109 Advances in Environmental Biology, 6(1): 109-124, 2012 ISSN 1995-0756

This is a refereed journal and all articles are professionally screened and reviewed ORIGINAL ARTICLE

Corresponding Author Saleh Arekhi, Watershed Management Department, Agriculture College, Ilam University, Ilam, Iran Tel: 0098-918-843-0318; Fax: 0098-841-2227015; E-mail: saleh148@yahoo.com

Mapping Soil Erosion and Sediment Yield Susceptibility using RUSLE, Remote Sensing and GIS (Case study: Cham Gardalan Watershed, Iran) 1Saleh Arekhi, 2Ali Darvishi Bolourani, 3Afshin Shabani, 4Hassan Fathizad, 5Salman Ahamdy-asbchin

1Department of Watershed Management, Agriculture College, Ilam University, Ilam, Iran, 2Department of Cartography, Geography College, University of Tehran, Tehran, Iran. 3MSc of Remote Sensing and GIS, Faculty of Geography, University of Tehran, Tehran, Iran, 4MSc Student of Combating desertification, 5Department of biology, Faculty of science, Ilam university, Ilam, Iran

Saleh Arekhi, Ali Darvishi Bolourani, Afshin Shabani, Hassan Fathizad, Salman Ahamdy-asbchin: Mapping Soil Erosion and Sediment Yield Susceptibility using RUSLE, Remote Sensing and GIS (Case study: Cham Gardalan Watershed, Iran)

ABSTRACT

The soil erosion is the most serious environmental problem in watershed areas in Iran. The main factors affecting the amount of soil erosion and sediment yield include vegetation cover, topography, soil, and climate. In order to describe the areas with high soil erosion and sediment yield risks and to develop adequate erosion preventation measures in watersheds of dams, erosion and sediment yield maps should be generated considering these factors. The purpose of this study was to investigate the spatial distribution of annual soil loss and sediment yield in Cham Gardalan watershed, Ilam Province, Iran, using Revised Universal Soil Loss Equation (RUSLE) model. Remote Sensing (RS) and Geographic Information System(GIS) technologies were used for erosion and sediment yield risk mapping, based on the this model. The R-, K-, LS-, C- and P- factors were obtained from monthly and annual rainfall data, soil map of the region, 50-meter Digital Elevation Model (DEM), Remote Sensing (RS) techniques (with use of NDVI), and GIS, respectively. The mean values of the R-, k-, LS-, C- and P- factors were 265.96 MJ mm ha−1 h−1 year−1, 0.29 t h MJ-1 mm-1, 14.31, 0.48 and 1 , respectively. The study indicated that the slope length (L) and slope steepness (S) of the RUSLE model (R2 = 0.84) were the most effective factors controlling soil erosion in the region. The average annual soil loss and sediment yield is predicted up to 38.81 and 19.01 (t h-1 year-1), respectively. The measured average annual sediment yield 16.58 (t h-1 year-1) was very close to estimated value (19.01 t h-1 year-1). Moreover, the results indicated that 47.46%, 11.22%, 9.69%, 11.29%, 20.34% of the study area was under minimal, low, moderate, high and extreme actual erosion risks, respectively. Since 31.63% of the region is under high and extreme erosion risk, adoption of suitable conservation measures seems to be inevitable. The RUSLE model integrated with RS and GIS technologies has great potential for producing accurate and inexpensive erosion and sediment yield risk maps in Iran.

Key words: GIS, RS, RUSLE, Sediment yield, Soil erosion, Cham Gardalan watershed

Introduction

Soil erosion in watershed areas and the

subsequent deposition in rivers, lakes and reservoirs are of great concern for two reasons. Firstly, rich fertile soil is eroded from the watershed areas. Secondly, there is a reduction in reservoir capacity as well as degradation of downstream water quality [17]. Although sedimentation occurs naturally, it is exacerbated by poor land use and land management practices adopted in the upland areas of watersheds. Uncontrolled deforestation due to forest fires, grazing, incorrect methods of tillage and unscientific

agriculture practices are some of the poor land management practices that accelerate soil erosion, resulting in large increases in sediment inflow into streams [41]. Therefore, prevention of soil erosion is of paramount importance in the management and conservation of natural resources [35,48].

Land degradation by soil erosion is a serious problem in Iran with an estimated soil loss of 2500 ×106 t year-1 and about 94% of arable lands and permanent rangelands are in the process of degradation [18,42,30]. In terms of erosion, Iranian soils are under a serious risk due to hilly topography, soil conditions facilitating water erosion (i.e. low

110 Adv. Environ. Biol., 6(1): 109-124, 2012

organic matter, poor plant coverage due to arid and semiarid climate), and inappropriate agricultural practices (i.e. excessive soil tillage and cultivation of steep lands). This widespread problem threatens the sustainability of Cham Gardalam watershed which is the main surface source of drinking water for Ilam city, Iran. Water and soil losses are the main reasons for sediment entering the reservoir and these processes potentially reduce water quality. Soil erosion in this area strongly influences the ecological health of the city.

