Methane variations on orbital timescales: a transient modeling experiment

Methane (CH<sub>4</sub>) variations on orbital timescales are often associated with variations in wetland coverage, most notably in the summer monsoon areas of the Northern Hemisphere. Here we test this assumption by simulating orbitally forced variations in global wetland emissions, using a simple wetland distribution and CH<sub>4</sub> emissions model that has been run on the output of a climate model (CLIMBER-2) containing atmosphere, ocean and vegetation components. The transient climate modeling simulation extends over the last 650 000 yr and includes variations in land-ice distribution and greenhouse gases. Tropical temperature and global vegetation are found to be the dominant controls for global CH<sub>4</sub> emissions and therefore atmospheric concentrations. The relative importance of wetland coverage, vegetation coverage, and emission temperatures depends on the specific climatic zone (boreal, tropics and Indian/Asian monsoon area) and timescale (precession, obliquity and glacial-interglacial timescales). Despite the low spatial resolution of the climate model and crude parameterizations for methane production and release, simulated variations in CH<sub>4</sub> emissions agree well with those in measured concentrations, both in their time series and spectra. The simulated lags between emissions and orbital forcing also show close agreement with those found in measured data, both on the precession and obliquity timescale. We find causal links between atmospheric CH<sub>4</sub> concentrations and tropical temperatures and global vegetation, but only covariance between monsoon precipitation and CH<sub>4</sub> concentrations. The primary importance of the first two factors explains the lags found in the CH<sub>4</sub> record from ice cores. Simulation of the dynamical vegetation response to climate variation on orbital timescales would be needed to reduce the uncertainty in these preliminary attributions