
34
35
171 An overview of the data analysis process is shown in Figure 1. It is a raster based spatial
36
vi

37 172 analysis applying all the mathematic equations presented in the Algorithms section.
38
39 173 Runoff is derived from soil data, land cover, and event rainfall. Peak flow is derived from
ew

40
41
42 174 data on land cover, DEM, and the previously calculated runoff. The LS factor is derived
43
44 175 from the DEM. Soil erodibility is a soil property available in SSURGO soil data. The
45
46 176 cover management factor is set based on land cover. The practice factor is a field specific
47
48
49 177 value. MUSLE is finally applied to spatially calculate the soil erosion amount for a given
50
51 178 rainfall event. User friendly interfaces were designed to facilitate ArcMUSLE usage and
52
53
54
179 Figure 2 shows the three consecutive interfaces. In the first interface, the user needs to
55
56 180 define the analysis environment which includes the watershed boundary, analysis
57
58
59
60
8
http://mc.manuscriptcentral.com/ldd
Page 9 of 21 Land Degradation & Development

1
2
3
4
5 181 resolution, and working directory. The second user interface is used to choose input
6
7 182 spatial data layers such as the DEM, soil, land cover, and 2-year, 24-hour rainfall data
8
9 183 and corresponding attribute fields. On the last user interface which is for parameters
10
11
12 184 setting, the user can designate the antecedent soil condition (dry, normal, or wet), the
13
14 185 amount of analysis rainfall, and drainage network size. As the P factor value is field
15
16
17
186 specific, a P factor raster layer can be used or a default value of one will be used if field
18
Fo
19 187 specific data are not available.
20
21 188 As mentioned previously, DEM, land cover, soil K factor, soil hydrologic group, and
22
r
23
24 189 2-year, 24-hour rainfall data are the necessary spatial input data for the model. The Iowa
Pe

25
26 190 Department of Natural Resources (DNR) compiled all of these spatial data in raster data
27
28 191 format for the entire state with a resolution of 30 meters and provided them to the public.
er

29
30
31 192 Land cover data include attributes on land cover category, C factor, and Manning
32
Re

33 193 roughness coefficient n. The C factor value is derived from land cover type. Soil data
34
35
194 include attributes on slope, K factor, and HGC. For Iowa soil erosion analysis, users only
36
vi

37
38 195 need to define a watershed boundary and the model can be easily run within ArcGIS®.
39
ew

40
41 196 ArcMUSLE was not compared with measured data for validation purpose in this
42
43 197 paper because MUSLE has been widely used and validated in many countries (Williams
44
45 198 1975b; Johnson et al. 1985; Loch et al. 2000; Erskine et al. 2002; Blaszczynski 2003;
46
47
48 199 Lim et al. 2005; Sadeghi et al. 2007). Other equations or algorithms including the curve
49
50 200 number method are from NRCS official documents (1986, 2004) and have been used
51
52
53
201 since 1975 (Ponce and Hawkins 1996; Blaszczynski 2003; Zhan and Huang 2004; Mishra
54
55 202 et al. 2006). Instead, ArcMUSLE was applied to a watershed located in Black Hawk
56
57
58
59
60
9
http://mc.manuscriptcentral.com/ldd
 Land Degradation & Development Page 10 of 21

1
2
3
4
5 203 County, Iowa, with an area of 24.2 km2, to demonstrate how the tool can be easily used
6
7 204 for effective soil erosion management. According to the Iowa DNR land cover
8
9 205 classification of 2002, the majority of the watershed is agriculture land (30% hay, 63%
10
11
12 206 corn or bean production). The Iowa DNR has listed the watershed as an impaired water
13
14 207 resource. In order to assist soil and water conservation projects within the watershed, it is
15
16
17
208 necessary to analyze the soil erosion risk and identify critical areas. All the input spatial
18
Fo
19 209 data are shown in Figure 3. Soil antecedent moisture condition was set as average and the
20
21 210 rainfall was set to the typical 24-hour precipitation for two year return period, i.e., 80.3
22
r
23
24 211 mm (3.16 inches). After running the model using this specified rainfall amount, the total
Pe

