Characterization And Partitioning Of CH4 And CO2 Eddy Flux Data Measured at NGEE-Arctic Sites

The high latitudes are experiencing rapid warming with permafrost ecosystems being highly vulnerable to this change. Since the advancement in Eddy Covariance (EC) measurements, the number of high latitude sites measuring greenhouse gases and energy (CO2, CH4 and H2O) fluxes is steadily increasing, with new sites being established each year. Data from these sites are not only valuable for annual carbon budget calculations, but also vital to the modeling community for improving their predictions of emission rates and trends. CH4 flux measurements are not as straightforward as CO2 fluxes. They tend to be less predictable or as easily interpretable as CO2 fluxes. Understanding CH4 emission patterns are often challenging. Moreover, gas flux fluctuations are spatially and temporally diverse, and in many cases event-based. An improvement in understanding would also contribute to improvements in the fidelity of model predictions. These rely on having high quality data, and thus will entail developing new QA/QC and gap-filling methods for Arctic systems, in particularly for CH4. Contributing to these challenges is the limited number of ancillary measurements carried out at many sites and the lack of standardized data processing, QA/QC, and gap-filling procedures, in particular for CH4. CO2, CH4, and energy flux measurements are ongoing at, both NGEE-Arctic/AmeriFlux, US-NGB (Arctic coastal plain), and US-NGC (subarctic tussock tundra) sites. The sites, with underlying continuous permafrost, show a high degree of inter-annual and seasonal variability in CH4 fluxes. In order to interpret this variability, we apply a variety of models, such as footprint characterization, generalized additive models, as well as artificial neural networks, in an attempt to decipher these diverse fluxes, patterns and events.