Proceedings
Integration of GIS-Rusle-Sedd Mod for Predicting Sediment Yield in Limboto Watershed
– Muhammad Ramdhan Olii, Ilyas Ichsan
6th International Seminar of HATHI
CASE STUDY
INTEGRATION OF GIS-RUSLE-SEDD MOD
FOR PREDICTING SEDIMENT YIELD
IN LIMBOTO WATERSHED
Muhammad Ramdhan Olii*, and Ilyas Ichsan
Departement of of Civil Engineering, Universitas Gorontalo
*[email protected]
Abstract
The limited information related to the amount erosion and sediment in watershed
scale causes the stakeholders did not need effective and appropriate steps to reduce
the amount of erosion and sediment yield that occurs. The deciding constraint is the
budget limitation in erosion and the direct sediment yield in the field. The most
popular method used in various parts of the world to overcome this was to integrate
GIS, erosion models and SDR models to predict sediment yield on a watershed
scale. RUSLE model is used to analyze the amount of erosion that occurs with the
grid by overlay all of soil erosion factors. The SEDD-Mod model was used to
analyze the ability of a watershed to deposit and transport eroded soils which will
then be the result of sediment yield. Integrating RUSLE, SEDD-Mod and GIS
allows for relatively easy, fast, and cost-effective estimation of spatially distributed
and amount erosion and sediment yield. The study provides a useful procedure that
can be used in watershed scale were data on sediment yield measurements for long
periods is lacking like Limboto Watershed. The cultivated lands exhibit much
greater sediment yield than the non-cultivated lands in Limboto Watershed.
Keywords: GIS, RUSLE, SEDD-Mod, sediment yield, Limboto Watershed.
INTRODUCTION
Soil erosion models namely Revised Universal Soil Loss Equation (RUSLE)
(Renard et al. 1997) estimate gross soil erosion rate at plot-scale. Soil erosion rates
predicted by RUSLE are often higher than those measured at watershed outlets.
Sediment Delivery Ratio (SDR) is used to correct for this reduction effect. SDR is
sediment transport efficiency, which estimation for the total of sediment yield that
is actually transported from the eroding sources to watershed outlet compared to
the total amount of soil erosion that is detached over the same area above that point
(Mutua and Klik, 2006). In relatively large watersheds, most sediment gets
deposited within the watershed and only slightly a fraction of the soil that is eroded
from the hillslope reaches the watershed outlet.
The integrated use of Geographic Information System (GIS) and SDR models has
been shown to be an effective approach to predicting the magnitude and distribution
of SDR value. Spatially distributed SDR value in a watershed scale has been
modeled by Ferro and Porto (2000) based on travel time concept eg. Sediment
Delivery Distributed (SEDD) model. This concept was integrated into a GIS by Jain
Kupang,
244 22-24 November 2019 1
Kupang, 22-24 November 2019
HATHI the 6th International Seminar
6th International Seminar of HATHI
and Kothyari (2000) due to it allows easy definition of spatial unit land as a grid of
relatively uniform properties. The identification of the spatially distributed soil
erosion, SDR and sediment yield makes possible the implementation of special
conservation practices on these source areas. Olii (2018) modified the SEDD Model
into SEDD-Mod model with several different variables, namely Curve Number
value, silt ratio and watershed-specific parameter, where those variables were
considered more representative of conditions in the field.
The purpose of this research is to present a methodology that integrates GIS with
RUSLE and SEDD-Mod model to predict the spatial distribution of soil erosion and
sedment yield at a large watershed, and to demonstrate the use of this methodology
by applying it to Limboto Watershed.
METHODOLOGY
Materials
To achieve the objective of study, the following datasets were collected: (1) Daily
precipitation of 6 rain-gauge stations, collected from Meteorological,
Climatological, and Geophysical Agency of Gorontalo. (2) 120 soil samples from
field were used to generate the soil fraction namely sand, silt and clay. (3) A Shuttle
Radar Topography Mission DEM with a spatial resolution of 30 m of the area was
downloaded from https://earthexplorer.usgs.gov. (4) Landuse map collected from
Watershed and Protection Forest Management Office of Bone Bolango and
download from http://tanahair.indonesia.go.id.
