In order to examine the underlying longer-term trends in greenhouse gases, that are driven for example by anthropogenic emissions or climate change, it is useful to remove the recurring effects of natural cycles and oscillations on the sources and/or sinks of those gases that have strong biological (e.g., CO2, CH4, N2O) and/or photochemical (e.g., CH4) influences on their global atmospheric cycles. We use global observations to calculate monthly estimates of greenhouse gas levels expressed as CO2 equivalents, and then fit these estimates to a semi-empirical model that includes the natural seasonal, QBO, and ENSO variations, as well as a second order polynomial expressing longer-term variations. We find that this model provides a reasonably accurate fit to the observation-based monthly data. We also show that this semi-empirical model has some predictive capability; that is it can be used to provide a reasonably reliable estimate of CO2 equivalents at the current time using validated observations that lag real time by a few to several months.

For more information, see the Carbon Counter Project.