Minimum aerosol layer detection sensitivities and their subsequent impacts on aerosol optical thickness retrievals in CALIPSO level 2 data products

摘要

Due to instrument sensitivities and algorithm detection limits, level 2 (L2)Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) 532 nm aerosolextinction profile retrievals are often populated with retrieval fill values(RFVs), which indicate the absence of detectable levels of aerosol withinthe profile. In this study, using 4 years (2007–2008 and 2010–2011) ofCALIOP version 3 L2 aerosol data, the occurrence frequency of daytime CALIOPprofiles containing all RFVs (all-RFV profiles) is studied. In the CALIOPdata products, the aerosol optical thickness (AOT) of any all-RFV profile isreported as being zero, which may introduce a bias in CALIOP-based AOTclimatologies. For this study, we derive revised estimates of AOT forall-RFV profiles using collocated Moderate Resolution ImagingSpectroradiometer (MODIS) Dark Target (DT) and, where available, AErosol RObotic NEtwork (AERONET) data. Globally, all-RFV profiles comprise roughly71 % of all daytime CALIOP L2 aerosol profiles (i.e., including completelyattenuated profiles), accounting for nearly half (45 %) of all daytime cloud-free L2 aerosol profiles. The mean collocated MODIS DT (AERONET) 550 nm AOT is found to be near 0.06 (0.08) for CALIOP all-RFV profiles. Wefurther estimate a global mean aerosol extinction profile, a so-callednoise floor, for CALIOP all-RFV profiles. The global mean CALIOP AOT isthen recomputed by replacing RFV values with the derived noise-floor valuesfor both all-RFV and non-all-RFV profiles. This process yields animprovement in the agreement of CALIOP and MODIS over-ocean AOT.