International Journal of Agriculture and Crop Sciences. Available online at www.ijagcs.com IJACS/2014/7-6/304-314 ISSN 2227-670X ©2014 IJACS Journal

The estimation of erosion and sediment by using the RUSLE model and RS and GIS techniques (Case study: Arid and semi-arid regions of Doviraj, Ilam province, Iran) Hassan Fathizad1, *Haji Karimi2, Seyed Maryam Alibakhshi3 1. M.Sc. Graduate of Combating desertification, Rangeland and Watershed Department, Ilam University, Iran 2. Associate Professor of Rangeland and Watershed Department, Agriculture College, Ilam University, Ilam, Iran 3. Ph.D. Student in water engineering (irrigation drainage), Faculty of Natural Resources, University of Ferdosi Corresponding Author email: [email protected] ABSTRACT: Nowadays, soil erosion is expressed as one of the important topics of watershed management at national and international level. The estimation of soil reduction risk and its spatial distribution is one of the important factors for the successful assessment of soil erosion. , The purpose of this study is prediction of annual loss potential of soil and sediment load. The soil reduction due to erosion can be obtained by using prediction model like the Universal Modified Equation of Soil Erosion (RUSLE). In this research, the RUSLE model is used for estimation the erosion and sediment yield in Doviraj watershed located in Ilam province, Iran. The RUSLE factors are included R, K, LS, C and P that are calculated from data of rainfall, region soil map, Digital Elevation Model (DEM) and remote sensing techniques, respectively. The mean values of R, K, LS, C and P factors were equal to -1 -1 -1 -1 -1 -1 209.04MJ mm ha h y , 0.51 Mg ha h ha MJ mm , 10.16, 0.53 and 1, respectively. The mean annual sediment yield was calculated as 273.6 ton/hectare/year in the study area that was close to the measured value from Doviraj sediment gauging station (253.42 ton/hectare/year). The calculated sediment value is 7.96% more than that of the measured sediment value (observed) (based on calculating the relative error). Therefore, there is no significant difference between rate of the observed and calculated sediment. Also the results showed that measurement of remote sensing techniques and GIS can be used for evaluate and calculate the rate of erosion and sedimentation. Key Words: soil erosion, RUSLE model, remote sensing, GIS, Ilam province. INTRODUCTION Erosion and sediment production are one of the important problems at the country watershed management. Erosion and its consequences, with intensification of the man exploitation of nature from the early twentieth century, are entered owns negative effects on the vital ecosystem (Lu et al., 2001).The main reason is population increase and overuse of land (Ahmadi, 1999). The scientific plans for soil conservation require to accurate information about the factors that caused erosion. According to the causative factor of erosion at each location, the accuracy of this information causing that take place the appropriate protection operations with it. Accurate information from the erosion is very low in Iran and also was observed very differences between the observations and calculations. Being young the research of this field and lack of longterm measurements of erosion is prevented to achieve the reliable numbers (Arab Khadri, 2005).As regards that the measurement of the watershed water erosion rate were done in the hydrometric stations and sediment survey, the absence of these stations creates problems for calculation the soil erosion rate at the watershed surface. The estimation close to reality of the erosion and sediment value in an area always has been considered by researchers. The lack of the accurate statistics of erosion and sediment value in many countries, the use of erosion and sediment calculation models is inevitable (Ahmadi, 1999). The watershed of Doviraj River is one of the largest watersheds of Ilam province. There is a reservoir dam in the output of this watershed. The outburst of this river in the wet seasons leads to create the destructive floods in downstream lands and Mosiyan fertile plain. Therefore, the study of erosion and sediment yield of this watershed has the great importance. Several models have been presented for the calculation and management plans development of soil erosion that the most important of these models, can be refer to: Universal Equations

