Field-scale simulation of methane emissions from coastal wetlands in China using an improved version of CH4MOD(wetland)

Coastal wetlands are important CH4 sources to the atmosphere. Coastal wetlands account for similar to 10% of the total area of natural wetlands in China, but the size of this potential CH4 source remains highly uncertain. We introduced the influence of salinity on CH4 production and CH4 diffusion into a biogeophysical model named CH4MOD(wetland) so that it can be used in coastal wetlands. The improved model can generally simulate seasonal CH4 variations from tidal marshes dominated by Phragmites and Scirpus. However, the model underestimated winter CH4 fluxes from tidal marshes in the Yellow River Delta and YanCheng Estuary. It also failed to capture the accurate timing of the CH4 peaks in YanCheng Estuary and ChongMing Island in 2012. The improved model could generally simulate the difference between the annual mean CH4 fluxes from mangrove sites in GuangZhou and HaiKou city under different salinity and water table depth conditions, although fluxes were systematically underestimated in the mangrove site of HaiKou city. Using the improved model, the seasonal CH4 emissions simulated across all of the coastal wetlands ranged from 0.1 to 44.90 g m(-2), with an average value of 7.89 gm(-2), which is in good agreement with the observed values. The improved model significantly decreased the RMSE and RMD from 424% to 14% and 314% to -2%, respectively, and improved the EF from -18.30 to 0.99. Model sensitivity analysis showed that CH4 emissions were most sensitive to P-ox in the tidal marshes and salinity in the mangroves. The results show that previous studies may have overestimated CH4 emissions on a regional or global scale by neglecting the influence of salinity. In general, the CH4MOD(wetland) model can simulate seasonal CH4 emissions from different types of coastal wetlands under various conditions. Further improvements of CH4MOD(wetland) should include the specific characteristics of CH4 processes in mangroves to decrease the uncertainty in estimating regional or global CH4 emissions from natural wetlands. (C) 2016 Elsevier B.V. All rights reserved