The role of aerosols in northern tropical Atlantic sea surface temperature anomalies.

摘要

Observations and models demonstrate that northern tropical Atlantic sea surface temperatures are sensitive to regional changes in stratospheric volcanic and tropospheric mineral aerosols. However, it is unknown to what extent the temporal variability of these aerosols shapes observed Atlantic temperature anomalies on all time scales. Simultaneously, 26 years of daily retrievals of aerosol optical thickness from the Advanced Very High Resolution Radiometer suggest strong seasonal to decadal local variability in aerosol optical thickness. However, to-date no study has used the information from these satellites in order to answer the question of the role of aerosols in observed temperature anomalies.;The aim of this thesis is to develop a method for using satellite data in conjunction with simple models to estimate the ocean mixed layer response to aerosol forcing, and then to understand those estimates in terms of recent observed sea surface temperature variability. To do this I create a northern tropical Atlantic climatology of the aerosol direct effect by calculating the change in downwelling surface solar and longwave flux, based on monthly and one-half degree satellite retrievals of aerosol optical depth. I next use this forcing climatology to drive a simple mixed layer model to estimate the integrated ocean temperature response to the aerosol radiative forcing. The output of this mixed layer model is then directly compared to observations of sea surface temperature in order to quantify the role of aerosols in recent ocean temperature variability.;The results from this analysis suggest that low frequency changes in northern tropical Atlantic temperatures are largely the mixed layer response to regional variability in mineral and stratospheric aerosols, with 70% of the upward trend in observed temperatures being directly attributed to aerosol variability, and 67% of the five year smoothed variability around that upward trend also being attributed to local changes in aerosols. Consequently, my results also imply that direct effects from other forcings (e.g., increases in greenhouse gasses or salinity driven changes in the deep ocean circulation) constitute no more than 23% of the variance in the long-term temperature signal, and no more than 30% of the total upward trend.