To estimate soil erosion and to develop optimal soil erosion management plans, many erosion models, such as Universal Soil Loss Equation (USLE) [57], Water Erosion Prediction Project (WEPP) [20], Soil and Water Assessment Tool (SWAT) [2], and European Soil Erosion Model (EUROSEM) [36], have been developed and used over the years. Among these models, the USLE model has remained as the most practical method for estimating soil erosion potential in fields and estimating the effects of different control management practices on soil erosion for nearly 40 years [15,23,38]. The other process-based erosion models require intensive data and thoughtfuly computation steps. The new version of the USLE model, called the Revised Universal Soil Loss Equation (RUSLE), a desktop-based model, was developed by modifying the USLE to more accurately estimate the R, K, C, P factors of soil loss equation, and soil erosion losses [46]. RUSLE is a field scale model, thus it cannot be directly used to estimate the amount of sediment reaching downstream areas because some portion of the eroded soil may be deposited while traveling to the outlet of the watershed, or the downstream point of interest. To account for these processes, the Sediment Delivery Ratio (SDR) for a given watershed should be used to estimate the total sediment transported to the outlet. Erskine et al. [16] compared RUSLE estimated soil loss with the measured sediment yield for 12 subwatersheds in Australia. The coefficient of determination was 0.88 for this comparison, although it did not consider the sediment delivery ratio in the estimated soil erosion using RUSLE. The SDR changes with the size of watersheds, thus, the SDR needs to be considered when RUSLE is applied for a large watershed.

The application of RS and GIS techniques makes soil erosion estimation and its spatial distribution to be determined at reasonable costs and better accuracy in larger areas [33,53]. A combination of RS, GIS, and RUSLE is an effective tool to estimate soil loss on a cell-by-cell basis [33]. Boggs et al. [10] assessed soil erosion risk based on a simplified version of RUSLE using digital elevation model (DEM) data and soil mapping units. Bartsch et al. [6] used GIS techniques to interpolate RUSLE parameters for determining the soil erosion risk in sample plots Camp Wiliams, Utah. Wilson and

Lorang [54] reviewed the applications of GIS in estimating soil erosion. Wang et al. [53] used a sample ground dataset, Thematic Mapper (TM) images, and DEM data to predict soil erosion loss through geostatistical methods. They showed that such methods provided significantly better results than using traditional methods. In general, remote-sensing data were primarily used to develop the cover-management factor image through land-cover classifications [33,47], while GIS tools were used for derivation of the topographic factor from DEM data, data interpolation of sample plots, calculation of soil erosion loss and sediment yield [13,6,53,39].

However, the above studies did not consider the sediment delivery ratio to estimate the sediment delivered to the downstream point of interest. Regional variations in sediment yields are very important since sediment delivery processes vary in space and time. Lin et al. [27] investigated the sediment delivery ratio based on the ratio receiving drainage length to total drainage length by using WinGrid system and computed soil erosion using USLE model. However, this system has separate component programs rather than being fully integrated with a GIS system. Hence, it is not readily available to soil erosion decision makers and it was developed for only research purposes. Instead of this approach, GIS-based Sediment Assessment Tool for Effective Erosion Control (SATEEC) system was used to estimate soil loss and sediment yield for any location within a watershed by a combined application of RUSLE and a spatially distributed sediment delivery ratio within the ArcView GIS software environment [26].

The input data for the SATEEC GIS system are R, K, DEM, C and P maps, which are the basic input maps to RUSLE. Thus, one of advantages of using the SATEEC GIS system is that no additional input data, other than those for RUSLE are required to operate the system. Moreover, all of the functions are fully integrated and automated within the ArcView GIS system. Thus, with several clicks of the mouse button with SATEEC menus, users can estimate the sediment yield for every cell within a watershed.

Keeping this in view, the study is planned to estimate the spatial soil loss and sediment yield in Cham Gardalan watershed to indentify the critical erosion prone areas for soil and water conservation work.

Materials and Methods Study area:

The study area is a mountainous watershed,

called Cham Gardalan watershed, located in the southeast of Ilam Province in the western Iran. The watershed area is 47976 ha and lies between 33°23' 32" to 33°38' 51" N latitudes and 46°20' 30" to 46°39' 33" E longitudes (Fig. 1). The elevations of

111 Adv. Environ. Biol., 6(1): 109-124, 2012

the highest and lowest points are 2582 m and 891 m above mean sea level, respectively. Climatic conditions in the area is semi-arid (typical Mediterranean climate) with mean annual rainfall of about 592.78 mm, and average temperature is 21.7 °C in summer and 4.7 °C in winter [37]. According to land use classification, the most common land use

type and land cover are rangeland (76%), forest(19.3%), rainfed agriculture(2.7%), residential area(0.7%), orchard(0.7%) and lake(0.6%) (Fig. 2 and Table 1). The Cham Gardalan Dam is one of the most important dams in the western Iran that supplies drinking water to the Ilam City.

Fig. 1: Location of the study area

Fig. 2: Land use/land cover of the study area Table 1: Landuse/land cover statistics of the study area

Landuse/land cover type Area(hectare) Area (%) Forest 9228 19.3 Lake 307 0.6 Rainfed Agriculture 1299 2.7 Orchard 357 0.7 Residential area 333 0.7

Rangeland 36452

76

112 Adv. Environ. Biol., 6(1): 109-124, 2012

2.2. Methods: In this study, soil erosion model RUSLE input

parameters and SATEEC GIS system were used to estimate spatial soil erosion and sediment yield of the Cham Gardalan watershed, Ilam Province, Iran. The RUSLE predicts soil loss for a given site as a product of six major erosion factors (equation 1), whose values at a particular location can be expressed numerically. The soil erosion is calculated as follows A = R. K. L. S. C. P (1)

Where A is the average soil loss per unit area by

erosion (t ha−1 year−1), R is the rainfall erosivity factor (MJ mm ha−1 h−1 year−1), K is the soil erodibility factor (t h MJ−1 mm−1), L is the slope length factor, S is the slope steepness factor, C is the

plant cover and management practice factor, and P is the conservation support practice factor. The L, S, C, and P values are dimensionless.

