Abstract
As a critical quality control step toward producing a stratospheric data assimilation system for volcanic aerosols, we conducted a comparison between Stratosphere Aerosol and Gas Experiment (SAGE) II aerosol extinction profiles and aerosol backscatter measured by five lidars, both in the tropics and midlatitudes, for the two‐year period following the 1991 Mt. Pinatubo eruption. The period we studied is the most challenging for the SAGE II retrieval because the aerosol cloud caused so much extinction of the solar signal that in the tropics few retrievals were possible in the core of the cloud. We compared extinction at two wavelengths at the same time that we tested two sets of conversions coefficients. We used both Thomason and Jäger's extinction‐to‐backscatter conversion coefficients for converting lidar backscatter profiles at 0.532 μm or 0.694 μm wavelengths to the SAGE II extinction wavelengths of 0.525 μm and 1.020 μm or the nearby ones of 0.532 μm and 1.064 μm respectively. The lidars were located at Mauna Loa, Hawaii (19.5°N, 155.6°W), Camagüey, Cuba (21.4°N, 77.9°W), Hefei, China (31.9°N, 117.2°W), Hampton Virginia (37.1°N, 76.3°W), and Haute Provence, France (43.9°N, 5.7°W). For the six months following the eruption the aerosol cloud was much more heterogeneous than later. Using two alternative approaches, we evaluated the aerosol extinction variability of the tropical core of the Pinatubo stratospheric aerosol cloud at the timescale of 1–2 days, and found it was quite large. Aerosol variability played the major role in producing the observed differences between SAGE II and the lidars. There was in general a good agreement between SAGE II extinction measurements and lidar derived extinction, and we conclude that all five lidar sets we compared can be used in a future data assimilation of stratospheric aerosols. This is the most comprehensive comparison yet of lidar data with satellite data for the Pinatubo period.