Reliable national estimates of CH4 emissions from natural wetlands depend on model validation based on site observations. We therefore evaluated the performance of the CH4MODwetland model in simulating CH4 emissions from 11 representative wetland sites in five regions of China. Model performance analysis showed that this method effectively simulates differences in the CH4 fluxes between different sites and regions. The model efficiency for estimating the daily CH4 fluxes in the northeastern China (NE), Inner Mongolia and northwestern China (NW), the North China plain and the Middle-Lower Yangtze Plain (E) and the Qinghai Tibetan Plateau (SW) was 0.51, 0.20, 0.52 and 0.65, respectively. The efficiency for estimating the annual mean CH4 fluxes in southern China (S) was 0.99. Systematic negative deviation between the simulated and observed CH4 emissions existed in all regions, especially in the NW region, which had a mean deviation (RMD) value of -36.7%. On the national scale, the root mean square error (RMSE), the RMD, the model efficiency (EF) between the simulated and observed seasonal values were 28.7%, -7.8% and 0.93, respectively. The CH4 emissions showed the highest sensitivity to air temperature in the NE and SW regions, and to water table depth in the E region. Based on the sensitivity analysis, future climate warming and wetting are likely to increase the wetland CH4 emissions at different levels in all regions of China.
Keywords: CH4MOD(wetland); China; Model performance; Natural wetland.
Copyright © 2017. Published by Elsevier B.V.