Abstract/Summary

In 2012, extreme anomalous climate conditions occurred around the globe. Large areas of North America experienced an anomalously hot summer, with large precipitation deficits inducing severe drought. Over Europe, the summer of 2012 was marked by strong precipitation anomalies with the UK experiencing its wettest summer since 1912 while Spain suffered severe drought. What caused these extreme climate conditions in various regions in 2012? This study compares attribution conclusions for 2012 climate anomalies relative to a baseline period (1964–1981) based on two sets of parallel experiments with different model configurations (with coupling to an ocean mixed layer model or with prescribed sea surface temperatures) to assess whether attribution conclusions concerning the climate anomalies in 2012 are sensitive to the representation of air-sea interaction. Modelling results indicate that attribution conclusions for large scale surface air temperature (SAT) changes in both boreal winter and summer are generally robust and not very sensitive to air-sea interaction. This is especially true over southern Europe, Eurasia, North America, South America, and North Africa. Some other responses also appear to be insensitive to air-sea interaction: for example, forced increases in precipitation over northern Europe and Sahel, and reduced precipitation over North America and the Amazon in boreal summer. However, the attribution of circulation and precipitation changes for some other regions exhibits a sensitivity to air-sea interaction. Results from the experiments including coupling to an ocean mixed layer model show a positive NAO-like circulation response in the Atlantic sector in boreal winter and weak changes in the East Asian summer monsoon and precipitation over East Asia. With prescribed sea surface temperatures, some different responses arise over these two regions. Comparison with observed changes indicates that the coupled simulations generally agree better with observations, demonstrating that attribution methods based on atmospheric general circulation models have limitations and may lead to erroneous attribution conclusions for regional anomalies in circulation, precipitation and surface air temperature.