The entire analytical methodology follows the steps shown in Fig. 3. The overall methodology involved the use of the RUSLE in a GIS environment, with factors obtained from meteorological stations (Meteorological Organization of Ilam Province), soil surveys (Watershed Management bureau of the Ministry of Agriculture of Iran), topographic maps with scale of 1:50000(Mapping Organization of Iran) and Landsat ETM+2001 images of the year 2001(was downloaded fromhttp:glcfapp.umiacs.umd.edu:8080/esdi/index.jsp). The layer of each factor was built in the ArcView and ILWIS software to predict soil loss and sediment yield in a spatial domain. The spatial resolution of the data was set to 50 meter.

Fig. 3: Methodology for estimation of soil loss and sediment yield of the study area 2.2.1. Rainfall erosivity factor (R):

Rainfall erosivity is defined as the ability of the

rain to cause erosion [25]. The most common rainfall erosivity index is the R factor of USLE [56,57] and RUSLE [44]. The R factor is considered to be the most highly correlated index to soil loss at many sites

throughout the world [4,8,11,19,22,29,32,45,50,55,57,58,59] The R factor for any given period is obtained by summing the product of total storm energy (E) and the maximum 30-min intensity (I30) for each rainstorm. Since pluviograph and detailed rainstorm data are rarely available at standard meteorological stations, mean

113 Adv. Environ. Biol., 6(1): 109-124, 2012

annual [7,45,60] and monthly rainfall amount (Ferro et al., 1991) have often been used to estimate the R factor for the USLE and RUSLE.

In this study, in order to estimate the R factor, the annual and monthly rainfall quantities were obtained from the records of 16 stations for 22 years (1986-2008) (Table 2). Then, with use of the following equations, Fournier index and R factor were estimated for all the stations. Fournier index [45] F, is defined as:

12

1

12

1

2

i

ii

P

PF (2)

Where, Pi, is the mean rainfall depth in mm of

month i and P is the mean annual rainfall in mm. The Fournier index for all the rainfall records in

the region was estimated (Table 2) using Eq.(2). In order to estimate the most appropriate R-factor based on the F index, the following R-F relationships [45] were used. It must be mentioned that monthly rainfall data were used to calculate R-factor annually because the watershed had no rainfall intensity records.

2.17/07397,0 847,1xFfactorR

When F < 55 mm (3)

2.17/477,0681,077,95 2xFxFfactorR

When F ≥ 55 mm (4) Using the above formulae and the daily rainfall

data of the 16 meteorological stations for the period of 1986-2008, the monthly rainfall erosivity for each station was calculated, and then, every month of each year was summed for the mounthly R factors. The average R-factor values for meteorological stations of the study watershed were obtained by averaging the yearly values from 1986 to 2008 and then, R-factor map layer was made in GIS. Table 2 listed the individual data sets and indicates the name of the station, longitude/latitude and elevation of the station, length of records in years, and some calculated parameters (Rainfall, F and R).

The R-factor map was obtained according to the relationship between elevation, rainfall and F factor. Initially, the relationship between elevation and annual rainfall (R2= 0.80) and annual rainfall and R factor was obtained (R2 =0.89). Then, the correlation between F and R factors was also obtained (R2 =1). Finally, with writing algorithm in ILWIS software, R factor map in grid format was generated. According to Arnoldus [3], the F index is a good approximation of R to which it is linearly correlated. The location of the meteorological stations for the study area is showed in Fig. 4. The R values for the study area are presented in Table 2.

Table 2: Calculated and estimated F, R and rainfall values of the rainfall stations

Stations Longitude

Latitude

Elevation (m)

Length of records (years)

Rainfall values (mm)

F values

R values (MJ mm ha-1 h-1y-1)

Arkavaz 648478 3695819 1290 21 643.50 116.54 391.02 Saleh Abad 610408 3704129 620 19 363.70 87.63 247.55 Sarab Kalan 659379 3716270 970 11 481.00 87.17 235.46 Taleghani 651900 3692580 1400 8 477.90 81.31 210.18 Tulab 647049 3713886 1600 10 522.50 96.86 281.51 Gholender 651010 3723966 1045 10 499.50 87.48 246.88 Mishkhas 642804 3708636 1250 13 590.20 112.00 363.83 Karezan 641860 3733617 1280 21 494.00 85.86 229.63 Gonbad 643307 3680459 860 19 474.00 86.74 233.55 Darageh 647344 3686969 1300 10 747.40 120.28 404.49 Ghajar 638768 3715554 1480 8 477.70 89.54 246.25 Siah Ab 639090 3689984 1220 11 486.00 93.72 266.04 Gol Gol 637736 3703826 1140 10 597.00 110.90 357.44 Gelan 617471 3698514 560 12 344.70 77.11 194.74 Ilam 633013 3716879 1337 22 597.70 111.44 360.55 Ema 633214 3702094 1090 19 511.50 96.72 280.83

2.2.2. Soil erodibility factor (K):

The soil erodibility factor (K) is the rate of soil

loss per rainfall erosion index unit as measured on a standard soil plot and often determined using inherent soil properties [40]. The K values are usually estimated using the soil erodibility nomograph method, which uses % silt plus very fine

sand (0.002 mm-0.1 mm), % sand (0.1 mm-2 mm), % organic matter and soil structure and permeability classes to calculate K [57]. However, there are lacks of structure and permeability class data in the soil survey data sources. Therefore, the following equation is adopted for calculations as recommended by MUSLE in the case of observation data lack [43]

114 Adv. Environ. Biol., 6(1): 109-124, 2012

Fig. 4: Location of the meteorological stations in the study area

))})./.)gD(log(exp(..{(*.K 27101065912104050003405947 (5)

)milnfi(sum.expgD 010

where: K, soil erodibility factor (t ha h ha-1 Mj-1 mm-1); Dg, mean geometric particle diameter (mm); Di, primary particle size fraction (%); Mi, arithmetic mean of the particle size limits of that size (mm) K values for the study area are presented in Table 3.