25
26 212 amount of predicted sediment delivered to waterbodies was 6,669 tons and the total
27
28 213 runoff volume was 765,370 m3 for the whole watershed.
er

29
30
31 214 Figure 4 illustrates the intermediate data products and analysis results (LS factor,
32
Re

33
34 215 curve number, runoff depth, peak flow, sediment, and sediment delivery) for this
35
36 216 hypothetical precipitation event. The relatively steep areas, in the upper part of the
vi

37
38 217 watershed, have high LS values. The curve number is based on land cover and hydrologic
39
ew

40
41 218 soil group. When compared with forest and grass lands, lands in corn and soybean
42
43 219 production generally have less evapotranspiration and water storage, and thus more run
44
45 220 off. Generally, runoff potential increases from hydrologic soil group A (sand, loamy
46
47
48 221 sand, sandy loam) to D (clay loam, silty clay loam, sandy clay, silty clay or clay). The
49
50 222 runoff depth map shows the overland flow which is calculated from the curve number
51
52
53
223 and rainfall amount. The rainfall amount is a constant in this analysis. The pattern in the
54
55 224 curve number and runoff depth maps match well with land cover and soil maps. There are
56
57
58
59
60
10
http://mc.manuscriptcentral.com/ldd
Page 11 of 21 Land Degradation & Development

1
2
3
4
5 225 higher curve number values and higher runoff depth for corn and soybean land. Lands
6
7 226 with hydrologic group D soil have the highest curve number values and highest runoff
8
9 227 depth. Peak flow is derived from six parameters including runoff depth, land cover, 2-
10
11
12 228 year 24-hour rainfall, flow length, drainage area, and slope. The sediment map shows
13
14 229 where soil sediment is eroded, the darker areas indicate larger amount of sediment
15
16
17
230 created. When comparing the sediment map with LS factor and runoff depth, in general,
18
Fo
19 231 the areas with the highest sediment yield are those having high LS factor values and
20
21 232 deeper runoff. There are areas that have high LS factor values, such as in the upper third
22
r
23
24 233 of the watershed, which have shallow runoff depth as the result of having hydrologic soil
Pe

25
26 234 group B and grass as the land cover. After field verification, corresponding Best
27
28 235 Management Practices (BMPs) such as contour farming, filter strips, or conservation
er

29
30
31 236 tillage can be applied to those areas. The sediment delivery map shows the amount of
32
Re

33 237 sediment entering waterbodies, the darker areas indicate larger amount. BMPs such as
34
35
238 sediment ponds or wetland restoration to trap sediment before the overland flow enters
36
vi

37
38 239 waterbodies can be applied in these areas. As a conclusion, soil erosion critical areas are
39
ew

40 240 identified in the sediment map and sediment delivery maps. The application of suitable
41
42
43
241 BMPs to these areas will improve the efficiency of soil and water conservation projects.
44
45
242 SUMMARY
46
47
48 243 The newly developed ArcMUSLE soil erosion modeling tool and its application to a
49
50
51 244 watershed in Iowa are reported to introduce this tool to the water and soil conservation
52
53 245 community. ArcMUSLE applies the MUSLE equation, curve number, and graphical peak
54
55 246 discharge methods within an advanced GIS analysis environment. Outputs of the tool
56
57
58
59
60
11
http://mc.manuscriptcentral.com/ldd
 Land Degradation & Development Page 12 of 21

1
2
3
4
5 247 include curve number, runoff, peakflow, and soil loss for a rainfall event within a
6
7 248 watershed. It can assist the prioritization of critical soil erosion areas and the
8
9 249 improvement of water and soil conservation efforts. The ArcMUSLE toolset was
10
11
12 250 originally designed for use in Iowa but with proper input data preprocessing it could be
13
14 251 applied in other areas. Further development will include improvement of functionality,
15
16
17
252 such as a curve number setting model that allows the user to specify CN’s for different
18
Fo
19 253 land cover categories.
20
21
22 254 ACKNOWLEDGEMENTS
r
23
24 255 This project was funded by the US National Aeronautics and Space Agency (NASA) at
Pe

