The Intergovernmental Panel on Climate Change assessment gives five "reasons for concern" (below) to aid policy makers in making their own determination about what constitutes "dangerous" climate change- the IPCC does not decide what is dangerous. The IPCC leaves out the grandfather of all concerns: Arctic carbon feedbacks, also oceans, and food security.
The IPCC 2014 AR5 diagram clearly shows 2C is far beyond dangerous, 1.5C is beyond dangerous and only 1.0 can be called a danger limit (as by James Hansen et al on dangerous climate 2013).
Our climate has many positive (i.e.bad) feedbacks. They tend to be self-reinforcing. The negative feedback is from air pollution aerosols. That is no help because it represents additional warming that will be unmasked down the road of stopping air global pollution. 'The big question for climate scientists then is: What is the balance between the positive and negative feedbacks? A consensus appears to have emerged over the last two years – not yet reflected in the recent IPCC reports – that the positive feedbacks are much stronger and more numerous than the negative ones.'
— Thomas Homer-Dixon, Professor, Centre for Environment and Business, Faculty of Environment, University of Waterloo, Canada
Incredibly, the position of the IPCC (scientists and policy makers) is that "dangerous interference with the climate system" (as expressed in the United Nations Framework Convention on Climate Change) cannot be defined. However, they do list five "reasons for concern."
'Each of the IPCC reasons for concern, by itself or in combination with the others, speaks to aspects of climate change that are dangerous. Although the IPCC said that "there is growing evidence of the risks of very large impacts on multiple-century time scales'
Here, excerpted from the November 2007 Synthesis Report, are the IPCC reasons for concern:
The most dangerous of all effects of global warming is carbon feedbacks because by adding more carbon to the atmosphere, these feedbacks accelerate the warming and all impacts. In a vicious cycle, warming leads to feedbacks, which lead to further warming, which leads to more feedbacks, until a state of runaway global warming leading to global climate catastrophe is inevitable.
None of the carbon feedbacks is included in the models chosen by the IPCC for the crucial most likely global temperature increase projections. The feedback danger is not included in the IPCC reasons for concern.
The most dangerous carbon feedbacks by far are from thawing permafrost in the Arctic and subsea coastal methane hydrates, which will destabilize with global warming.
The most immediate dangers to large human populations and humanity are the many adverse effects of global climate change on agriculture. Food security or agriculture is not included as a Reason for Concern. It is only very captured by inference under Extreme Weather Events. How ever in the IPCC main chart of climate change impacts extreme weather events does get included. Clearly there needs to be strong voice for human rights and health protection in the assessments.
The other huge danger from the many impacts of global climate change is their additive effects. As these cannot be modelled, the IPCC does not list additive effects as a major reason for concern. Clearly, however, all additive effects will increase damages and risks to a huge degree. They will also be cumulative increasing over time.
The IPCC gives computer model ranges for all impacts. More committed impacts from IPCC AR5 WG2 are below: