Debate on Stratospheric Temperature Trends from SSU Measurements

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Thursday, 6 February 2014: 1:45 PM
Room C102 (The Georgia World Congress Center )
Cheng-Zhi Zou, NOAA/NESDIS, College Park, MD

Warming of the troposphere and cooling of the stratosphere are two central features of global warming resulting from anthropogenic forcing. Although cooling of the lower stratosphere has been well documented with satellite observations from the Microwave Sounding Unit (MSU), changes in the middle and upper stratosphere were not well understood due to a lack of well documented satellite datasets as well as other in situ observations. NOAA/STAR has recently released a well documented stratospheric temperature time series derived from the Stratospheric Sounding Unit (SSU) observations onboard historical NOAA polar orbiting satellites. The stratospheric temperature trends derived from this dataset has been compared with an earlier version developed by UK Met Office (UKMO) and with climate model simulations from the Coupled Model Intercomparison Project Phase 5 (CMIP5) and Chemistry-Climate Model Validation phase 2 (CCMVal2) projects. Extensive differences were found in absolute magnitudes and temporal patterns, both within and between models and observations. The results have been published in Nature by a group of scientists, which further sparked debate on the causes of these differences.

NOAA STAR is recalibrating the SSU observations for improved understanding of the stratospheric temperature trends. Firstly, a recalibrated level-1c radiance dataset is being developed for improved data assimilation in climate reanalyses. Secondly, trend decomposition for various SSU bias adjustment effects is being analyzed to understand stratospheric temperature trend differences between different versions. This presentation will introduce initial results on SSU level-1c recalibration and NOAA's Version 2.0 SSU dataset. Reasons for trend differences between NOAA and UKMO SSU data will be discussed. It was found that cell pressure correction differences may explain a large part of the trend differences in channels 1 and 2. Reasons for differences in spatial trend patterns will also be discussed in this presentation.