Table 3: Soil texture types and estimated soil erodibility(K) factor for the watershed studied Texture Clay(%) Sand(%) Silt(%) Area(ha) K Factor Sandy clay loam 60 10 30 13330 0.23 Silty clay loam 34 15 51 15906 0.32 Silt 25 25 50 8253 0.39 Loam 35 25 40 10252 0.26 Silt loam 5 15 80 235 0.5

2.2.3. Topographic factor (LS):

These factors in RUSLE reflect the effect of

topography on erosion. It has been demonstrated that increases in slope length and slope steepness can produce higher overland flow velocities and correspondingly higher erosion [21]. Moreover, gross soil loss is considerably more sensitive to changes in slope steepness than to changes in slope length [31]. Slope length has been broadly defined as the distance from the point of origin of overland flow to the point where either the slope gradient decreases enough where deposition begins or the flow is concentrated in a defined channel [57].

The specific effects of topography on soil erosion are estimated by the dimensionless LS factor

as the product of the slope length (L) and slope steepness (S). In this study, LS-factor map was derived from the DEM map of the region using the “Terrain Analysis” extension of “ArcView 3.2”, which was developed by Schmidt in 2002, based on the seminal work of Moore et al. [34] as follows:

4.0

13.224.1

sA

L (6)

3.1

0896.0

sin

S (7)

115 Adv. Environ. Biol., 6(1): 109-124, 2012

Where, AS is specific watersheds area (m2/m) and is slope angle (degree).

2.2.4. Cover management factor (C): The C factor represents the effect of cropping

and agricultural management practices as well as the effect of ground, tree and grass covers on reducing soil loss in non-agricultural situation [24]. As the vegetation cover increases, the soil loss decreases. According to Benkobi et al. [7] and Biesemans et al. [9], C and LS are the most sensitive factors to soil loss. In the USLE/RUSLE model, the C factor is derived based on empirical equations with measurements of ground cover, aerial cover, and minimum drip height [57].

The most widely used indicator of vegetation growth based on the RS technique is the Normalized Difference Vegetation Index (NDVI), which for Landsat-ETM is given by the following equation:

RNIR

RNIRNDVI

(8)

Where NIR and R are near infrared and red

bands, respectively. This index is an indicator of the energy reflected by the Earth related to various cover intensities. NDVI values range between -1.0 and +1.0. When the measured spectral response of the earth surface is very similar to both bands, the NDVI

values approach to zero. A large difference between the two bands results in NDVI values at the extremes of the data range.

Photosynthetically active vegetation presents a high reflectance in the near IR portion of the spectrum, in comparison to the visible portion. Therefore, The NDVI values for photosynthetically active vegetation will be positive. Areas with low vegetation cover or without vegetative cover (such as bare soil and urban areas) as well as inactive vegetation(unhealthy plants) will usually display NDVI values fluctuating between -0.1 and +0.1. Clouds and water bodies will give negative or zero values.

We used a scene of Landsat ETM+ images acquired on April 15 2001 (as this date coincides with maximum stage of vegetation growth) with a spatial resolution of 28.5 meter. The NDVI was used to calculate the spectral ground-based data, showing the highest correlation with the above-ground biomass [28]. After a reversal linear transformation derived from training samples, the relationship between C and NDVI can be established as C = ((1 − NDVI)/2), by which the C value in each grid cell can be specified (Fig. 5). As the C factor ranges between 0 and 1, a value of 0 was assigned to a few pixels with negative values and a value of 1 to pixels with value greater than 1. Fig. 5 depicts the C factor from the NDVI calculation.

Fig. 5: Cover Management factor (C) from NDVI calculation compared with land cover information [14]

2.2.5. Support practice factor (P):

Since the P factor reflects the impact of support

practices dealing with the average annual erosion rate. It is the ratio of soil loss with contouring and/or strip cropping to that with straight row farming up-and-down slope. The lower the P value, the more effective the conservation practice is deemed to be at reducing soil erosion. As with the other factors, the P-factor differentiates between cropland and

rangeland or permanent pasture. As the study of this research work to estimate soil erosion using RUSLE modeling was applied in the area of non-agriculture or on natural (geological) erosion, it was considered that there was no conservation practices(P) in non-agricultural areas. Therefore as the conservation practice factor (P) value ranges from 0.0-1.0 and highest value is assigned to areas with no conservation practice, the maximum value for P, that is 1.0, is assigned to this research work area [35].

116 Adv. Environ. Biol., 6(1): 109-124, 2012

Results And Discussions 3.1. RUSLE-factors:

The values of the R, K, LS, C and P factors are shown in Table 4. Fournier indices and rainfall erosivity values were calculated using Eqs. (2), (3) and (4) for the rainfall intensity data of 16 stations. A very high and acceptable determination coefficient (R2 = 1) was obtained between R factor values and F indices.R value estimation equation was calculated as: R=0.0277F2-0.3535F+5.568.

The average annual R factor value varies from 194.74 to 404.49 MJ mm ha−1 h−1 year−1. Its total mean and standard deviation (SD) is 265.96 MJ mm ha−1 h−1 year−1 and 38.07, respectively(Table 4). More rainfall erosivity was observed in the north,

northeast and east of the watershed that coincides with higher elevation and ridge of study area and this is shown with green color in Fig. 6. The decreasing R factor has a strong relationship with the decreasing trend of elevation and rainfall from the north, northeast and east ridges to the south and southwest of watershed. The spatial distribution of R-factor map for the study area is presented in Fig. 6.

