Origin of Aeolian Sediments Using Fingerprinting Approach in Abu Ghovir Dehloran, Ilam Province

More than two thirds of Iran's area is covered by arid and semi-arid lands. Lack of rainfall in these areas has reduced its ecological diversity and low-density vegetation has been established on it. Lack of vegetation allows the wind to easily erode the soil surface and annually carry large amounts of topsoil from one point to another (Ahmadi, 1999). Finding the source of sediment is one of the basic principles of controlling and combating soil erosion, because by identifying these areas, instead of addressing the problems, the causes can be identified and control erosion activities can be concentrated in source areas (Feng et al., 2011). Knowing the origin of sand dunes is one of the most important examples of soil management in order to optimize exploitation, reduce degradation and conduct wind erosion control plans. Due to the problems of traditional methods, the fingerprinting method as an alternative and appropriate method has been considered by various researchers. Therefore, the main purpose of this study, considering the importance of land uses in erosion and sedimentation process, is investigating the role and importance of dominant land uses in Abu Ghovir region, in Dehloran city of Ilam province, in producing sediment of sand dunes by using fingerprinting method.

The study area is located in Abu Ghovir plain with an approximate area of 19650 hectares in the southeast of Ilam province and on the banks of Doviraj River, at 47° 31' 29'' to 47° 55' 01'' east longitude and 32° 10' 06'' to 32° 24' 19'' north latitude. First, by studying the wind rose of the Dehloran synoptic station during the 25-year statistical period (1992-2017) for the prevailing wind in the study area and using the images received from Google Earth, the study area was determined.Then, by using Landsat 8 satellite imagery and land use map, the existing land uses in the study area including sandy dunes, agricultural lands, Doviraj River and rangelands were determined. After that, 5 soil samples from each land use including rangeland, agriculture and riverside lands as sources of sediment and from sand dunes as sedimentation area from a depth of 0-5 cm were collected.After drying in the open air, the soil samples were passed through a 2 mm sieve and physical and chemical tests of the soil were performed by the following methods. Soluble cations including Ca and Mg by complexometric titration, Na and K soluble by atomic flame emission method (flame photometry) (Rhoades, 1982), soluble chlorine by sedimentation titration using silver nitrate and also carbonate and bicarbonate by simple titration of acid and base were calculated. The elements Fe and Cu were also read by the atomic absorption apparatus (Lindsay, 1979).In order to select the initial detectors, the normality of the data is first checked by the Smirnov Kolmogorov test. If the source characteristics are normal, the quality of the source characteristics of sediment source in the study area is applied by using a two-stage statistical method (Collins et al., 2001) to select the optimal combination of traces in sediment source.

Examination of Kolmogorov-Smirnov test values in the studied land uses shows that the distribution of the studied variables is normal, so parametric statistics can be used to compare these elements in different land uses. To evaluate the ability of detectors in separating work units including rangeland, river and agriculture, Kruskal-Wallis statistical test and stepwise detection function test were performed. The results of Kruskal-Wallis statistical test was not suitable, but the stepwise detection function test was introduced to measure the optimal detector for separating the three sources from the 9 detectors (Na, Cu, Fe, Mg, Ca, Cl, Sulfate, Carbonate, Bi carbonate) in order to select the source of sand dunes including rivers, pastures and agricultural lands. Based on the results, Na and Cu elements were selected as optimal tracers and, in general, 80% of the source samples were correctly classified by these two optimal tracers. To evaluate the different patterns of spatial displacement of the three sediment sources, the distribution diagrams of the first and second functions calculated by the regression method of detection analysis based on the optimal combination of two geochemical tracers were used. The results showed that all three sediment sources were well separated based on these detectors.

The results of granulation and studies conducted showed that there was a very close genetic relationship between sediments in the source area and wind deposits, which indicated the localization of the particles origin and their displacement in the field. Statistical tests showed that Cu and Na detectors have the ability to distinguish different land uses; therefore, these elements are the best source characteristics for the region. The elements in this composition are geochemical elements, so only these elements can be used for source origin studies and determining the share of land uses, and there is no need for other elements used in this research. Finally, by using this combination and hybrid models, the contribution of each land use in the sediment producing process was determined. The results of hybrid models showed that the share of river, rangeland and agricultural land uses was 99.24, 0.76 and 0%, respectively. The relative error of the hybrid model was calculated to estimate the share of different land uses for sediment samples, which was equal to 1%, and the efficiency coefficient of the model was 99%. Due to the fact that the calculated relative error was low and the efficiency coefficients were close to 1, the accuracy and efficiency of the model was confirmed. According to the results, river, rangeland and agricultural land uses had the highest and lowest share in the sedimentation of sand dunes in Abu Ghovir region, respectively. In general, the method of wind sediments origin was able to determine and properly separate the share of land uses in the study area.