SBUV Version 4 PMC Data and Trend Analysis: Did the Trend Change in the mid-1990’s?

We have developed an updated Version 4 (V4) polar mesospheric cloud (PMC) data set from 35 years of SBUV measurements. The detection threshold has been revised to better represent the expected variation of PMC brightness with scattering angle. We now recommend the use of 273 nm directional albedo values to reduce uncertainties associated with instrument noise. We present our analysis of the V4 PMC data set in terms of ice water content (IWC), using a parameterization of regression results from a microphysical model to determine IWC from observed albedo. While we derive a local time adjustment as a function of IWC in order to use all SBUV measurements, we also find that our long-term trend analysis is insensitive to the nature of this function.

In contrast to our earlier published work assuming a linear long-term trend of the SBUV time series, our new multiple linear regression results indicate that the anti-correlated response to solar activity and the increasing trend were both stronger in the first part of our data record (through ~1996) compared to the latter part. The use of a two-part linear trend for this analysis provides a long-term forcing similar to observed variations in upper stratospheric ozone and lower stratospheric temperature, as suggested earlier by Lübken et al [2011]. The fraction of total variance in the IWC time series explained by the multiple regression model is significantly smaller for all Southern Hemisphere latitudes compared to the corresponding Northern Hemisphere latitude. This suggests that interannual variability (not captured by our simple model) plays a larger role in the Southern Hemisphere. Possible explanations for the mid-1990’s ‘break’ in the long-term record are offered.

Lübken, F.‐J., and U. Berger (2011), Latitudinal and interhemispheric variation of stratospheric effects on mesospheric ice layer trends, J. Geophys. Res., 116, D00P03, doi:10.1029/2010JD015258.