Proxies for soil organic carbon derived from remote sensing

•An approach to estimate Soil Organic Carbon (SOC) in subtropical forest soils from remote sensing based proxies was introduced.•Two regression models were fitted.•The remotely sensed variables consisted of elevation, above ground biomass (based on a novel method using object based classification) and forest type.•The model with only remotely sensed measured variables have a bit higher AIC and explain 68% variation of SOC.