Der erste Teil dieser Arbeit konzentriert sich auf die systematische Synthese von Informationen über Anpassung, Anpassungsstrategien, Einschränkungen und Grenzen auf globaler Ebene durch die Aufarbeitung der wissenschaftlichen Literatur und mit Hilfe von Techniken des maschinellen Lernens. In einem zweiten Schritt werden mittels einer Online-Umfrage und halbstrukturierter Interviews die Wahrnehmungen von Einschränkungen der Anpassungsmöglichkeiten in den kleinen Inselstaaten der Karibik bewertet, wobei der Schwerpunkt auf den Wechselwirkungen zwischen verschiedenen Arten von Einschränkungen und dem erheblichen Einfluss der fehlenden Anpassungsfinanzierung liegt. Schließlich wird ein zukunftsorientierter Ansatz verfolgt, bei dem Bottom-up- und Top-down-Daten kombiniert werden, um zu veranschaulichen, wie sich sozioökonomische Dimensionen im Zusammenhang mit Einschränkungen bis zum Jahr 2100 unter verschiedenen Szenarien der künftigen sozialen und wirtschaftlichen Entwicklungen entfalten könnten. Die Ergebnisse zeigen, dass die Anpassung derzeit schrittweise und weitgehend fragmentiert erfolgt. Die Erkenntnisse über die Anpassungspolitik nehmen rasch zu, aber es bestehen weiterhin geografische Ungleichheiten. Finanzen und Regierungsführung sind weltweit die größten Einschränkungen, wobei die kleinen Inselstaaten sowie Mittel- und Südamerika die meisten Einschränkungen und Grenzen melden. Die Wechselwirkungen zwischen den Einschränkungen können zu Grenzen und zusätzlichen Verlusten und Schäden führen. Dies unterstreicht die Notwendigkeit, die Einschränkungen parallel anzugehen und eine nachhaltige und gezielte Anpassungsfinanzierung zu gewährleisten. Mögliche Szenarien für die künftige Entwicklung zeigen, dass selbst in den optimistischsten Szenarien sozioökonomische Schlüsseldimensionen wie schwache Regierungsführung die Anpassung bis weit in die zweite Hälfte des 21. Jahrhunderts hinein erschweren werden. / The first part of this thesis focuses on systematically synthesising information on adaptation, adaptation policies, constraints and limits on the global level by reviewing scientific literature and with the support of machine learning techniques. Secondly, through an online-survey and semi-structured interviews, perceptions of constraints in Caribbean Small Island Developing States are assessed, by focusing on the interactions between different types of constraints and the significant influence of the lack of adaptation finance. Finally, a forward-looking approach combining bottom-up and top-down data is taken to illustrate how socio-economic dimensions related to constraints could evolve by 2100, under various scenarios of future development. The results show that adaptation is currently incremental and largely fragmented. Evidence on adaptation policy is rapidly growing but geographic inequalities persist. There is negligible data on adaptation reducing climate change risks. Finance and governance are found to be the most prominent constraints globally, with Small Island Developing States and Central and South America reporting most constraints and limits. Indeed, Caribbean Small Island Developing States face large financial constraints which in turn closely interact with governance, human capacity and information constraints and result in vicious circles. Interactions between constraints can lead to limits and additional losses and damages. This highlights the need to address constraints in parallel and for sustained and dedicated adaptation finance. Potential scenarios of future development show that, even under the most optimistic scenarios, key socio-economic dimensions such as weak governance will challenge adaptation well into the second half of the 21st century. The persistence of constraints, in particular in the most vulnerable regions, calls for stringent mitigation, improved adaptation and increased efforts to address losses and damages.

Agricultural adaptation to climate change is often simulated by changes in land use over time. Land use is commonly optimised in economic models, which rests on the neoclassical economic assumption of rational choice among farmers. A wealth of experimental and empirical evidence demonstrates that rational choice can be a poor approximation of human decision making. Models simulating adaptation by optimising producers’ behaviour are in effect simulating adaptive potential. Much evidence demonstrates that adaptive potential does not necessarily translate into adaptation. This investigation focuses on the ways by which farmers’ real-world adaptive behaviours depart from those assumed by the dominant economic models of agricultural responses to climate change. These departures are characterised as adaptation constraints, and they are assessed through an empirical case study of adaptive behaviours in the Hikurangi catchment, New Zealand. Data are collected using a mixed methodology comprising an extensive survey of rural decision making, to which this study contributes, and a suite of semi-structured interviews. The interviews give an understanding of the origins and processes of adaptation constraints, while the surveys provide information about the extent to which they impact adaptive propensity. These adaptation constraints are then formalised as mathematical rules and written into an existing agent-based model of land use change, which is substantially modified for the purposes of this study. Different combinations of constraints are then tested in order to produce estimates of their economic impacts. The constraints on adaptation are found to significantly reduce profits relative to a specification that assumes rational choice among farmers. This is understood to be the first empirically derived estimate of the extent of the adaptation deficit. The size of the deficit identified in this study implies that current economic models are likely to significantly underestimate the costs of adaptation to climate change, the benefits of climate change mitigation, and the residual loss and damage climate change will cause.