PattiMichelle's Scientific Research

Discussion
It was somewhat surprising when the post-transient data shown in the figure entitled “Global Land Surface Mean Annual Temperature Change” was averaged for the various scenarios and gave rise to the plot entitled “Global Land Surface Annual Data, Averaged 2031-2039.” That is, a simple linear fit to the model results yielding, with low variance, a mean temperature rise coefficient of 0.01 C per gigaton of added methane. It is also remarked that this was the first and only analysis attempted, and was chosen “by eye” to be the region of approximate temperature stabilization following the transient introduction of uniformly dispersed methane into the earth’s atmosphere. Simplified experiments such as this can have a useful clarity.

The clarity of the estimate of this coefficient is likely at least partially a result of the simplicity of the model input conditions. The spatially uniform distribution of methane in the atmosphere is not highly-realistic and likely contributes to the simplicity and linearity of the result. Both the Arctic/taiga snow/ice fields and the ocean surface temperatures in this model are prescribed (and therefore do not change in consecutive years); however, it may be unlikely that the ocean and arctic can respond at the same short time scales as the atmosphere given a large methane release. A more complete analysis would likely incorporate a fully-coupled AOGCM; however, it is not clear that existing AOGCMs could reliably model the Arctic/taiga snow/ice responses to a sudden methane increase given the failure of the AR5 models to accurately predict well-observed changes in the Arctic snow/ice.

It should also be noted that in this experiment, the post-transient stabilization of the atmospheric methane burden implies that additional methane is released at a slower rate following the rapid transient release in order to maintain the atmospheric methane concentration since the model chemistry continuously degrades methane, for instance, into carbon dioxide and water. Such additional releases are not an unphysical scenario for a general atmospheric circulation model as the Arctic oceans continue to warm following the phase change from ice-cover to open-ocean.

The effect of decreasing Arctic insolation ~20% is also of interest given the recent rapid decrease in Arctic and taiga snow and ice and the importance of the polar regions to global weather patterns and circulations. A slight Arctic mean land surface cooling was observable in the data when the annual mean data points were boxcar-averaged over at least 4 years, and was observable only in the Arctic.