Forecasts from the deep past: using paleoclimate analogues to inform climate projections
OA Version
Citation
Abstract
This dissertation lies at the intersection of philosophy of climate science, philosophy of the historical sciences, philosophy of model-based science, and philosophy of data. I focus on epistemological and methodological issues concerning how scientists use episodes of climate change in Earth’s deep past that significantly resemble incipient climate change ("paleoclimate analogues") to inform their climate projections. I argue that we should conceptualize paleoclimate analogues as kind of climate model - a naturally-occurring, full-size, concrete model – and that viewing paleoclimate analogues through this lens has important philosophical implications. It follows, for instance, that the usefulness of models does not depend on our ability to experimentally manipulate them (we cannot manipulate the deep past). I also argue that the historical sciences can be predictive: they can use the past as a guide to the future. I give several recommendations for scientists, including some guidance on how to manage, process, and evaluate paleoclimate data in order for it to be used to inform our climate projections, despite various sources of error and uncertainty inherent to our reconstructions of the deep past. I close with some general reflections on "using the past as a guide to the future," within the context of paleoclimate analogues and beyond.
Description
2023