B043-03
FLUXNET-CH4 and seasonality of methane emissions from freshwater wetlands
FLUXNET-CH4 and seasonality of methane emissions from freshwater wetlands
Wednesday, 9 December 2020: 19:08
Virtual
Abstract:
Methane seasonality has not been characterized across wetlands globally, despite wetlands producing roughly half of all natural methane emissions. Methane seasonality, i.e., the start and end of seasonally-elevated emissions, peak timing, and magnitude, can vary considerably across wetlands and is expected to be driven by changes in temperature, moisture, and substrate availability. We have established a global, open-access database named FLUXNET-CH4, with eddy covariance data from 83 sites (including 47 freshwater wetlands). Here we use FLUXNET-CH4 data to provide the first assessment of global variability and predictors of wetland CH4 flux seasonality. For extra-tropical wetlands, we find that the spring onset of increasing methane emissions starts about 3 days earlier for every +1 degree C mean annual temperature, and seasonally elevated emissions last about 3 months longer in the warmest wetlands compared with the coolest. On average, the onset of methane emissions lags soil warming by one month, while roughly half the sites experience the spring onset of rising methane emissions prior to the spring increase in gross primary productivity. In contrast, the timing of peak seasonal methane emissions do not correlate with either peak temperature or peak gross primary productivity timing. We also find that tropical systems, though not as well represented in FLUXNET-CH4, have unique seasonality patterns that differ substantially from temperate and northern wetlands. For example, they have a lower ratio of amplitude to peak CH4 flux, and much wider variability in season start and peak timing. Our results address knowledge gaps in broad-scale variability and predictors of wetland CH4 seasonality and provide insights to guide future modeling efforts. We acknowledge the FLUXNET-CH4 contributors for the data provided in these analyses.