On the Radiative Quasi-Equilibrium Between Historical Sea-Surface Temperatures and Anthropogenic

SUPPLEMENTAL MATERIAL:

Intensification of seasonal extremes given a 2 C global warming target

Bruce T. Anderson

Fig.S1 Fractional change of ensemble-mean, 10-year average grid-point temperature increases for June-August (JJA) at the time the 10-year average global-mean temperature increases by 1.2 °C relative to base period. Base period starts at 2000-2009 and is increased in increments of 1 year up through 2030-2039. Fractional change given by standard deviation of estimated 10-year average grid-point temperature increases divided by mean 10-year average grid-point temperature increases. Values shown are percentages (%); shading interval is 10%.

Fig.S2 Average fractional change in the 10-year standard deviations of simulated grid-point June-August (JJA) temperature variations at the time the 10-year average global-mean temperature increases by 1.2 °C relative to the base period. Base period starts at 2000-2009 and is increased in increments of 1 year up through 2030-2039. Fractional change given by the average difference in 10-year grid-point temperature standard deviations, divided by the average initial 10-year grid-point temperature standard deviations. Values shown are percentages (%). Minimum value is +/-10%; shading interval is 10%. Global-mean temperature increases, standard deviations, and standard deviation differences at the time the 1.2 °C threshold is reached are calculated separately for each ensemble run. Overall, 120 separate changes are calculated before the averages are taken.

Fig.S3Standard deviation of number of 3-month seasonal-mean temperature values during a given 10-year period that exceed the hottest 3-month mean temperature experienced at the given grid-point during the period 1950-2009. Standard deviation based upon 1000 stochastically-generated decadal time-series given a global-mean temperature increase of 1.2 °C relative to 2000 – see text for details. Exceedence values estimated using the CCSM3 ensemble-mean grid-point temperature trends. Shading interval is seasons/decade. Data in (a) taken from all 3-month mean temperatures within the decade. Data in (b) represent only the number of June-August (JJA) mean temperature values that exceed the hottest JJA mean temperature experienced at the given grid-point during 1950-2009.

Fig.S4(a) Observed 10-year average grid-point June-August (JJA) temperature changes for 2000-2009, relative to 1950-1959. Shading interval is 0.1K/decade. (b) Same as (a) except for grid-point JJA temperature changes derived using the linear trend of JJA temperatures from 1950-2009. (c) Same as (b) except derived using the linear trend of JJA temperatures from 1979-2005, as in IPCC (2007a).