The average K value varies from 0.23 to 0.50 and its total mean is 0.29 t h MJ-1 mm-1 (Table 4). The standard deviation parameter is 0.06. K value map was generated to show spatial distribution of erodibility (Fig. 7). It can be seen that higher amounts of K values coincides with Gurpi and Pabdeh formation that have the greatest sensitivity to erosion as shown with green color. The spatial distribution of K factor is given in Fig. 7.

Table 4: Values of R, K, LS, C and P factors

Parmeter R K LS C P Max.a 404.49 0.50 80.90 1 1.0 Min.b 194.74 0.23 0 0 1.0 Meanc 265.96 0.29 14.31 0.48 1 SDd 38.07 0.06 17.39 0.10 0.0

a: Maximum;b: Minimum;c: Mean and d : Standard Deviation

Fig. 6: Spatial distribution of R-factor The LS factor was calculated by Eqs. (6) and (7)

by using DEM of the watershed as well as considering the interactions between topography and flow accumulation (Fig. 8). It can be seen that the LS factor varies from 0 to 80.90 and its total mean is 14.31. Maximum area of the watershed has LS values of less than 5. However, some specific areas with steep slopes have greater LS values(such as ridges). The case study area is characterized by decreasing elevation values from north, northeast and east ridges to the southwest(outlet) and south, with a

maximum drop of 891 m. The spatial distribution of topographic factor (LS-factor) is presented in Fig. 8.

In general, the C factor has completely inverse relationship with NDVI. The C-factor value varies from 0.0 to 1 and its total mean is 0.48(Table 4). As seen from the Fig. 9, the greatest amount of this parameter coincides with lake in southwest part of the watershed (dark color) whereas the least amount is related to the area with dense forest vegetation (lighter color). The NDVI based C factor map is presented in Fig. 9.

117 Adv. Environ. Biol., 6(1): 109-124, 2012

Fig. 7: Spatial distribution of K-factor

Fig. 8: Spatial distribution of LS-factor

118 Adv. Environ. Biol., 6(1): 109-124, 2012

Fig. 9: Spatial distribution of C-Factor

Finally, In this study, no major conservation practices are followed. Hence, the conservation practice of 1 is

assigned to this watershed. Fig. 10 depicts the spatial distribution of P-factor in the study area

Fig. 10: Spatial distribution of P-Factor 3.2. Annual Soil loss:

Average annual soil loss was estimated by

multiplying R, K, LS, C and P factors with use of SATEEC extension in ArcView GIS software environment. The resulting map for the study area is presented in Fig. 11. Soil loss values (Table 5) range

between 0 and 6369 t ha-1 year-1 at the pixel level, with mean value of 38.81 (t ha-1 year-1). The SD parameter is 110.41.

Statistical correlations and regression relationships between RUSLE factors and the annual soil loss value for the study area are provided in Table 6. It is clear that the strongest correlation is

119 Adv. Environ. Biol., 6(1): 109-124, 2012

between the LS factor and annual soil loss value (R2=0.84) while the correlations between remaining

factors of RUSLE and annual soil loss are found to be very low.

Fig. 11: Spatial distribution of soil loss (gradual) Table 5: The minimum, maximum, mean and SD soil loss values for the study area

Parameter soil loss (t ha-1 year-1)

Maxa 6369 Minib 0 Meanc 38.81 SDd 110.41

a, b, c and d are maximum, minimum, mean and standard deviation, respectively Table 6: The correlation and regressions among the annual soil loss and RUSLE factors for the study area

RUSLE factors Annual soil loss R r2 = 0.0024, y = -0.3526x + 195.16 K r2= 0.0051, y = 270.91x + 198.17 LS r2= 0.841, y = 20.337x + 45.145 C r2= 0.0666, y = 503.16x - 67.268 P r2= 0.0993, y = 1642.6x - 1071.3

As seen from Fig. 12, the average annual soil

loss in most of the area lies between 0 to 25 t ha-1 year-1. With regard to the spatial variation, the north, northeast and east part of the watershed has more erosion than the south, southwest and northwest parts. However, it should be noted that areas with amount of erosion greater than 80 t ha-1 year-1 have been distributed in the watershed non-uniformly at the pixel level (Fig. 12 and Table 7). Reason for this high soil loss is related to its close relationship with topography (LS factor). This area needs suitable conservation measures on a priority basis. The Spatial distribution of soil loss is given in Fig. 12. 3.3. Assessment on soil erosion risk zone:

In this study, the standard deviation method was

used and the spatial distribution of soil loss was divided into five classes. In this method, natural

breaks of histogram are detected, so that sum of variance into each class is minimum. As seen from the Table 7, most areas of the watershed fall within the minimal (47.46%) and low erosion category (11.22%) which is mostly seen in the west and southwest parts of the watershed. About 31.63% of the watershed is categorized from high to extreme erosion risks which are mostly found in southeast, east and northeast parts of the region (Table 7). The reason for this high soil loss is related to its close relationship with the slope length (L) and slope steepness (S) (R2=0.84). In this area, priority must be given to protection of forest and afforestation of bare lands to reduce soil loss. 3.4. Annual sediment yield:

In this study, three area based methods i.e. USDA SCS [51], Boyce [12] and Vanoni [52] were used in SATEEC to compute the SDR map. The

120 Adv. Environ. Biol., 6(1): 109-124, 2012

average SDR values for the study area range from 0.23 to 0.49 and the average total sediment yields in the outlet of the watershed range from 8.92 to 19.01 t ha-1 year-1 as shown in Table 8.