Methods
Soil Erosion Model
Revised Universal Soil Loss Equation (RUSLE) (Renard et al., 1997) is an
empirically based model developed on the Universal Soil Loss Equation (USLE)
(Wischmeier and Smith, 1978). RUSLE model enables prediction of an average
annual rate of soil erosion for a site of interest for any number of scenarios involving
cropping systems, management techniques, and erosion control practices. In the
application of RUSLE on GIS environment, soil loss is determined within the raster
GIS. Raster models are grid-based representations of map features, which offer
analytical capabilities for continuous data and allow fast processing of map layer
overlay operations. The average annual soil loss was found in RUSLE by
multiplying the six factors indicated in the following equation:
A R K LS C P Ga (1)
where :
A : the average annual soil loss (tons/year)
R : the erosivity factor (MJ mm/ha h year)
K : the erodibility factor (ton h/MJ mm)
LS : the slope length factor
C : the cover management factor
P : the support practice factor
Ga : the grid area (ha)
Kupang, 22-24November
Kupang, 22-24 November 2019
2019 2452
Integration of GIS-Rusle-Sedd Mod for Predicting Sediment Yield in Limboto Watershed
6th International Seminar of HATHI
In Indonesia the number of automatic rainfall recorders is limited and the erodibility
index has to be approximated. A generally accepted and widely used approximation
of the R factor is the Bols index (Bols (1978) in Vis, 1987). The formula developed
by Bols has the following form:
R 6,119 Pm H 0,47 Pmaks 0,53
1,21
(2)
where :
R : the mean monthly rainfall erosivity index unit (MJ mm ha-1 h-1 year-1)
Pm : the monthly rainfall (mm),
Ni
Pmax : the maximum-daily-rainfall in a month (mm).
The soil erodibility (K) is the rate of soil loss per rainfall erosivity index unit as
measured on a standard plot and often determined using inherent soil properties.
The K factor was calculated by means of the following formula which were
developed from global data of measured K values, obtained from 225 soil classes
(Renard et al. 1997).
1 Log Dg 1, 659 2
K 7,594 0, 0034 0, 0405exp (3)
2 0, 7101
Dg exp 0,01 f r ln mr (4)
where:
fr : the particle size fraction in percent of class i;
mr : the arithmetic mean of the particle size limits of that size.
The slope length factor (LS) is defined as a ratio of soil loss under given conditions
to that at a site with the ‘‘standard’’ slope steepness of 9% and slope length of 22.13
m (Renard et al., 1997). It has been demonstrated that increases in slope length and
slope steepness can generate higher overland flow velocities and correspondingly
higher soil erosion (Haan et al., 1994). The LS factor was calculated using the
following relationship (McCool et al., 1987).
m
L 10.8sin 0.03 for slope percent < 9% (5)
22.13
m
L 16.8sin 0.50 for slope percent ≥ 9% (6)
22.13
m
sin 0.0896 2.96 sin 0.79
0.56 (7)
1 sin 0.0896 2.96 sin 0.56
0.79
where:
m : a variable slope length exponent;
λ : the slope length (m)
θ : the slope angle (o)
Kupang,
246 22-24 November 2019 3
Kupang, 22-24 November 2019
HATHI the 6th International Seminar
6th International Seminar of HATHI
The crop management factor (C) expresses the effect of cropping and management
practices on the soil loss rate (Renard et al. 1997). The C-factor for a particular
landuse type is the weighted average of those soil loss ratios and ranges between 0
and 1. The C factor for each land use (Table 1) was assigned based on the Minister
of Forestry Regulations Number: P. 32 / MENHUT-II / 2009.
Table 1. Cover management factor (C) values for different land use classes of the
study area
Landuse C
Irrigated rice 0.01
Maize 0.64
Bare soil 1.0
Estate crops, good ground cover 0.1
Estate crops, poor ground cover 0.5
Shrub land, undisturbed 0.01
Forest, undisturbed, sparse litter 0.005
Forest, undisturbed, good litter 0.001
The P factor is a supporting practice factor that reflects the effects of practices that
will reduce the amount and rate of the water runoff and thus reduces the amount of
erosion, the higher the supporting practice, the lower the value of the P factor. In
this study, P factor is related to the land use identified by land use types (agriculture
and non-agriculture) and slope (Farhan, et al., 2013).
