There's an ongoing debate in the scientific community regarding the threshold value, tipping point, for frozen grounds in the Arctic, permafrost, to start thawing irreversibly. And whether released methane from the permafrost will occur gradually over time or more abruptly. There is more stored carbon in frozens soils than we currently have in the atmosphere.

There are basically two camps, some believe the permafrost to be stable with a threshold value around <3°C while others claim 1,5°C is enough to start thawing large parts of the frozen grounds and lakes in the Arctic. 

For a lay person this is quite confusing, but it simply means that there isn't enough data to know for sure and so some scientists are more or less conservative in their estimates. Then there is the question of using climate models to try and predict potential threshold values or doing actual fieldwork and extrapolating conclusions from that. To my knowledge, climate models have a pretty bad track record of capturing highly non-linear dynamics in the climate system. For example, Arctic sea ice passed a tipping point in 2007 and is now in a death spiral but models had predicted sea ice to remain until the end of this century. Pretty high margin of error if you ask me. Also, we are learning that there seems to be differences in how permafrost soils and lakes thaw. 

According to a recent field research study funded by NASA of thermokarst lakes, formed by thaw of permafrost below the soil, in Alaska and Siberia the potential for abrupt thaw (decades) is now likely and irreversible. As the Arctic warms more of these lakes are appearing and growing in size which expands the thaw below. It has been estimated that they now cover about 20% of northern permafrost regions. This could double the release from terrestrial landscapes by the 2050s. A carbon cycle feedback that is not yet included into climate models.

This is bad news for climate change mitigation efforts. This feedback is significant because methane is about 30 times more potent than carbon dioxide as a heat-trapping gas. And the lakes are expected to thaw even under the lowest IPCC emissions scenario, adding further warming. Since we most likely are already committed, warming yet to come from current emissions, to 1,5-2°C this extra warming from the permafrost reinforcing feedback could take us above the 2C threshold for potentially catastrophic warming. Unless we rapidly decarbonize our economy and try to take out carbon from the atmosphere by for example large-scale reforestation efforts. Time is not on our side. We need a climate emergency plan.

Permafrost is soil at or below the freezing point of water 0 °C that stays frozen for two or more years. Permafrost comprises 24% of the land surface in the Northern Hemisphere and can be found at Arctic ocean shelves and floor. It contains large quantities of trapped greenhouse gases such as methane and carbon dioxide, and is usually regarded as a carbon sink. Researchers have become increasingly worried that with climate change large permafrost areas could start to thaw and ultimately melt, releasing massive amounts of carbon that would exacerbate global warming. 

A small team of scientists working in Greenland have now found evidence that as microbes become active in permafrost, they produce heat, which can increase the rate of permafrost melt. In their paper published in Nature Climate Change, Hollesen et al. (2015) describes computer simulations that showed possible impacts of microbe activation in permafrost areas. Previous attempts at predictions for permafrost melt through modeling are now looking like they will have to be revised. 

Suspecting that microbes in the soil might have an impact on warming permafrost, Hollesen et al. collected 21 samples of permafrost soil from six locations across Greenland. They then exposed the samples to different temperatures in a laboratory. By monitoring the heat production from microbes they were able to gather enough information to create a computer simulation. That simulation revealed that as global temperatures rise, a feedback loop occurs in permafrost areas. Heat causes melting which stimulates the microbes that start decomposing organic material, and producing heat, which adds to the increased temperatures, on and on until the permafrost melts, releasing massive amounts of carbon into the atmosphere far earlier than previous models have predicted. 

One immediate consequence of thawing permafrost in Greenland is the potential for destruction of unexcavated archeological findings. The National Museums of Denmark and Greenland have now started several projects where decomposing wooden artifacts and bones from the first people on Greenland will help identify areas most threatened. This is mainly happening because the average temperature has risen by 2-3°C in Greenland. Thawing of protective permafrost leads to archaeological material rotting because the amount of oxygen rises and the decomposition process accelerate. Other concerns are related to coastal erosion, resettlement and infrastructure damage.

More long-term consequences of permafrost thaw is of course increasing greenhouse gas concentration in the atmosphere, and in turn a warmer climate. Previous estimates have pointed to 120 ± 85 Gigatonne of carbon emissions from thawing permafrost by 2100, which could increase global temperatures by 0.29 ± 0.21 °C. However, we now know that permafrost starts to thaw much earlier than expected, so we need to start including this knowledge into climate models or we risk overshooting the 2°C warming limit. For example, the most recent knowledge on permafrost/carbon feedbacks are not included into IPCC climate projections.