Graph of the Day: Distribution of temperature anomalies in Australia and the western tropical Pacific, 1881-2013Posted by Jim at Tuesday, October 14, 2014
September 2014 (BAMS) – Seasonal and annual temperature anomalies around the globe were highly skewed toward positive (warm) extremes in 2013, as in the recent few decades. Although global warming has been described as “pausing” since 2000, global temperatures remain at anomalously high levels, and warm annual and seasonal temperature extremes continue to far outpace the occurrence of cold annual extremes. Two examples of regions with extreme (record or near-record) annual warmth during 2013 include much of Australia and a region of the far western tropical Pacific. In both regions, a contribution of anthropogenic forcing to an observed long-term warming trend was detected. The annual mean anomalies for 2013 were either completely outside of, or extremely rare in, the distributions of modeled natural variability. Thus, the fraction of risk of these extreme events attributable to anthropogenic forcing was 100% or close to 100%, according to the Coupled Model Intercomparison Project phase 5 (CMIP5; Taylor, et al., 2012) models. These results reinforce the notion of a potentially high signal-to-noise ratio for anthropogenic warming signals for seasonal and annual anomalies—even at the subcontinental scale in some cases. They further suggest that even if the global warming “hiatus” continues, further extreme (record or near-record) seasonal or annual mean warm anomalies at the regional scale can be anticipated, though the particular regions with such extremes change from year to year (e.g., comparing the present study with our 2012 analysis, Knutson et al. 2013b).
Fig. 8.2. (a,b) Time series of annual averaged surface temperature anomalies (°C) averaged over regions of (a) Australia, left column, and (b) the far western tropical Pacific, right column. The black curves depict the observed (HadCRUT4) anomalies; the dark red (dark blue) curves depict the multi-model ensemble anomalies from the CMIP5 All-Forcing (Natural Forcing-only) runs, with each of the 23 (10) available models weighted equally; the orange curves are individual All-Forcing ensemble members. The green curve in (b) is the 7-yr running mean observed global mean temperature anomaly. The three blue circles labelled “Sensitivity Tests” depict low, medium, and high estimates of the Natural Forcing-only response for 2013 (see Supplemental Material). The All-Forcing simulations for these regions included both anthropogenic and natural forcings from about 1860 to the present, with data from RCP4.5 runs used to extend the time series through 2013 where necessary. Only 10 models had Natural Forcing runs available to us through 2012. All time series shown are adjusted to have zero mean over the period 1881–1920.
(c,d) Trends [°C (100 yr-1)] in the area-averaged annual-mean surface temperature series in (a,b) as a function of starting year, with all trends ending in 2013. The black curves show trends from observations (HadCRUT4), with the black shading depicting the 5th–95th percentile range for the 100-member HadCRUT4 observed ensemble (Morice, et al., 2012), giving one indication of the observational uncertainty in these results. The red curves show the inter-model mean of ensemble mean trends from the 23-member CMIP5 All-Forcing ensemble. The pink region represents the ‘All Forcing’ hypothesis—the 5th–95th percentile range of trends from the All-Forcing runs. The green-shaded region shows the 5th to 95th percentile range of the alternative “Internal Variability Only” hypothesis estimated from the pre-industrial control runs. Purple shading indicates where the pink- and green-shaded regions overlap. The white spaces in the curves denote years where the initial “start year” was missing due to inadequate spatial or temporal coverage. Temporal coverage was assessed as in Fig. 8.1, and the spatial coverage was assessed for each year by requiring at least 33% non-missing annual means for the region.
(e,f) Distribution of annual mean anomalies in the CMIP5 Natural Forcing-only runs (blue) and for the All Forcing runs (red) for 2013. The observed temperature anomalies for 2013 are depicted as dark black vertical lines, with anomalies for another recent similarly extreme year shown by the gray vertical lines.