Table 2. Support practices factor (P) value based on land use and slope
Landuse Slope (%) P Factor
Agriculture 0–5 0,10
5 – 10 0,12
10 – 20 0,14
20 – 30 0,19
30 - 50 0,25
50 - 100 0,33
Non- Agriculture all slope 1.00
Sediment DElivery Distributed (SEDD) Model
SDRi, the fraction of the gross soil erosion from grid i-th that actually reaches a continuous
river system, is determined following (Ferro and Minacapilli 1995) as a function of travel
time. Olii (2018) modified SEDD Model into SEDD-Modification Model. This is
expressed as:
li 1
SDRSEDD exp 1,6
(8)
Modi
s CN Sri 0,25
i i
The watershed -specific parameter β depends primarily on watershed
morphological data (Olii, 2018) and can be estimated with Equation 9.
1,6
L
0.00002 D 0.00002 u
1,6
d
(9)
Ac
Kupang, 22-24November
Kupang, 22-24 November 2019
2019 2474
Integration of GIS-Rusle-Sedd Mod for Predicting Sediment Yield in Limboto Watershed
6th International Seminar of HATHI
where:
Dd : drainage density (m/m2)
β : watershed-specific parameter
Ac : watershed t area (m2)
Lu : total stream length (m)
li : flow length from grid i-th to the nearest river (m)
si : slope of the grid (m/m)
CNi : curve number of the grid
Sri : the percent ratio distribution of the silt fraction to the percentage of clay
and sand fraction of the grid
Sediment Yield
Sediment yield is the gross (total) erosion minus the sediment deposited en route to
the point of concern. To Identify the major source areas of sediment reaching the
stream network, the soil erosion and SDR coverages are overlaid. The sediment
yield from each grid is computed as follows.
n
SY A SDR
i 1
i i (10)
where:
i : i-th grid
SY : the average annual sediment yield (tons/year)
Eai : the average annual soil erosion (tons/year)
SDR : sediment delivery ratio (-)
RESULT AND DISCUSSION
Soil Erosion
The all factors of soil erosion were discretized into 30 x 30 m2 (resolution of the
DEM) grid cells, which were presumed to exhibit homogeneous properties. By
using the values of the R factor estimated for each investigated rain-gauge (6 rain
gauges) and a IDW interpolation method, the is erosivity map of the study area is
produced as is shown in Fig. 1. The high values were along mountainous areas.
Statistically, the R factor shows that the value of R factor ranges from 1430.97 to
1944.56 MJ mm/ha h year. The mean value of R factor is observed to be 1549.81
MJ mm/ha year, and the standard deviation is 93.82 which implies that the rainfall
in the study area was unstable and obviously fluctuated. By using the values of the
K factor estimated for 120 soil samples by using the soil fraction obtained from
laboratory analysis and an IDW interpolation method. The spatial distribution of K
factor is given in Fig. 1. Soil erodibility factor ranged from 0.013 to 0.044 ton h/MJ
mm. The mean value of K factor is 0.032 ton h/MJ mm, and the standard deviation
is 0.0035. LS factor was directly related to topography relief. The high values were
along the stream courses and in the mountainous areas (Fig. 1). The average value
for this area was 27.25 with a large standard deviation of 36.05 which indicates the
LS values were relatively unevenly distributed. The high values of C factor were
Kupang,
248 22-24 November 2019 5
Kupang, 22-24 November 2019
HATHI the 6th International Seminar
6th International Seminar of HATHI
scattered on the low vegetation-covered area, such as lake, river, cultivated land,
and settlement. The average value was 0.2 with a standard deviation of 0.13. Owing
large area of forest and grassland, the average value was lower. The P factor (Fig.
1) was computed from the slope and land use. The slope map (%) was prepared
using DEM in ARC GIS and it was overlain with land use map using ‘intersect
tool’. The P factor in the Limboto Watershed vary from 0.12 to 1, and the mean
value is 0.67.