25
26
27 256 the GeoTREE Center of University of Northern Iowa (grant #NNX06AC69G). Iowa
28
er

29 257 DNR provided all the spatial data. The authors appreciate the technical support from
30
31 258 Chris Ensminger and Adam Kiel, Iowa DNR. Also the authors would like to thank two
32
Re

33
34 259 anonymous reviewers for helping improving this paper
35
36 260
vi

37
38 261 REFERENCES
39
ew

40 262 Blaszczynski J. 2003. Estimating watershed runoff and sediment yield using a GIS
41
42
43 263 interface to curve number and MUSLE models. Soils and Geology, Resources Notes
44
45 264 No. 66, National Science and Technology Center, Denver, Colorado, US.
46
47
48
265 Erskine WD, Mahmoudzadeh A, Myers C. 2002. Land use effects on sediment yields and
49
50 266 soil loss rates in small basins of Triassic sandstone near Sydney, NSW, Australia.
51
52 267 Catena 49: 271-287.
53
54
55
56
57
58
59
60
12
http://mc.manuscriptcentral.com/ldd
Page 13 of 21 Land Degradation & Development

1
2
3
4
5 268 Johnson CW, Gordon ND, Hanson CL. 1985. Northwest Rangeland sediment yield
6
7 269 analysis by the MUSLE. Transactions of the American Society of Agricultural
8
9 270 Engineers 28: 1889-1895.
10
11
12 271 Jones AJ, Lal R, Huggins DR. 1997. Soil erosion and productivity research: A regional
13
14 272 approach. American Journal of Alternative Agriculture 12: 185-192.
15
16
17
273 Kinnell P. 2005. Why the universal soil loss equation and the revised version of it do not
18
Fo
19 274 predict event erosion well. Hydrological Process 19: 851–854. DOI:
20
21 275 10.1002/hyp.5816
22
r
23
24 276 Lal R. 2001. Soil degradation by erosion. Land Degradation & Development 12: 519-
Pe

25
26 277 539. DOI: 10.1002/ldr.472.
27
28 278 Lim JK, Sagong M, Engel BA, Tang Z, Choi J, Kim K. 2005. GIS based sediment
er

29
30
31 279 assessment tool. Catena 64: 61-80.
32
Re

33 280 Loch RJ, Connolly RD, Littleboy M. 2000. Using rainfall simulation to guide planning
34
35
281 and management of rehabilitated areas: part II computer simulations using
36
vi

37
38 282 parameters from rainfall simulation. Land Degradation & Development 11: 241-255.
39
ew

40 283 Lu D, Li G, Valladares G, Batistella M. 2004. Mapping soil erosion risk in Rondonia,

41
42
43
284 Brazilian Amazonia: using RUSLE, remote sensing and GIS. Land Degradation &
44
45 285 Development 15: 499–512. DOI: 10.1002/ldr.634.
46
47 286 Merritt WS, Letcher RA, Jakeman AJ. 2003. A review of erosion and sediment transport
48
49
50 287 models. Environmental Modeling and Software 18: 761-799. DOI: 10.1016/S1364-
51
52 288 8152(03)00078-1
53
54
55
56
57
58
59
60
13
http://mc.manuscriptcentral.com/ldd
 Land Degradation & Development Page 14 of 21

1
2
3
4
5 289 Mishra SK, Tyagi JV, Singh VP, Singh R. 2006. SCS-CN-based modeling of sediment
6
7 290 yield. Journal of Hydrology 324:301-322. DOI: 10.1016/j.jhydrol.2005.10.006
8
9 291 Miller GA, Amemiya M, Jolly RW, Melvin SW, Nowak PJ. 1988. Soil erosion and the
10
11
12 292 Iowa soil 2000 program. Iowa State University Pm-1056. 8 pages.
13
14 293 Moore I, Burch G. 1986a. Physical basis of the length-slope factor in the Universal Soil
15
16
17
294 Loss Equation. Soil Science Society of America Journal 50: 1294-1298.
18
Fo
19 295 Moore I, Burch G. 1986b. Modeling erosion and deposition: topographic effects.
20
21 296 Transactions of American Society of Agricultural Engineers 29: 1624-1630, 1640.
22
r
23
24 297 Neitsch SL, Arnold JG, Kiniry JR, Williams JR, King KW. 2005. SWAT theoretical
Pe

