Journal of Spatial Analysis

Investigation of desertification trend needs an understanding phenomena creating changes singly or action and reaction together in the manner that these changes end up in land degradation and desertification. Since Lampery reported the south advancement of the Sahara Desert at an average rate of 5-6 km/year in the period 1958-1975 in the UNESCO/UNEP conference in 1975, “desert encroachment” or “desert advancing” or “desert marching” in Africa had drawn the attention of a great number of institutions, individual scientists and governments and “desertification” has become one of the major subjects of dryland environmental research in the world in the past decades (Wu et al., 2008). Land degradation occurs everywhere but is defined as desertification when it occurs in the drylands, resulting from various factors, including climatic variations and human activities (UNCCD, 2012). Investigation of desertification potential in Alluvial fans located in the south east of the Rude short watershed area who consist of fans originated from different susceptible geological formations to erosion using nonparametric methods and attributes affecting desertification is very crucial. The aim of this research is presenting indices to determine the role of pedological criterion affecting rangeland desertification and prioritization of desertification potential in research area using nonparametric methods of Friedman and Kendalls' W rank tests.
Rude-shoor watershed area is about 17000 square kilometers. 42 percent of the area is plain and the other is highland. This area has been located in the geographic extent of 48º 30׳ to 51º (East) and 35º 21׳ to 36º 30׳ (North) and between two different geological systems and structures of south Alborz and central Iran.
To investigate pedological criterion affecting land degradation and desertification in alluvial fans, first, a part of the Rud-e-shoor watershed area was selected. After distinguishing target area, maps of slop classes, land use and geology were created, then a map of units was created by overlaying and crossing them. To create map of slope classes, first, digital elevation model (DEM) was created using ArcGIS 10.3 software, then slop map was created. It was shown as the classified map using slope classes of 0-1 (class 1), 1-2 (class 2), 2-4 (class 3), 4-8 (class 4), 8-15 (class 5), and >15 (class 6). Land use map of “Watershed Atlas” (SCWMRI, 2008) was also used after monitoring and investigating the target area using Google Earth, Landsat 8, and field studies. After creating three maps of rock unit, slope classes and land use, map of work units was created using ArcGIS by overlaying and crossing them. After creating units, grids of 1000*1000 square meters in the target area were created using an extension of ET GeoWizards in ArcGIS software to create more units.
In this research three indices of erodibility, salinity and permeability of soil were considered, finally according to the seven level scales, each of them was shown as the classified map. In this research, Linguistic variables of criteria are very low, low, rather low, medium, rather high, high and very high.  Because of the presence of linguistic attributes such as resistance coefficient to erosion, nonparametric tests is a useful method to determine desertification potential. First, Research area from the viewpoint of resistance coefficient to erosion was classified into seven classes of very low (0-2), low (2-4), rather low (4-6), medium (6-8), rather high (8-10), high (10-12) and very high (>12) on the basis of the soil texture samples and coefficient of erodibility derived from Morgan table (1986) and also the limits of resistance coefficient to erosion derived from the Feiznia method (1995). Then, Research area zonation from the viewpoint of the salinity index to be done. After sampling, the electrical conductivity of saturated mud of 159 samples was measured by EC-meter by mmhous.cm^-1 and finally classification of salinity with considering seven classes of salinity (very low (0≤ ECe <2), low (2≤ ECe <4), rather low (4≤ ECe <6), medium (6≤ ECe <8), rather high (8≤ ECe <10), high (10≤ ECe <12) and very high (12≤ ECe)) to be done according to the USSL method (1954). Finally, Research area zonation from the viewpoint of the permeability coefficient to be done. To classify the area from viewpoint of permeability coefficient, sampling of 174 samples by brazen rings to be done and permeability coefficients of them were measured according to Darcy׳s law by meters per day  (m.day-1), and finally classification of permeability based on seven classes of permeability coefficient of very low (<0.069 cm.min -1), low (0/069-0/900 cm.min -1), rather low (0/900-1/388 cm.min -1), medium (1/388-6/944 cm.min -1), rather high (6/944-10/000 cm.min -1), high (10/000-12/500 cm.min -1) and very high (>12/500 cm.min -1) to be done according to the Bouwer classification (1978).
 In this research, nonparametric methods of Friedman and Kendalls' W rank tests were used to prioritize the options, in the manner that after defining crisp numbers in seven level scales, the rank means were calculated.
The rank means obtained from Friedman and Kendalls' W rank tests alter from 7/00 to 17/33. Obtained results from the nonparametric methods showed that 74/18 % (4245/77 ha) and 25/82 % (1477/67 ha) of the area were classified into two classes of medium and very heigh desertification potential, respectively. The results also showed that the youngest and oldest gravel fans of the area have desertification potential of very high (798/75 ha) and low (152/23 ha), respectively. Results also showed that alluvial fans originated from different susceptible geological formations to erosion, such as malls, have very high desertification potential.
Desertification can also be determined using multiple criteria decision making techniques such as the fuzzy simple additive weighting or fuzzy logic models.