Data on the concentration of particulate organic carbon (POC) and its endmembers provide a basis for the characterisation of lake biogeochemical cycles. Here, a novel remote sensing strategy (the SCPOC algorithm) was developed to determine total POC concentrations, as well as terrestrial and endogenous POC concentrations in lakes. This strategy provides a successful example for the combination of isotope tracer and remote sensing technology. First, we obtained the terrestrial and endogenous POC concentration at the sampling point based on isotope tracing technology. Afterwards, we established a relationship between the phytoplankton absorption coefficient and the endogenous POC concentration (Cend), and applied a semi-analytical algorithm to invert the Cend value. Finally, the POC source ratio model and Cend value were combined to obtain the POC concentration (CPOC) and terrestrial POC (Cter). The results of synchronisation verification based on ocean and land colour instrument (OLCI) images show that the SCPOC algorithm has high Cend, Cter, and CPOC inversion accuracy, with MAPE values of 26.07%, 30.43%, and 42.28%, respectively. In fact, the SCPOC algorithm not only improved the accuracy of lake POC mapping, but also fills the gap of optical retrieval of POC endmember concentrations. Additionally, data from the OLCI images indicated that the studied lakes were dominated by external POC. However, because of the greater contribution of algal blooms to POC, this dominant advantage weakens in summer, although the terrestrial organic carbon carried by rainfall runoff also affects lake POC composition. Different POC sources have different ecological roles in lakes, and the superior POC end-element estimation capability of the SCPOC algorithm can not only be used as a supplement to traditional tracing methods, but also provides accurate spatial data for lake management.