**9. Research methods**

Different phases of the research and the methods which were followed are listed below:


Different phases of the research and the methods which were followed are listed below: - Collecting necessary basic data including air photos, landsat images, geology and

geologic map and climatic classification using Domarten method.

of air photos and numerous field controls and checking were used.

is important in erosion and sediment studies (Saghafian, 2003).





Fig. 5. Geological map of Sarab Drainage Basin

topographic maps and climatic data of the area.

**9. Research methods** 

These units are similar in view points of geology, climate, slope, the kind of morphological processes and prominent erosional features (Ahmadi, 2006). Among polygons of each unit, the polygon that is accessible, has higher area and is good indicator of other similar polygons were selected for rainfall simulator tests. In each land unit, rainfall simulator analyses were performed in three repetitions (total 69 cases) and the produced runoff and sediments were collected and measured (Picture 2). After the analyses, the depths which were wetted by penetration of rainfall, were measured in 3 points of the rainfall simulator plot. Rainfall simulator unit produced rainfalls with constant intensity of 32 millimeters/hour with 30 minutes duration using urban water of Kalaleh in each plot. The time of field tests was late summer after harvesting of dryfarming wheat and before ploughing of cultivated lands for next cultivation. In these times residue of agricultural lands are used by livestock. In rangelands, the growth period was terminated and exploitation of rangeland was started. In most parts of the dry-lands, wheat has been cultivated and the tests of agricultural lands were performed in these areas.


Table 1. Stratigraphic characteristics and area of geological units of Sarab Drainage Basin

Investigation of Effective Factors on Runoff Generation and

Picture 3. Measurement of the depth of A-horizon in each land unit


Zare Chahouki, 2008).

models), Kolmogrove-Smirnov Test was used (Kalantari, 2006).

slopes classes dually (Bihamta and Zare Chahouki, 2008).

samples, percentages of clay, silt and sand in surface materials by hydrometric method, acidity using pH meter in saturated mud, electrical conductivity using Electro conductivity meter in saturated extract, percentage of organic matter by oxidation method, percentage of gypsum, percentage of equivalent calcium carbonate by Calcimetry, cations by flame photometry and titration with EDTA (Ethylene Diamine Tetra Acetic acid), anions by Titrometry, cation exchange capacity by Bour Method and determination of kind and percentage of clay minerals by X-Ray Diffraction Analysis.


2. For investigating the effect of two predominant landuses of rangeland and dry-farming cultivation on the amount of sediment, Independent T-Test was used (Bihamta and

3. For investigating the changes of runoff and sediment production of loesses in different slope classes, One Way Variance Analysis (ANOVA) was used. For determining the fact that which slope classes are different from each other, the method of Clustering of Means used. For this, Duncan Method was used in order to be able to compare all

For determining logical relationships among different variables, correlation and regression analyses were used. Correlation investigates the relationship between variables with sediment and runoff productions and regression investigates the relationship of effective factors with the amount of sediment produced. Due to the fact

Sediment Yield of Loess Deposits Using Rainfall Simulator 125

Picture 2. Collecting runoff at the outlet of plot



land unit, a profile was dug to determine the depth of A-horizon (Picture 3).




Picture 2. Collecting runoff at the outlet of plot

sediments at the end of rainfall was measured.

accurately measured.

Picture 3. Measurement of the depth of A-horizon in each land unit

samples, percentages of clay, silt and sand in surface materials by hydrometric method, acidity using pH meter in saturated mud, electrical conductivity using Electro conductivity meter in saturated extract, percentage of organic matter by oxidation method, percentage of gypsum, percentage of equivalent calcium carbonate by Calcimetry, cations by flame photometry and titration with EDTA (Ethylene Diamine Tetra Acetic acid), anions by Titrometry, cation exchange capacity by Bour Method and determination of kind and percentage of clay minerals by X-Ray Diffraction Analysis.


For determining logical relationships among different variables, correlation and regression analyses were used. Correlation investigates the relationship between variables with sediment and runoff productions and regression investigates the relationship of effective factors with the amount of sediment produced. Due to the fact

Investigation of Effective Factors on Runoff Generation and

drops which fall without initial velocity under gravity force.

Picture 4. Rainfall Simulator which was used in this research

**10. Specification of rainfall simulator** 

Sediment Yield of Loess Deposits Using Rainfall Simulator 127

In general, these equipments are classified into two main types. In some of them, the initial falling velocity of raindrops is zero. In other type, rain drops fall from nozzles which are under pressure and have initial velocity. The equipment which was used in this research is type one from Soil Conservation and Watershed Management Research Center which has been designed and fabricated according to climatic conditions of Iran and can simulate intensities of 5 to 80 mm per hour (Picture 4). It has a plot with dimensions of 84 cm× 120 cm in which all of the resulted runoff and sediments are directed to the outlet where they can be collected(Picture 5). The equipment can be regulated and installed in different slopes. The specifications of the rainfall simulator are shown in Table 3. This equipment produces rain


Table 2. Descriptive information of the areas in which rainfall simulator tests were performed

that the data have interval scale nature, for determination of correlation between variables with sediment and runoff productions, Pearson Correlation Coefficient was used. This coefficient usually changes between +1 and -1, in which the number indicts the slope and the sine shows the direction of linear relationship (Bihamta and Zare Chahouki, 2008).

4. For predicting the variations of the amount of produced sediments from physical and chemical characteristics and determination of the share of each on explanation of sediment production, multiple regression method was used. For doing this, first the factors that have significant correlation with the amount of produced runoff and sediment were selected as independent variables and the amount of produced runoff and sediment of each plot as dependent variable in Stepwise method regression analysis. In this method, the most important variable are entered into regression equation one by one and this function is continued until the error of significant test reaches %5(Kalantary,2006). For observation of co-linearity between the present variables in the extracted models, Variance Inflation Factor (VIF) was considered (Montgomery et al., 2001).
