The United States and over 180 other signatories to the United Nation Framework Convention on Climate Change have committed to prevention of “dangerous anthropogenic interference with the climate system.” This dissertation expands the knowledge base necessary to reach this goal, presenting a broad definition of “dangerous anthropogenic interference” (DAI) and characterizations of specific categories of DAI. The magnitude and spatial distribution of future climate change are uncertain, and the first study presented here examines the range of potential “dangerous” climate impacts to produce a probabilistic metric for DAI that can be applied to projections of future climate change, demonstrating that application with a probabilistic analysis of the likelihood of DAI with and without climate policy controls based on the results of a simple integrated assessment model. Systematic evaluation of climate impacts and their potential for “danger” is a requirement to refine future analyses of the potential for DAI, and the remaining studies present such evaluation. Included studies examine abrupt nonlinear behavior and multiple equilibria within the climate system and in the interactions between social and natural systems, the increase in stringency of modeled climate policy controls when incorporating the potential damages from such abrupt climatic changes, and the detection of an unbroken chain of correlation from anthropogenic climate change to changes in species traits during the 20th century. The policy message from these studies is clear and consistent. Human-induced climate change is already creating measurable impacts, and action can be defended even though there is uncertainty in future changes. Future climate change will include important components that are not predictable, reversible, or gradual. Instead, it is likely to include some unexpected, irreversible, abrupt changes. Climate policy controls will work to slow the changes, increasing the potential for successful adaptation to climatic changes for humans and other species of the Earth, and reducing the probability of reaching critical thresholds of DAI.