25
26 298 documentation, Grassland, Soil and Water Research Laboratory. 494 pages.
27
28 299 Natural Resources Conservation Services (NRCS). 1986. Urban hydrology for small
er

29
30
31 300 watersheds, Technical Release 55, 210-VI-TR-55, 728. US Department of
32
Re

33 301 Agriculture, Washington DC, US. 164 pages.

34
35
302 NRCS. 2004. National Engineering Handbook. Part 630 Hydrology, 210-VI-NEH-
36
vi

37
38 303 630.10. US Department of Agriculture, Washington DC.
39
ew

40 304 Pimentel D, Harvey C, Resosudarmo K, Sinclair K, Kurz D, McNair M, Crist S, Shpritz

41
42
43
305 L, Fitton L, Saffouri R, Blair R. 1995. Environmental and economic costs of soil
44
45 306 erosion and conservation benefits. Science 267: 1117–1123.
46
47 307 Ponce VM, Hawkins RH. 1996. Runoff curve number: has it reached maturity? Journal
48
49
50 308 of Hydrologic Engineering 1: 11-19.
51
52
53
54
55
56
57
58
59
60
14
http://mc.manuscriptcentral.com/ldd
Page 15 of 21 Land Degradation & Development

1
2
3
4
5 309 Sadeghi SHR, Mizuyama T, Miyata S, Gomi T, Kosugi K, Mizugaki S, Onda Y. 2007. Is
6
7 310 MUSLE apt to small steeply reforested watershed? Journal of Forest Resources 12:
8
9 311 270-277. DOI: 10.1007/s10310-007-0017-9
10
11
12 312 Williams JR. 1975a. Sediment routing for agricultural watersheds. Water Resource
13
14 313 Bulletin 11: 965-974.
15
16
17
314 Williams JR. 1975b. Sediment-yield prediction with universal equation using runoff
18
Fo
19 315 energy factor. In Present and prospective technology for predicting sediment yield
20
21 316 and sources. ARS.S-40, US Gov. Print. Office, Washington, DC. 244-252.
22
r
23
24 317 Williams JR, Berndt HD. 1977. Sediment yield prediction based on watershed hydrology.
Pe

25
26 318 Transactions of the American Society of Agricultural Engineers 20:1100-1104.
27
28 319 Xu Y, Shao X, Kong X, Peng J, Cai Y. 2008. Adapting the RUSLE and GIS to model soil
er

29
30
31 320 erosion risk in a mountains karst watershed, Guizhou Province, China
32
Re

33 321 Environmental Monitoring and Assessment 141: 275-286. DOI: 10.1007/s10661-

34
35
322 007-9894-9
36
vi

37
38 323 Zhan X, Huang M. 2004. ArcCN-Runoff: an ArcGIS tool for generating curve number
39
ew

40 324 and runoff maps. Environment Modelling & Software 19: 875-879. DOI:
41
42
43
325 10.1016/j.envsoft.2004.03.001
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
15
http://mc.manuscriptcentral.com/ldd
 Land Degradation & Development Page 16 of 21

1
2
3
4
5 326 Figure Captions
6
7 327
8
9 328 Figure 1 Flow chart of the ArcMUSLE calculation process
10
11 329
12
13 330 Figure 2 ArcMUSLE’s graphical user interfaces. The first is used to set the analysis
14
15 331 environment, the second is used to set input layers, and the third one is used to set
16
17 332 analysis parameters.
18
Fo
19
20 333
21
22 334 Figure 3 Study watershed location and input data. (Clockwise: location map, land cover,
r
23
24 335 2-year, 24-hour rainfall, hydrologic soil group (HSG), soil K factor, and DEM.
Pe

