Stratocumulus sensitivity to aerosols and dynamics

A new technique is presented for quantifying the impacts of aerosols on clouds while controlling for variations in meteorology. Few observational studies have quantified aerosol-cloud impacts at regional or larger scales. The recent work of Kaufman et al (PNAS, 2005) has shown observational evidence for large aerosol effects on clouds. We present work that builds on these results by separating aerosol from meteorological effects on cloud forcing. A new technique is presented, which uses parcel back-trajectories to account for differences in cloud history. Observations are obtained from the MODIS, CERES, and AMSR-E instruments aboard the Aqua and Terra satellites, and are supplemented with ECMWF reanalyses. Geographic and seasonal biases are removed so that climatological variations cannot contribute to false correlations between aerosols and cloud properties. The present work is focused specifically on the stratocumulus cloud regions of the Northeast Pacific and Northeast Atlantic for June through August 2002, the season of maximum cloud cover. Trajectories are grouped using cluster analysis, and evaluated for systematic aerosol-meteorology correlations. Results show statistically significant differences in the meteorology of high and low aerosol cases, indicating that variations in the dynamics are contributing to the observed correlation between aerosols and cloud forcing. Partial derivatives are obtained by compositing data into high, low, and middle terciles. This technique is used to obtain preliminary estimates of cloud sensitivity to aerosol and meteorological properties, which show a reduced aerosol sensitivity, relative to previous estimates. We conclude that meteorological variations must be accounted for in assessing aerosol impacts on cloud forcing.