Figure 1. The spatial distribution of soil erosion factor and average annual soil
erosion in Limboto Watershed
The average annual soil erosion in the Limboto Watershed was multiplied by map
algebra tool the six RUSLE factor values in ArcGIS. The soil erosion map of the
study area is shown in Fig. 1. As the study revealed, the mean soil erosion amount
in the study area was 4 tons/year. Annual soil loss ranged from 0 to 789 tons/year
in the study area. As seen from Fig. 1, the majority of the study area experiences
Kupang, 22-24November
Kupang, 22-24 November 2019
2019 2496
Integration of GIS-Rusle-Sedd Mod for Predicting Sediment Yield in Limboto Watershed
6th International Seminar of HATHI
soil erosion between 0 and 5 tons/year, especially where the slope is very low and
forest. However, excessive soil erosion is observed in northern and southern parts
of the watershed. In these regions, soil loss is more than 5 tons/year. Only very
small areas in the study area experience severe soil erosion, which is in excess of
50 tons/year.
Sediment Delivery Ratio
The SDR averaged for all grids in the Limboto Watershed was 0.11. The SDR
spatial distribution (Fig. 2) showed a relationship between the grid distance and the
nearest stream (<500 m). This is easily explained by Eq. 8), which assumes that
SDR has an inverse relationship with flow length to nearest stream, surface
roughness, soil characteristics. Therefore, the same distance does not imply that
they will have the same SDR because they may have different surface roughness
and overland slope. Randhir, et al. (2001) found from their research that longer
travel times of runoff inclined to occur in areas with rougher surfaces (vegetated
areas) compared to areas with impervious and open land surfaces. The SDR spatial
distribution based on grid is very important for identifying the critical sediment
source and delivery areas as well as the site-specific implementation of proper
management practices within a watershed (Fernandez, et al., 2003). Dai and Tan
(1996) note that the SDR values imply the integrated capability of a watershed for
storing and transporting the eroded soil.
Unlike soil loss, the SDR values obtained for the Limboto Watershed did not exhibit
a clear relation with land use type (Figure 2). This result may be explained by the
argument that SDR tends to be affected more by the character of the drainage
system than by land use type (Novotny and Chesters, 1989). However, different
land use show distinctness in average SDR (Table 3). This might be because the
different land use have different roughness and distribute at specific locations, with
the latter producing different overland slopes and distances to stream channel (Fu,
et al., 2009).
Sediment Yield
The average annual sediment yield for the Limboto Watershed, calculated as an
average of the sediment yields from all the grids to its nearest channel cell, was 0.29
tons/year and total average of the sediment yields was 305.284 tons/year. It includes
sheet and rill erosion minus channel-type erosion (gullies, valley trenches,
streambank erosion, etc.). The spatial variation of the sediment yield across the
entire watershed is shown in Figure 2. The sources of sediment in the watershed
coincide well with agricultural and steep areas. About 90.37 % of the sediment yield
reaching the Limboto Lake was produced in the dryland farming and plantation
(Table 3). This means that agricultural land was still the most important area
producing sediment yield in the Limboto Watershed.
The second important source of sediment yield to the stream was limited production
forest and bare land. It contributed about 4.74 % of the entire sediment yield with
an average of 3.45 tons/year. This might be explained by the fact that it is located
near the stream channel and the fact that it has steeper slopes.
Kupang,
250 22-24 November 2019 7
Kupang, 22-24 November 2019
HATHI the 6th International Seminar
6th International Seminar of HATHI
Figure 2. The spatial distribution of SEDD-Mod factor; SDR; and average annual
sediment yield in Limboto Watershed
Table 3. Soil erosion, sediment yield, and SDR for different land use types
Soil Erosion Sediment Yield
Landuse SDR
(tons/year) (tons/year)
Production Forest 4,482 0.14 332
Limited Production Forest 232,958 0.12 12,811
Limited Production Forest Convertible 313 0.24 50
Dryland Farming 1,735,239 0.13 135,014
Shrubs 168,857 0.15 13,547
Jungle 6,437 0.09 204
Plantation 2,446,941 0.11 140,857
Bare Land 22,636 0.17 1,660
Nature Reserve Forest / Nature Conservation Area 268 0.01 1
Protected Forest 14,857 0.09 644
Rice Field 2,747 0.09 165
Total 4.635.736 - 305.284
Kupang, 22-24November
Kupang, 22-24 November 2019
2019 2518
Integration of GIS-Rusle-Sedd Mod for Predicting Sediment Yield in Limboto Watershed
6th International Seminar of HATHI
CONCLUSION AND RECOMMENDATION
The Integrating of GIS, soil erosion and SDR models is an efficient procedure for
determining the spatial distribution of sediment yield under a variety of simulation
scenarios. The study provides a useful procedure that can be used in large watershed
scale were data on sediment yield measurements for long periods is lacking and
limited time and cost.
