Examination of the relation between cover management factor with vegetation indices in the loess slope land (case study: Wheat farm in toshan catchment)

Background and objectives: Erosion rate in the Golestan province has a high rate due to geographical location, climatic and resource degradation. Determining a suitable method for measuring soil erodibility is one of the priorities in the prevention of destructive erosion. Wishmeyer and Smith equation (1978) is one of the important methods for calculating soil erosion, that is known the Universal Soil Loss Equation. In this equation, cover management (C) is one of the six factors affecting soil water-erosion that it`s measuring is not simply possible. One of the widely used methods for estimating this factor is using satellite imagery. Researcher provided many ways to evaluate vegetation management factor using NDVI to assess soil losses with USLE method (Lin et al., 1999; DeJong, 2002). These methods use regression model to analyze the correlation between the amount of measured C factor at the farm.
So, in this research scrambled to derive relevant indicators from satellite images of different in the Golestan province agronomic main crop growth and obtain their relations between them with real erosion in the farm.

Materials and methods: Evaluation C factor was measured by using simulated rainfall with 2 intensity of 32 and 105 mm/h in duration 20 minute in 6 consecutive times. Cover management factor also was measured using CERL method, and vegetation indices (SAVI & NDVI) were computed from of Landsat 8 images in 6 consecutive times using ERDAS IMAGIN 2011. Then, the relations between CSERL factor and vegetation indices (SAVI & NDVI) were obtained using Excel software and Adjusted coefficient of determination, Regression coefficient and Correlation were obtained using SAS software were used as evaluation indices.
Results: According to the results, ) in order to predict CSERL factor using satellite images NDVI index (R2 = 0.76 is more suitable comparing SAVI index (R2 = 0.54). The vegetation indices derived from satellite images have higher relation with C factor derived from higher rain intensity. By examining trends of CSERL changes in 6 consecutive time on the rainfall intensity with 105mm at duration of 20 minutes per hour rainfall was also found that this factor was 0.16 and 0.03 in December and June respectively.

Conclusion: In this survey, observed that there is significantly relation between C factor and NDVI index (R2 = 0.76 at 99% probability level. According to the results, NDVI index had a best estimation with CSERL in higher intensity (R2 = 0.76) and so this index is offered as a method with proper accuracy.