As seen from the Table 8, the sediment yield (19.01 t ha-1 year-1) estimated using the Boyce [12] method was almost close to the measured sediment yield at the hydrometric station of the Water Resources Department of Ilam Province (16.58 t ha-1 year-1) and the deviation is only 14 percent. Therefore, in this study, Boyce [12] method was selected for calculation of SDR and generation of sediment yield map. In this method, the average annual SDR and sediment yield varies from 0.08 to 0.90 and 0 to 912354 t year-1 for total area of the watershed (average sediment yield is 19.01 t ha-1

year-1), respectively (Fig. 13 and 14). The estimated results are matching well with other studies and local data [1,37], which demonstrates that is a feasible method and technical approach to apply the GIS technology and RUSLE Model to estimate sediment yield in Ilam Province.

Fig. 12: Spatial distribution of soil loss (ordinal)

Table 7: Area and amount of soil loss of each soil erosion risk category for the study area

Erosion categories Numeric range (t ha-1 year-1) Area (%) Minimal 0-5 47.46 Low 5-25 11.22 Moderate 25-50 9.69 High 50-80 11.29 Extreme 80-6369 20.34

Table 8: SDR methods, amount of delivery ratio and sediment yield in the study area

SDR Delivery Ratio(average) sediment yield (t ha-1 year-1)

Vanoni (1975) 0.23 8.92 Boyce (1975) 0.49 19.01 USDA SCS (1972) 0.35 13.58

121 Adv. Environ. Biol., 6(1): 109-124, 2012

Fig. 13: Spatial distribution of sediment delivery ratio [12]

Fig. 14: Spatial distribution of sediment yield [12]

4. Conclusions: A quantitative assessment of soil loss and

sediment yield on grid basis was carried out using the well-known RUSLE model with a view to identify the critical soil erosion prone zones for conservation planning in a study watershed in the Western Iran. Detailed data for the computation of the R, C and P factors were not available; therefore these parameters were estimated either by means of general or approximation formulae (i.e. R factor) or by processing available satellite images (i.e. for the extraction of C factor) and GIS(P factor). Moreover,

the estimation of LS factor was performed with the use of a GIS automated technique to generate L and S factors. All the maps of R, K, LS, C and P were integrated to generate erosion and sediment yield risk map to find out spatial distribution of soil loss and sediment yield within GIS environment. The average annual soil erosion for Cham Gardalan watershed was found to be 38.81 t ha-1 year-1. About 58% of the watershed area was found out to be under minimal and low erosion class which was covering northwest and southwest parts of the watershed. About 31.63% of the watershed is categorized under high to extreme erosion risks which were mostly found in northeast,

122 Adv. Environ. Biol., 6(1): 109-124, 2012

east and southeast parts of the region. Since 31.63% of the region is under high and extreme erosion risk, adoption of suitable conservation measures seems to be inevitable. So, generated soil loss map was able to indicate high erosion risk areas to soil conservationist and decision maker. Major factors effecting soil erosion in the study area were found to be L and S parameters of RUSLE (R2=0.84). In this area, priority must be given for protection of forest and afforestation of steep barren lands and maximizes plant coverage. Meanwhile, the estimated average annual sediment yield (19.01 ton h-1 year-1) was close to measured average annual (16.58 t ha-1 year-1). The results of RUSLE model demonstrated that this is a feasible method to apply the RS, GIS and RUSLE integrated model to estimate the sediment yield of the watershed.

Establishing of database through conventional methods is time consuming, tedious and difficult to handle. In the present study, attempt was made to utilize RS data for obtaining land use/land cover data which are essential prerequisites for generation of RUSLE factors. Thus, RS and GIS techniques can play significant role in generation of parameters from remote areas of watersheds for the purpose of soil erosion and sediment yield modeling as well as watershed management. The strength of GIS relies on its ability to handle spatial data and attribute information at a higher level of resolution. Once the input information is entered into the GIS environment, analyses can be readily performed to yield accurate data without relying on tedious, expensive manual methods. High-quality output, easy updating capabilities, and the possibility for testing management options make GIS particularly useful for providing information for decision-makers. Thus, RS and GIS techniques can assist in developing management scenarios and provide options to policy makers for handling soil erosion problem in the most efficient manner for prioritization of watershed areas for treatment. The representation of model input and output also facilitates examination of a wider range of alternatives than would be possible by using standard methods. References

1. Azami, Ayad, 2006. Comparison of various

methods for estimating sediment yield in Cham Gardalan watershed. Msc Thesis.Gorgan University.

2. Arnold, J.G., R. Srinivasan, R.S. Muttiah and J.R. Williams, 1998. large area hydrologic modeling and assessment: Part I. Model development. Journal of the American Water Resources Association, 34(1): 73-89.

3. Arnoldus, H.M.J., 1980. An approximation of the rainfall factor in the Universal Soil Loss Equation. In: De Boodt M, Gabriels D (eds)

Assessment of erosion. Wiley, Chichester, pp: 127-132.

4. Aronica, G. and V. Ferro, 1997. Rainfall erosivity over the Calabrian region. Hydrological Science Journal, 42(1): 35-48.

5. Banasik, K. and D. Goˆrski, 1994. Rainfall erosivity for South-East Poland. In: Rickson RJ (ed) Conserving soil resources. European perspectives. Lectures in soil erosion control, Silsoe College, Cranfield University, UK, pp: 201-207.

