Wednesday, October 04, 2017, 10:30AM - 11:30AM
SPSC, East Campus
In the past decade or so, a technique has arisen which purportedly solves the problem of how to evaluate the realism of future climate simulations. Since observations of future climate obviously do not yet exist, so-called "emergent constraints" rely instead on strong simulated relationships between an observable element of current climate and a relevant aspect of future climate. Such relationships are seen to "emerge" from ensembles of climate simulations organized by international scientific bodies. Models can be evaluated in how well they simulate the element of the current climate, which presumably allows for inferences about their ability to simulate the relevant aspect of the future. In this talk, I will survey the emergent constraints that have been proposed, and discuss their credibility and meaning. I will discuss how the technique can (and cannot) be used to reduce climate model spread and weight models. Finally, I will also assess the near-term prospects for constraining critical elements of future climate change, including physical and biogeochemical feedbacks.