Introduction to the Special Issue on Climate Adaptation: Improving the connection between empirical research and integrated assessment models

1 The U.S. Department of Energy, Office of Science, Biological and Environmental Research Program, Integrated Assessment Program, Grant No. DE-SC0005171. Integrated assessment models (IAMs), models that couple the human and natural systems, have been widely used by the climate change research community to project the emissions consequences of economic activity and the technical potential and cost of mitigation options; to perform cost-benefit analysis (CBA) to determine the optimal future path of GHG emissions and mitigation costs; and to assess the magnitude and incidence of climate impacts and associated economic cost of climate damages. DICE (Nordhaus, 1994), the first fully coupled IAM to account for the feedbacks of climate change on the economy, introduced the device of a climate damage function which was global in scope but with a highly simplified and aggregated treatment of either the meteorological drivers of impacts (global mean temperature change) or their physicalmanifestations across different endpoints, economic sectors and geographic regions. In the intervening two decades a succession of IAMs has followed this lead, incorporating climate feedbacks with limited complexity—or more commonly disregarding them altogether, even as computational advances have made possible increasingly detailed representations of the economic activity to which climate change poses risks. As a consequence there has been slow progress in modeling climate adaptation responses, and, to the best of our knowledge, no study has accounted for the implications of impacts and adaptation for the climate stabilization strategies. Omitting climate change impacts and adaptation responses from IAM studies can affect their results in important ways. For one, climate feedbacks can reduce the pool of mitigation options which can understate the cost of mitigation policies; e.g., water shortages can limit the potential for hydropower and biofuel production could be limited by negative impacts on crop productivity. Omitting adaptation responses from the analysis can also bias the results in a number of ways; e.g., (a) adaptation responses could dampen the economic cost of climate impacts; (b) adaptation responses could change the baseline emissions trajectory (e.g., greater air conditioning in response to higher temperatures could lead to higher emissions), making it more difficult to reach stabilization targets; and (c) adaptation investments could crowd out mitigation investment making it more costly to mitigate. Given the importance of climate feedbacks and adaptation responses, a natural question is why have they been largely left out of IAMs? A reason that is commonly given is the inadequate empirical basis for characterizing the responses to be inserted into models. For the few studies that do attempt to incorporate these responses, engineering or natural science process simulations are often used to fill the gap; however, the latter models typically are computationally expensive, capture only a single link of the causal chain from meteorology to