Presenter: Matthias Aengenheyster (University of Oxford)

Description:
Climate change will have large impacts on societies partly through modified frequencies and magnitudes of extreme events. Large model ensembles with atmospheric models, forced with SSTs, are commonly used to investigate the factors that drive extreme events and attribute them to climate change. This approach lacks realistic representation of air-sea interactions, but coupled models are often prohibitively expensive. Here, we present the new HadSM4 model that is now functional on the climateprediction.net system, combining the HadAM4 atmospheric model at N144 resolution with a Slab ocean.  We use HadSM4 to investigate the dynamics during the 2013-14 winter which was characterized by strong ocean heat loss over the North Atlantic and wet conditions in southern England. Comparing ensembles of traditional SST-forced runs, Slab-forced runs and SST-forced runs with SST diagnosed from the Slab runs, each of >1000 members, we attempt to answer two questions: 1. How does the introduction of more physical air-sea interactions change the statistics of atmospheric dynamics and extreme events? 2. What is the impact of a change in Q-Flux (as could result from low-frequency oceanic variability) on atmospheric dynamics and extreme events? We find that employing the Slab configuration improves the representation of dynamic lead-lag relationships, e.g. between SSTs and the NAO, while not strongly altering the statistics of NAO and jet regimes themselves. A negative Q-Flux anomaly in the North Atlantic subpolar gyre (designed to remove the 2013-14 anomalous positive Q-Flux anomaly) is damped locally through air-sea interaction. It leads to reduced (extreme) winter precipitation over parts of Western Europe (e.g. in Western Germany a 20-year wet DJF becomes a 44-year event), implying that changing ocean heat convergence can be important for seasonal extreme events in Europe.