6. Bartsch, K.P., H. Van Miegroet, J. Boettinger. and J.P. Dobrwolski, 2002. Using empirical erosion models and GIS to determine erosion risk at Camp Williams.Journal of soil and water conservation, 57: 29-37.

7. Benkobi, L., M.J. Trlica. and J.L. Smith, 1994. Evaluation of a refined surface cover subfactor for use in RUSLE. J Range Manage, 47: 74-78.

8. Bergsma, E., 1980. Provisional rain-erosivity map of The Netherlands. In: De Boodt M, Gabriels D (eds) Assessment of erosion. Wiley, Chichester.

9. Biesemans, J., M.V. Meirvenne, D. Gabriels, 2000. Extending the RUSLE with the Monte Carlo error propagation technique to predict long-term average off-site sediment accumulation. J Soil Water Conserv, 55: 35-42.

10. Boggs, G., C. Devonport, K. Evans and P. Puig, 2001. GIS-based rapid assessment of erosion risk in a small catchment in the wet/dry tropics of Australia. Land Degrad. Dev., 12: 417-434.

11. Bolinne, A., A. Laurant. and P. Rosseau, 1980. Provisional rain-erosivity map of Belgium. In: De Boodt M, Gabriels D (eds) Assessment of erosion. Wiley, Chichester, pp: 111-120.

12. Boyce, R.C., 1975. Sediment routing with sediment delivery ratios. Present and Prospective Technology for ARS. USDA, Washington, D.C.

13. Cerri, C.E.P., J.A.M. Dematte, M.V.R. Ballester, L.A. Martinelli, R.L. Victoria and E. Roose, 2001. GIS erosion risk assessment of the Piracicaba River Basin,southeastern Brazil,Mapping sciences and remote sensing, 38: 157-171.

14. Chao-Yuan, Lin., Lin. Wen-Tzu, Chou. Wen-Chieh, 2002. Soil erosion prediction and sediment yield estimation: the Taiwan experience, Soil and Tillage Research, 68: 143-152.

15. Dennis, M.F., M.F. Rorke, 1999. The relationship of soil loss by interill erosion to slope gradient. Catena, 38: 211-222.

16. Erskine, W.D., A. Mahmoudzadeh and C. Myers, 2002. Land use effects on sediment yields and soil loss rates in small basins of Triassic sandstone near Sydney, NSW, Australia. Catena, 49: 271-287.

17. European Environment Agency, 1995. CORINE soil risk and important land resources-in the

123 Adv. Environ. Biol., 6(1): 109-124, 2012

southern regions of the European Community, Commission of the European Communities.

18. FAO/UNEP, 1994. Land degradation in south Asia: its severity causes and effects upon the people. FAO, UNEP and UNEP project, Rome.

19. Ferro, V., G. Giordano and M. Lovino, 1991. Isoerosivity and erosion risk map for Sicily. Hydrol Sci J 36(6): 549–564.

20. Flanagan, D.C. and M.A. Nearing, 1995. USDA water erosion prediction project: hillslope profile and watershed model documentation. NSERL Report No. 10. USDA-ARS National Soil Erosion Research Laboratory, West Lafayette, IN 47907-1194.

21. Haan, C.T., B.J. Barfield. and J.C. Hayes, 1994. Design hydrology and sedimentology for small catchments. Academic Press, San Diego, pp: 588.

22. Hussein, M.H., 1986. Rainfall erosivity in Iraq. J Soil Water Conserv, 41: 336-338.

23. Kinnell, P.I.A, 2000. AGNPS-UM: applying the USLE-M within the agricultural non point source pollution model. EnvironmentalModelling and Software, 15(3): 331–341.

24. Kouli, M., P. Soupios and F. Vallianatos, 2009. Soil erosion prediction using the Revised Universal Soil Loss Equation(RUSLE) in a GIS framework, Chania, Northwestern Crete, Greece, Environ Geol., 57: 483-497.

25. Lal, R., 1990. Soil Erosion in the Tropics. Principles and Management. McGraw-Hill, New York. pp: 580.

26. Lim, KJ., J. Choi, K. Kim, M. Sagong and B.A. Engel, 2003. Development of sediment assessment tool for effective erosion control (SATEEC) in small scale watershed. Transactions of the Korean Society of Agricultural Engineers, 45(5): 85-96.

27. Lin, C., W. Lin. and W. Chou, 2002. Soil erosion prediction and sediment yield estimation: the Taiwan experience. Soil and Tillage Research, 68: 143-152.

28. Lin, C.Y., 1997. A study on the width and placement of vegetated buffer strips in a mudstone-distributed watershed.J.china. Soil water conserve.29 (3), 250-266(in Chinese with English abstract).

29. Lo, A., S.A. El-Swaify, E.W. Dangler and L. Shinshiro, 1985. Effectiveness of EI30 as an erosivity index in Hawaii. In: El-Swaify SA, Moldenhauer WC, Lo A (eds) So/7 erosion and conservation. Soil Conservation Society of America, Ankeny, pp: 384-392.

30. Masoudi, M., A.M. Patwardhan. and S.D. Gore, 2006. Risk assessment of water erosion for the Qareh Aghaj subbasin, southern Iran. Stoch Environ Res Risk Assess., 21: 15-24.

31. McCool, D.K., L.C. Brown and G.R. Foster, 1987. Revised slope steepness factor for the

universal soil loss equation. Trans Am Soc Agric Eng., 30: 1387-1396.

32. Mikhailova, E.A., R.B. Bryant, S.J. Schwager. and S.D. Smith, 1997. Predicting rainfall erosivity in Honduras. Soil Sci Soc Am J 61: 273-279.

