Cold‐Season Methane Fluxes Simulated by GCP‐CH<sub>4</sub> Models
Akihiko Ito, Tingting Li, Zhangcai Qin, Joe R. Melton, Hanqin Tian, Thomas Kleinen, Wenxin Zhang, Zhen Zhang, Fortunat Joos, Philippe Ciais, Peter O. Hopcroft, David J. Beerling, Xiangyu Liu, Qianlai Zhuang, Qing Zhu, Changhui Peng, K.‐Y. Chang, Etienne Fluet‐Chouinard, Gavin McNicol, Prabir K. Patra, Benjamin Poulter, Stephen Sitch, W. J. Riley, Qing Zhu
Abstract
Abstract Cold‐season methane (CH 4 ) emissions may be poorly constrained in wetland models. We examined cold‐season CH 4 emissions simulated by 16 models participating in the Global Carbon Project model intercomparison and analyzed temporal and spatial patterns in simulation results using prescribed inundation data for 2000–2020. Estimated annual CH 4 emissions from northern (>60°N) wetlands averaged 10.0 ± 5.5 Tg CH 4 yr −1 . While summer CH 4 emissions were well simulated compared to in‐situ flux measurement observations, the models underestimated CH 4 during September to May relative to annual total (27 ± 9%, compared to 45% in observations) and substantially in the months with subzero air temperatures (5 ± 5%, compared to 27% in observations). Because of winter warming, nevertheless, the contribution of cold‐season emissions was simulated to increase at 0.4 ± 0.8% decade −1 . Different parameterizations of processes, for example, freezing–thawing and snow insulation, caused conspicuous variability among models, implying the necessity of model refinement.