Intl J Agri Crop Sci. Vol., 7 (6), 304-314, 2014

of Soil Erosion (Wischmeier and Smith, 1978), Water Erosion Prediction Project (Flanagan and Nearing, 1995), Soil and Water Assessment Tool (Arnold et al., 1998) and European Soil Erosion Model (Morgan et al, 1998). During the recent 40 years, USLE model due to the simple calculations, has been mostly used for the estimation of potential soil erosion and different management operations effects (Kinel, 2000), and subsequent, the new version of USLE model has been developed with name of the Revised Universal Soil Loss Equation (RUSLE) that does more precise estimations of R, K, C, P factors and soil erosion (Renard et al., 1991; Van Remortel et al., 2004). Due to the capability of Geographic Information System (GIS) and Remote Sensing (RS) in geospatial data analysis; RUSLE/USLE models are combined with GIS and RS (Ouyang and Bartholic, 2001; Lufafa et al., 2003). Other advantages of using the GIS and RS techniques, is the possible of soil erosion estimation and erosion geospatial distribution with the acceptable cost and accuracy at the large areas (Mill ward and Mersey, 1999; Wang et al., 2003). For example, by combining RS techniques, GIS and this model, the network (distributor) soil erosion potential can be estimated (Mill ward and Mersey, 1999). Boggs et al. (2001) estimated the soil erosion risk based on RUSLE revised model, DEM data and land units map. Bartsch et al. (2002) used the GIS techniques for calculation the required factors of RUSLE for determination of soil erosion risk at Camp Williams. Considering the limitations of previous studies, Wilson and Lorang (2002), studied the GIS applications for soil erosion estimation and proved that the GIS provides the extraordinary ability for improvement and soil erosion estimation. Wang et al. (2003) have attempted using the geocentric data, Landsat images (TM) and DEM for soil erosion prediction by geo statistical methods. These researchers showed that these methods give significantly better results than that of the traditional methods. Arekhi and Niazi (2010) used the RUSLE model, RS techniques and GIS for soil erosion estimation and sediment yield of Ilam dam. The results are shown that the mean of annual sediment yield was calculated as 14.75 ton/hectare/year in the study area which is close to the obtained value of sedimentation station of Ilam dam (16.58 ton/hectare/year). The result of this research also showed, LS factor with correlation coefficient of 0.77 has been the maximum effect for the annual soil erosion estimation by the RUSLE model. This study proved the effectiveness of RS and GIS for quantitative estimation of soil erosion values, sediment yield and also erosion management. The purpose of this research is the modeling of soil erosion rate and sediment yield of Doviraj watershed by combination of the RUSLE model, GIS and RS. METHODS Characteristics of the study area The Doviraj watershed is located at the geographic range of47° 16 ́- 47° 40 ́eastern length and 32° 34 ́-33° 05 ́north latitude in south of Ilam province, west of Iran. This watershed is beside the Karkheh River and limited to the heights of Kabir Kouh at north. The Black and Kase Mast mountains are the most important mountains of watershed. The area of study region is 118837.77 hectare and the maximum and minimum elevations of the study area are 2163 and the 177m above sea level respectively. The Ilam-Khuzestan road is located at the east to west direction and south of watershed. Figure 1 is shows the location of the study area at Ilam province and Iran.

Figure 1. The location of the study area

MATERIALS The early studies are included the interpretation of topographic maps(1:50000), land capability map (1:80000), the study of rainfall statistics of weather stations, examination of detailed study report of Doviraj + watershed of Ilam, the satellite image ETM (2007) and ArcView 3.2,Envi 4.8,Idrisi ands, ArcGIS 9.3 and ILWIS3.3soft wares. 304

Intl J Agri Crop Sci. Vol., 7 (6), 304-314, 2014

THE RESEARCH METHOD Universal Soil Loss Equation (USLE) and its revised version (RUSLE) are used for prediction of soil erosion and design of protective programs (Renard et al., 1991; Sadeghi et al., 2004). The RUSLE model is a water erosion estimation model that related to the following six erosion factors (Wischmeier and Smith, 1978) (equation 1): A= R. K. L. S. C. P(1) -1 -1 -1 In this equation, A is mean soil erosion at area level (ton/hectare/year),rain erosivity factor MJ mm ha h y -1 -1 -1 (R), soil erodibility factor Mg ha h ha MJ mm (K), slope length factor (L),slope degree factor (S), management operations and coatings factor (C) and protective operations (P). L, S, C and P values are no unit. In this study, the general method is involved using of RUSLE model in GIS area. Next paragraphs also explains how to estimate R, K, C, P and LS factors that obtained from rainfall data, soil maps, processing of satellite images(C and P factors) and Digital Elevation Model. Rain erosivity factor Rain erosivity is defined as the rain compressional power in incidence of erosion (Lal, 1990). The most common index of the rain erosivity is R factor which is related to USLE and RUSLE models. The study of scientific papers showed, in many parts of the world, R factor has high correlation with soil erosion (Ferro et al., 1991; Renard and Freimund, 1994; Wischmeier and Smith, 1978; Yu and Rosewell, 1996). The R factor for different periods is obtained from the product of rain kinetic energy (E) in the maximum of 30-minute rainfall intensity I 30 . Since the rainfall graph and shower detailed data (intensity of rainfall) are available rarely on

 

water stations, mostly monthly and annual mean values of rain is used for estimation of R factor in USLE and RUSLE models (Renard and Freimund, 1994;YuandRosewell, 1996; Ferro et al., 1991). For determination of R factor, after determination of index stations at the study area, monthly and annual rainfall was reconstructed at these stations and the study period. Figure 2 shows 12 stations which are used in this study. In the next stage, by using the following equation, the Fournier index and R factor for all stations are obtained. Fournier index (F) value from equation (2) is obtained (Renard and Freimund, 1994): 12

F

 pi i 1 12

2

p

(2)

i 1

In this equation pi is the average rainfall in month i and p is the average annual rainfall (mm). In this study, by using equation (3 and 4), Fournier index was calculated for all station and then with substitution Fournier index (equation 5) in the following equations that proposed by Renard and Freimund (1994) for the regions with lack of shower detailed data (intensity of rainfall), R factor value estimated for index stations.

R  factor  ( 0 .07397  F 1 .847 ) / 17 .2

(3)

F