33. Millward, A.A. and J.E. Mersey, 1999. Adapting the RUSLE to model soil erosion potential in a mountainous tropical watershed. Catena, 3: 109-129.

34. Moore, I.D. and J.P. Wilson, 1992. Length-slope factors for the Revised Universal Soil Loss Equation: Simplified method of estimation. Journal of Soil and Water Conservation, 47: 423-428.

35. Morgan, R.P.C., 1995. Soil erosion and conservation. Longman, London, pp: 23-37.

36. Morgan, R.P.C., J.N. Quinton, R.E. Smith, G. Govers, J. Poesen, K. Auerswald, G. Chisci, D. Torri. and M.E. Styczen, 1998. The European Soil Erosion Model (EUROSEM): a dynamic approach for predicting sediment transport from fields and small atchments. Earth Surface Processes and Landforms, 23: 527-544.

37. Nepta Advisory Coopeation, 2001. Detail project of Ilam Dam watershed, Jahad and Agriculture Organization of Ilam province., pp: 135.

38. Pandey, A., V.M. Chowdary and B.C. Mal, 2007. Identification of critical erosion prone areas in the small agricultural watershed using USLE, GIS and Remote Sensing. Water Resour. Manage (Springer)., 21(4): 729-746.

39. Pandey, A., V.M. Chowdary and B.C. Mal, 2009. Sediment Yield Modelling of an Agricultural Watershed Using MUSLE, Remote Sensing and GIS. J. Paddy Water Environment (Springer)., 7(2): 105-113.

40. Parysow, P., G.X. Wang, G. Gertner and A.B. Anderson, 2003. Spatial uncertainty analysis for mapping soil erodibility based on joint sequential simulation. Catena, 53 :65-78.

41. Pimental, D., 1998. Ecology of soil erosion in ecosystems. Ecosystems (1),pp: 416-426.

42. Programme and budget organization, 1996. First national report on human development of Iran.

43. Renard, K.G., G.R. Foster, Weesies, G.A, D.K McCool and D.C. Yoder, 1997. Predicting soil erosion by water: a guide to conservation planning with the Revised Universal Soil Loss Equation (RUSLE). Agriculture Handbook No. 703, USDA-ARS.

44. Renard, K.G., G.R. Foster, G.A. Weesies, D.K. McCool, 1996. Predicting soil erosion by water. A guide to conservation planning with the revised universal soil loss equation (RUSLE). Agric. Handbook 703. US Govt Print Office, Washington, DC.

45. Renard, K.G. and J.R. Freimund, 1994. Using monthly precipitation data to estimate the R

124 Adv. Environ. Biol., 6(1): 109-124, 2012

factor in the revised USLE. J Hydrol., 157: 287-306.

46. Renard, K.G., G.R. Foster, G.A. Weesies, J.P. Porter, 1991. RUSLE: revised universal soil loss equation. Journal of Soil and Water Conservation, 46(1): 30-33.

47. Reusing, M., T. Schneider and U. Ammer, 2000. Modeling soil erosion rates in the Ethiopian Highlands by integration of high resolution MOMs-2/D2-stereo-data in a GIS.International Journal of Remote sensing, 21: 1885-1896.

48. Sadeghi, S.H.R., B. Ggaderi Vangah and N.A. Safaeeian, 2007. Comparision between effects of open grazing and manual harvesting of cultivated summer rangelands of Northern Iran on infiltration,runoff and sediment yield. Land degradation and development (DOI/10,1002:1dr.799).

49. Schmidt, F. and A. Persson, 2002. Comparison of DEM Data Capture and Topographic Wetness Indices. Precision Agriculture., 4(2): 179-192.

50. Stocking, M.A. and H.A. Elwell, 1973. Prediction of sub-tropical storm soil losses from field plot studies. Agric Meteorol., 12: 193-201.

51. USDA, 1972. Sediment Sources, Yields, and Delivery Ratios. National Engineering Handbook, Section 3 Sedimentation.

52. Vanoni, V.A., 1975. Sedimentation Engineering, Manual and Report No. 54. American Society of Civil Engineers, New York, N.Y.

53. Wang, G., G. Gertner, S. Fang and A.B. Anderson, 2003. Mapping multiple variables for predicting soil loss by geostatistical methods

with TM images and a slope map.photogrammetric Engineering and remote sensing, 69: 889-898.

54. Wilson, J.P. and M.S. Lorang, 2000. Spaatial models of soil erosion and GIS.In spatial models and GIS.New potential and new models,Fotheringham AS, Wegener M(eds).Taylor and Francis:Philadelphia, PA; 83-108.

55. Wischmeier, W.H., 1959. A rainfall erosion index for a universal soil loss equation. Soil Science Society of America Proceedings, 233: 246–249.

56. Wischmeier, W.H. and D.D. Smith, 1965. Predicting rainfall-erosion losses from cropland east of the Rocky Mountains. Agricultural Handbook, vol. 282. USDA, Washington.

57. Wischmeier, W.H. and D.D. Smith, 1978. Predicting rainfall erosion. losses: a guide to conservation planning. Agriculture Handbook, vol. 537. US Department of Agriculture, Washington,DC, pp: 58.

58. Yu, B. and C.J. Rosewell, 1996a. An assessment of a daily rainfall erosivity model for New South Wales. Aust J Soil Res., 34: 139 152.

59. Yu, B. and C.J. Rosewell, 1996b. Rainfall erosivity estimation using daily rainfall amounts for South Australia. Aust J Soil Res., 34: 721-733.

60. Yu, B. and C.J. Rosewell, 1996c. A robust estimator of the R factor for the Universal Soil Loss Equation. Trans Am Soc Agric Eng., 39(2): 559-561.