Predicting soil organic carbon concentrations in a low relief landscape, eastern Iran

• Published in: Geoderma Regional Vol. 15; p. e00195
• Main Authors: Pahlavan-Rad, Mohammad Reza, Dahmardeh, Khodadad, Brungard, Colby
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.12.2018
• Subjects: agricultural management; clay; clay fraction; Digital soil mapping; environmental factors; Iran; land use; Landsat; landscapes; planning; prediction; Random forest; remote sensing; rivers; soil organic carbon; Soil organic carbon (SOC); soil sampling; soil surveys; soil texture; topography; wind erosion; Zahak; Iran; Random forest; Zahak; Digital soil mapping; Soil organic carbon (SOC)
• ISSN: 2352-0094; 2352-0094
• DOI: 10.1016/j.geodrs.2018.e00195

Understanding the spatial distribution of soil organic carbon (SOC) concentrations in arid areas is necessary for the management of these regions. In the present study, the distribution of SOC concentrations was investigated using digital soil mapping for an area of ~41,000 ha in Zahak county, Sistan and Baluchestan province, eastern Iran. 417 soil samples were collected from 0 to 30 cm depth across a 750 m by 750 m grid. Random Forest (RF) technique was used to model soil organic carbon (SOC) concentrations. Topographic and remote sensing indices, and land use and soil texture fractions maps were used as environmental variables. 10-fold cross-validation was used to evaluate models. Results showed that the mean SOC concentration is very low with a value of 0.33%. The most important variables for predicting SOC concentration variations were the clay map, Landsat 8 bands 7, 4, 6, 3 and distance from the river. Remote sensed data showed high potential to predict SOC in this low relief region. Root mean square error (RMSE) and mean absolute error (MAE) were 0.21 and 0.16%, respectively. We concluded that soil clay content, agricultural management, and the processes of wind erosion and deposition determine the SOC distribution in Zahak region, eastern Iran. The predicted map is consistent with the realities of the study area. These maps could be used in comprehensive planning to improve SOC in the study area.