Introduction. In 2016 a quarter of the ocean surface experienced either the longest or most intense marine heatwave (Hobday et al. 2016) since satellite records began in 1982. Here we investigate two regions--Northern Australia (NA) and the Bering Sea/Gulf of Alaska (BSGA)--which, in 2016, experienced their most intense marine heat waves (MHWs) in the 35year record. The NA event triggered mass bleaching of corals in the Great Barrier Reef (Hughes et al. 2017) while the BSGA event likely fed back on the atmosphere leading to modified rainfall and temperature patterns over North America, and it is feared it may lead to widespread species range shifts as was observed during the "Blob" marine heat wave which occurred immediately to the south over 2013-15 (Belles 2016; Cavole et al. 2016). Moreover, from a climate perspective it is interesting to take examples from climate zones with very different oceanographic characteristics (high-latitude and tropics). We demonstrate that these events were several times more likely due to human influences on the climate.
Data and methods. Observations consisted of sea surface temperatures (SSTs) from the daily NOAA OI SST v2 0.25[degrees] gridded dataset over 1982-2016 (Reynolds et al. 2007). We also used the in situ-based monthly HadISST Io gridded dataset over 1900-2016 (Kennedy et al. 2011a,b). SST time series were generated by spatially averaging over (20[degrees]-5[degrees]S, 110[degrees]-155[degrees]E) for NA and (50[degrees]-65[degrees]N, 178[degrees]-127[degrees]W,) for the BSGA (Figs. 9.1a,b, black boxes). Anomalies were calculated relative to a base period of 1961-90. Daily climatologies were calculated from NOAA OI SST over the period 1982-2005, and in order to reference this to the chosen base period, we offset by the mean warming from 1961-90 to 1982-2005 calculated from HadISST (+0.19[degrees]C for both NA and BSGA; see Oliver et al. 2017 for more details).
Marine heat waves were defined as periods when SSTs were above the seasonally varying 90th percentile for at least five consecutive days (Oliver 2015; Hobday et al. 2016). We considered two MHW metrics: duration (time between the start and end dates) and maximum intensity (peak temperature anomaly).
We employed Coupled Model Intercomparison Project Phase 5 (CMIP5; Taylor et al. 2012) global climate model simulations of historical and projected future climates. We used daily SST outputs from the historicalNat (representing historical conditions without anthropogenic influence; models are forced by natural volcanic and solar forcing only) and the historical and RCP8.5 experiments (representing historical conditions with anthropogenic influence; models include anthropogenic greenhouse gas and aerosol forcing in addition to natural forcing) from seven models (Table ES9.1). Model climatologies were calculated using a base period of 1961-90; RCP8.5 anomalies were defined relative to the historical run climatology. The nonseasonal daily SST variance (i.e., after removing the climatology) was bias-corrected for each model based on the ratio between the standard deviations of the daily observations and the daily historical runs (see Oliver et al. 2017 for more details). MHWs were then identified in all model experiments using the Hobday et al. (2016) definition.
The fraction of attributable risk (FAR) methodology (Lewis...
This is a preview. Get the full text through your school or public library.