Spatial Heterogeneity of Greenhouse Gases at Howland Forest, Maine

Abstract:

Natural and human emissions of greenhouse gases (GHG’s)- carbon dioxide (CO₂), methane gas (CH₄), and nitrous oxide (N₂O)- are important contributors to climate change. These GHG’s are released naturally from forested soils. Carbon dioxide flux tends to have a relatively low coefficient of variation (CV) compared to CH₄ and N₂O. Methane and nitrous oxide are often characterized by “hot spots” and “hot moments”, thus their CV are often very high compared to CO₂, even in seemingly homogenous soil environments. The objective of this study is to determine the sampling intensity necessary for confidence estimates from forested soils of all three of these important GHGs. On July 2015, the emission of GHGs was measured from 36- randomly distributed soil flux- collars at the Howland Forest, Maine. A data analysis was run to determine how many chamber fluxes are needed to adequately estimate the mean and variance of GHGs fluxes. It was determined that both CO₂ and CH₄ had a 100% confidence estimate that a random 10,000 randomly sample group of an example of 8 fluxes had means within 50% of the full population mean and 14% confident for N₂O. The confidence estimate that CO₂, CH₄, and N₂O had means within 25% the full population mean was 97%, 92% and 9% and about 62%, 57% and 3% had means within the 10% full population mean. This analysis can be used to guide researchers when determining how many flux measurements may be necessary to achieve a desired level of confidence in their GHG estimates for their site.