Title: Reducing Uncertainty Surrounding Climate Change Using Emergent Constraints
Abstract: Many aspects of future climate change remain highly uncertain as different climate models can exhibit large and persistent spreads in their projections. This uncertainty stems from a variety of sources including natural variability, differing emissions pathways, and differences in how models represent key physical processes. Efforts to reduce uncertainty in climate change projections primarily focus on the latter. A popular approach for this task is emergent constraints (ECs), which use strong intermodel relationships between historical and future climate metrics in conjunction with observations to narrow model spread. Here, two recent examples are discussed. The first one targets the large uncertainty in projected changes to global precipitation extremes. We find that the observed trend in heavy rainfall frequency over recent decades can be used to substantially narrow this model spread. Second, we focus on cloud feedback, which remains the primary source of model uncertainty in projections of future global warming. We show that climatological biases in extratropical cloud properties can be used to constrain tropical mean cloud feedback across three generations of models. Lastly, I will also briefly discuss some ongoing analysis, which seeks to develop ECs for various aspects of regional climate change across the western United States (e.g. flood timing, drought).