Read here and here. As with every unsubstantiated, speculative prediction made by alarmists and climate models, there is a grain of truth regarding melting tundra/permafrost releasing stored carbon dioxide. But researchers have discovered that when these melted areas are thawed, the explosion of new growth of vegetation becomes a positive CO2 sink that sequesters carbon dioxide in greater quantities than that released from the thaw. So instead of permafrost melting being a positive warming feedback, it actually becomes a negative feedback - funny how the climate always seems to do that in the end.
"In light of these compelling observations, it would appear that even if global warming were to accelerate and reach a tipping point that led to the demise of much of the world's permafrost, the subsequent "terrestrialization" of these regions would actually lead to more carbon being stored in the soils and vegetation of these parts of the world, rather than -- as climate alarmist claim -- more being lost."
And what about the release of methane (CH4) from melting?
Just another reason why climate models are not an accurate representation of how the real-world climate actually works.
"Delisle throws in another fast ball regarding methane (CH4) at the end of the article by stating “A second, rarely touched upon question is associated with the apparently limited amount of organic carbon that had been released from permafrost terrain in previous periods of climatic warming such as e.g. the Medieval Warm Period or during the Holocene Climatic Optimum. There appear to be no significant CH4-excursions in ice core records of Antarctica or Greenland during these time periods which otherwise might serve as evidence for a massive release of methane into the atmosphere from degrading permafrost terrains.”"