Mapping Soil Particle Size and Organic Matter Distribution in Different Land Uses in Bardeh Watershed (Chaharmahal and Bakhtiari Province) Using Remote Sensing and Geostatistics Techniques

Soil is one of the natural and capital resources of each country that should be seriously considered in land-use planning at national and local scales.   The purpose of this study was to investigate the effect of land use on organic matter content and soil particle size distribution in Bardeh watershed. To fulfil the objective, 140 composite and random surficial soil samples were taken from a depth of 0-5 and 0-20 cm. Geostatistical analysis of the spatial structure of soil characteristics showed that their distributions followedan exponential model and had medium to intermediate spatial dependencies. The results also showed that there were significant Pearson correlation coefficients between reflection in visible, infrared, and thermal bands with soil texture components. In order to map different variables, the interpolation and zoning were done by Kriging method. The maps of different soil properties showed that the distribution pattern of organic matter and particle size distribution were mainly due to the variability of soil inherent factors (such as soil material) and partly by management practices (such as soil management). Lastly, using Landsat 8 images, the soil properties were compared with radiation, reflection and degree. Radiation heat was investigated using regression models. The results showed that at the two depths of sampling 0-5 and 0-20 cm, and in different applications, the coefficient of variation did not show a significant difference.  Correlation coefficients showed that the strongest relationship existed between the percentage of organic matter, clay, silt, sand, with the reflection in the visible light range and near infrared and the long-wave infrared panchromatic wavelength of the electromagnetic spectrum. Comparison of the power of correlation between the bands reflection and the percentage of organic matter in the section (NIR) shows greater correlation. Validation showed that the maps obtained by geostatistics and remote sensing well characterized the distribution of features.