In the case study of the Limboto Watershed, the cultivated lands eg. the dryland
farming and plantation exhibit much greater soil erosion rates and sediment yield
than the non-cultivated lands. They contribute >90% of the total sediment yield. A
reduction in soil erosion up to 30% can be expected when combined soil erosion
control practices are implemented.
ACKNOWLEDGEMENT
This work is part of a research project funded by Beginner Lecturer Research Grant
in 2019 of Ministry of Research, Technology, and Higher Education, Indonesia .
REFERENCES
Dai, D. and Tan. Y, 1996. Soil erosion and sediment yield in the Upper Yangtze
River Basin, Proceedings of the Exeter Symposium Erosion and Sediment
Yield Global and Regional Perspectives, International Association of
Hydrological Sciences Publication, 236; 191-203.
Farhan, Y., Zregat, D., and Farhan, I., 2013, Spatial Estimation of Soil Erosion Risk
Using RUSLE Approach ,RS , and GIS Techniques : A Case Study of
Kufranja Watershed, Northern Jordan, Journal of Water Resource and
Protection, 5:1247–1261.
Fernandez, C., Wu, J., McCool, D., and Stoeckle, C., 2003, Estimating Water
Erosion and Sediment Yield with GIS, RUSLE, and SEDD, Journal of Soil
and Water Conservation, 58:128–136.
Ferro, V and Minacapilla, M., 1995, Sediment Delivery Processes at Basin Scale,
Hydrological Sciences Journal, 40(6):703-717.
Ferro, V. and Porto, P., 2000, Sediment Delivery Distributed (SEDD) Model,
Journal of Hydraulic Engineering, 5(4): 411-422.
Fu, B., Newham, L. T. H., and Field, J. B., 2009, Modelling Erosion and Sediment
Delivery from Unsealed Roads in Southeast Australia, Mathematics and
Computers in Simulation, 79:2679-2688, 2009.
Haan, C.T., Barfield, B.J., and Hayes, J.C., 1994, Design Hydrology and
Sedimentology for Small Catchments. Academic Press, San Diego,
California, USA.
Jain, M. K. and Kothyari, C. 2000, Estimation of Soil Erosion and Sediment Yield
Using GIS, Hydrological Sciences Journal, 45(5):37–41.
Kupang,
252 22-24 November 2019 9
Kupang, 22-24 November 2019
HATHI the 6th International Seminar
6th International Seminar of HATHI
McCool, D. K., Brown, L. C., Foster, G. R., Mutchler, C. K. and Meyer, L. D.,
1987, Revised slope steepness factor for the Universal Soil Loss Equation.
Trans. ASAE, 30:1387-1396.
Minister of Forestry Regulations Number: P. 32 / MENHUT-II / 2009 concerning
Procedures for the Preparation of Technical Plan for Forest and Land
Rehabilitation in Watersheds.
Mutua, B.M, and Klik, A., 2006, Estimating Spatial Sediment Delivery Ratio on A
Large Rural Catchment, Journal of Spatial Hydrology, 6(1): 64-80.
Novotny, V., and Chesters, G., 1989, Delivery of sediment and pollutants from
nonpoint sources: a water quality perspective, J. Soil Water Conserv, 44:
568-576.
Olii, M.R., 2018, Model sediment delivery ratio untuk daerah aliran sungai,
Disertasi, Universitas Gadjah Mada, Yogyakarta.
Randhir, T.O., O’Connor, R.O., Penner, P.R., and Goodwin, D.W., 2001, A
watershed-based land prioritization model for water supply protection.
Forest Ecology and Management, 143: 47-56.
Renard, K. G., Foster, G. R., Weesies, G. A., McCool, D. K. and Yoder, D. C.,
1997, Predicting soil erosion by water: A guide To Conservation Planning
with The Revised Universal Soil Loss Equation (RUSLE), USDA Agricultural
Handbook 703, USA.
Vis, M., 1987, A Procedure For The Analysis of Soil Erosion and Related Problems
in Water and Land Resources Management Studies, Reserach Report,
Waterloopkundig Laboratorium Delft Hydraulics Laboratory.
Kupang, 22-24November
Kupang, 22-24 November 2019
2019 10
253
