Accurate carbon storage estimation for the salt marsh ecosystem based on Bayesian maximum entropy approach - A case...

The blue carbon ecosystem, including the salt marsh ecosystem, possesses a significant carbon sequestration potential. Therefore, accurately quantifying the carbon storage within such ecosystems is crucial for the adequate accounting of carbon sequestration. The present work chose a Spartina alterniflora ecosystem in the Xiaogan Island (China) as the study area (approximately 11 ha), and employed the Bayesian maximum entropy (BME) approach to assimilate both hard organic carbon (OC) data and soft OC data measured from 2 cm and 10 cm stratified samples. A 3-dimensional model was developed for space-time OC estimation purposes based on the sediment chronology results. The 10-fold BME cross validation results demonstrated a high estimation accuracy, with the R², RMSE and MAE values equal to 0.8564, 0.1026 % and 0.0748 %, respectively. A noteworthy outcome was the BME-generated carbon storage density maps with 1 m spatial resolution. These maps revealed that the carbon storage density at the top 30 cm sediment depth in the stable zone (with elder stand age of S. alterniflora) was higher than that in the rapid expansion zone, i.e., 71.79 t/ha vs. 69.82 t/ha, respectively. Additionally, the study found that the averaged carbon burial rate and the total carbon storage at the top 30 cm sediment depth across the study area were 266 g C/m²/yr and 781.50 t, respectively. Lastly, the proposed BME-based framework of carbon storage estimation was found to be versatile and applicable to other blue carbon ecosystems. This approach can foster the development of a standardized carbon sink metrological methodology for diverse blue carbon ecosystems.