Comparison of updated SSU temperatures with chemistry climate

Comparison of updated SSU temperatures with chemistry climate model simulations Nathan Gillett Canadian Centre for Climate Modelling and Analysis, Victoria, BC

Introduction • Thompson et al. (2012), The mystery of recent stratospheric temperature trends: The new data call into question our understanding of observed stratospheric temperature trends and our ability to test simulations of the stratospheric response to emissions of greenhouse gases and ozone-depleting substances. • WMO Ozone Assessment 2014, Ch 4: However, in the middle stratosphere, the two data sets exhibit differences in SSU channels 1 and 2 of as large as 0. 5 K decade-1, with the NOAA STAR data set showing a global mean cooling almost twice as large as the Met Office data set (Wang et al. , 2012; Thompson et al. , 2012). The full reasons for these discrepancies remain unknown (e. g. , Nash and Saunders, 2013), but planned updates to the data sets may resolve the discrepancies in part.

Thompson et al. (2012)

Seidel et al. (2011)

Comparison of updated SSU with CCMVal 2 temperatures SSU 3 UKMO NOAA vn 1 dotted vn 2 solid CCMVal 2 SSU 1 (SSU 1+SSU 3)/2 -SSU 2

RMS differences between global mean SSU and CCMVal mean 0, 3 0, 25 0, 2 RMS Difference (K) 0, 15 0, 1 0, 05 0 CCMVal 2 UKMO 1 UKMO 2 NOAA 1 NOAA 2 RMS difference in global mean anomalies relative to CCMVal mean across SSU 1, 2, and 3: mean for individual CCMVal 2 models, and four SSU obs datasets.

Zonal mean trends • WMO Ozone Assessment 2014, Ch 4: However…. the Met Office SSU temperature trends have a different latitudinal structure to the NOAA STAR data (Thompson et al. , 2012), meaning that the BDC trend inferred from each data set would be different. Thus, there are large uncertainties in the changes in the upper branch of the BDC inferred from temperature observations.

Zonal mean trends SSU 3 SSU 2 UKMO NOAA vn 1 dotted vn 2 solid CCMVal 2 SSU 1

Conclusions • Updated NOAA and UKMO global mean SSU temperatures agree better with CCMVal 2 mean than original versions. • NOAA 2 agrees with CCMVal mean as well as individual CCMVal models. • NOAA 2 shows vertically consistent SSU temperature anomalies. • NOAA 2 zonal mean trends agree with CCMVal zonal mean trends much better than original NOAA and UKMO datasets.

Conclusions • The mystery of recent stratospheric temperature trends…. . solved by Cheng-Zhi.