25
26
27 336
28
er

29 337 Figure 4 Intermediate data products and analysis results of ArcMUSLE (clockwise: LS
30
31
32
338 factor, CN, runoff depth, peakflow, sediment, and sediment delivery).
Re

33
34 339
35
36
vi

37
38
39
ew

40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
16
http://mc.manuscriptcentral.com/ldd
Page 17 of 21 Land Degradation & Development

1
2
3
4 Soil
5 hydro- Curve Runoff (Q0.56)
6 number
7
group
8 ×
9 Land Event Peak flow (qp0.56)
10 cover rainfall
11 ×
12 Manning
13 coeff.. Soil erodibility (K)
14
15 2-yr ×
16
24-hr Topographic factor (LS)
17
18 rainfall
×
Fo
19
20 Flow Cover management factor(C)
21 length
22
×
r
23
24 Drainage Practice factor (P)
Pe

25 area
26 ×
27
DEM Slope 11.8
28
er

29
30 Flow =
31 accu. Rain event soil loss map
32
Re

33
34
35 Figure 1 Flow chart of the ArcMUSLE calculation process
36
vi

37
38
39
ew

40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
http://mc.manuscriptcentral.com/ldd
 Land Degradation & Development Page 18 of 21

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Fo
19
20
21
22
r
23
24
Pe

25
26 A
27
28
er

29
30
31
32
Re

33
34
35
36
vi

37
38
39
ew

40
41
42
43
44
45
46
47
48
49
50 B
51
52
53
54
55
56
57
58
59
60
http://mc.manuscriptcentral.com/ldd
Page 19 of 21 Land Degradation & Development

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Fo
19
20
21
22
r
23
24
Pe

25
C
26
27
28
er

29 Figure 1 ArcMUSLE’s graphical user interfaces. The first is used to set the analysis
30
31 environment, the second is used to set input layers, and the third one is used to set
32
Re

33
34 analysis parameters.
35
36
vi

37
38
39
ew

40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
http://mc.manuscriptcentral.com/ldd
 Land Degradation & Development Page 20 of 21

1
2
3
4
± ± ±
5
Land cover
6
Water
7
Forest
8
Grass
9 2-yr 24-hr rainfall
Corn
10
Beans 80.3 mm

Fo
11
Artifical/urban
12
Barren
13 0 0.5 1 2 Km
0 0.5 1 2 Km

rP
14 Study watershed, Iowa, U.S.A.
15
16
17
18
19 ee
rR
20

± ± ±
21
22
23
24
25 HSG
ev K factor
0

iew
26 A DEM
27 0.15
28 B 0.17 312 m
29 C 0.2
30 0.24 265 m
D
31
0.26
32
33 0 0.5 1 2 Km 0 0.5 1 2 Km 0 0.5 1 2 Km
34
35
36
37
38
39 Figure 3 Study watershed location and input data. Clockwise:
40
41 location map, land cover, 2-year, 24-hour rainfall, hydrologic soil
42 group (HSG), soil K factor, and DEM.
43
44
45
46 http://mc.manuscriptcentral.com/ldd
47
48
 ± ± ±
Page 21 of 21 Land Degradation & Development
0 0.5 1 2 km 0 0.5 1 2 km 0 0.5 1 2 km
1
2
3
4
5
6
7
8
9
LS factor CN
10 Runoff depth (mm)

Fo
11 92
1.97 58.7
12
13

rP
14
15 0 0 0
16
17
18
19 ee
rR
20

± ± ±
21
22 0 0.5 1 2 km 0 0.5 1 2 km 0 0.5 1 2 km
23
24
25
ev
iew
26
27
28
29
30
31 Peakflow (cms) Sediment delivery (t)
32 Sediment (t)
33
3.62 0 - 10
1.832
34 10.1 - 30
35
36 0 30.1 - 60
0
37 60.1 - 175.4
38
39
40
41 Figure 4 Intermediate data products and analysis results of ArcMUSLE
42
43 (clockwise: LS factor, CN, runoff depth, peakflow, sediment, and sediment delivery).
44
45
46 http://mc.manuscriptcentral.com/ldd
47
48