2024, Vol. 7, Special Issue 9

author(s)
Amarnath AT, Gopal V Dasar and Gowri B Gowda

abstract
The amount of soil organic carbon (SOC) varies with time and space and essential for plant growth and productivity. The study was aimed for Digital Soil Mapping (DSM) of soil organic carbon stock in sirsi forest division using appropriate remote sensing techniques which is labour and time efficient and identifying patterns, trends and potential factors influencing variations in the soil organic carbon. A 30 soil samples were collected from different forest types in two depths (0-30, 0 – 30 – 60 cm) using stratified random sampling method. Estimated and geospatially mapped soil organic carbon stock among different depths (0-30, 30-60 and 0-60 cm) by using Random Forest (RF) algorithm model in R studio software with correlating to the different environmental and spectral indices which are derived Sentinel 2A satellite data using Google Earth Engine (GEE). The RF model predicted SOC stock (0-30 cm) ranging from 59.45 Mgha^-1 to 76.59 Mgha^-1, with an average of 68.02 Mgha^-1.In 30 to 60 cm the SOC stock was ranging from 53.4 Mgha^-1 to 71.87 Mgha^-1, with an average of 62.63 Mgha^-1.The total SOC stock from 0 to 60 cm is ranging from 114.75 Mgha^-1 to 145.89 Mgha^-1, with an average of 130.3 Mgha^-1.The SOC stock prediction map for 0-30cm, 30-60 cm and 0-60 cm soil depth indicated higher SOC content in the western parts as compared to the eastern parts, because of dense and abundant vegetation with favorable climatic conditions in western parts of the study area.