Thirteenth ARM Science Team Meeting Proceedings, Broomfield, Colorado, March 31-April 4, 2003 Inter-Comparison and Synergy Between the Two Long-Term Global Aerosol Products Derived from AVHRR and TOMS M.-J. Jeong and Z. Li Department of Meteorology University of Maryland College Park, Maryland D. A. Chu and S.-C. Tsay National Aeronautics and Space Administration Goddard Flight Center Greenbelt, Maryland Introduction Eighteen years of satellite-based monthly aerosol products have been derived from the advanced very high resolution radiometer (AVHRR) and total ozone mapping experiment spectrometer (TOMS) sensors. The two products differ in many regards rendering a great potential for developing an integrated product for climate studies. Presented here are some preliminary results of inter-comparison and synergy analyses. Global Aerosol Climatology Satellite-based long-term aerosol climatology shows considerable spatial and temporal variability, but distinct regional and seasonal distribution features are clearly seen. They are useful for climate studies such as the earth’s radiation budget. In general, satellite-based aerosol retrievals suffer from much larger uncertainties for instantaneous values than long-term means due to compensating errors from various sources (Mishchenko et al. 1999). These long-term dataset can be further improved by rescaling through comparisons with estimates from current and future satellites designed specifically for aerosol retrieval by reducing some biases, which were not removed by averaging. Several regions were selected in light of their distinct characteristics for further analysis. Figure 1 shows time series of monthly mean values of AVHRR aerosol optical thickness (AOT), Angstrom exponent, and TOMS AOT over the oceans covering the latitudinal bands of 30S~30N, 30S~EQ, and EQ~30N. As shown, seasonal variations are clear for all the variables, and decadal variations also exist. Decadal variations of AOTs are related with two major volcanic eruptions (Mt. El Chichon in 1982; Mt. Pinatubo in 1991). No visually discernable trend was found. It is hard to differentiate any subtle trend from uncertainties due to sensor calibration, cloud contamination, etc. It has been reported that any gentle linear trends of the global mean derived must be construed in the context of potential long- term drift in ISCCP calibration (Brest et al. 1997; Rossow and Schiffer 1999). The very low-frequency 1