Assessment of aerosol optical depth and estimation of lidar ratio from CALIOP measurements

Abstract

The Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) is a space-borne lidar system onboard the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) satellite. CALIOP is a unique instrument that can provide aerosol extinction profile and AOD on a global scale for both day and night. However, the uncertainty in CALIOP AOD is relatively high compared to other passive sensors. In order to assess the uncertainty in CALIOP AOD and investigate potential sources of uncertainty, CALIOP AOD was compared with MODIS-Aqua (hereafter referred to MODIS) AOD over ocean by separating aerosol type defined in CALIOP level 2 algorithm. Such comparisons have been performed for five different aerosol subtypes classified by CALIOP algorithm, namely clean marine, dust, polluted dust, polluted continental, and biomass burning, over the ocean from June 2006 to December 2010. MODIS AOD at 550 nm (0.111 ± 0.079) for the collocated data pairs is about 63% higher than CALIOP AOD at 532nm (0.068 ± 0.073). For clean marine, MODIS AOD (0.110 ± 0.064) is almost twice the CALIOPAOD (0.056 ± 0.038), and the difference between the AOD values has a strong latitude dependence likely related to the surface wind speed over the ocean. The difference in AOD for dust (13%) is observed to be the lowest among the five aerosol types under consideration, but it shows a slight regional variation. The discrepancy of AOD for dust also shows strong dependency on the layer mean of the particulate depolarization ratio. CALIOP AOD is higher than MODIS AOD for both polluted dust and polluted continental by 15% and 29%, respectively, for most of the ocean. One of the possible reasons for the difference is the misclassification of clean marine (or marine + dust) as polluted dust and polluted continental in the CALIOP algorithm. For biomass burning, uncertainty in the layer base altitude is thought to be one of the main reasons for the lower value of CALIOP AOD. Current CALIOP level 2 algorithm uses pre-determined S_aer for classified aerosol types. However, not only using single values of S_aer for each aerosol type but also aerosol classification algorithm could be sources of uncertainties in aerosol extinction retrieval. In this study, to improve currently used S_aer in CALIOP level 2 algorithm, S_aer is determined from CALIOP by using AOD from AERONET and MODIS-Aqua as a constraint without any assumption. Using 4-year measurements (2006-2010) of elastic-backscatter lidar and SKYNET sun/sky radiometer at Seoul National University of Seoul, Korea, mean lidar ratio is estimated to be 61.7±16.5 sr. Lidar ratios are also retrieved using ground-based lidar and AERONET sun/sky radiometer during Distributed Regional Aerosol Gridded Observation Networks (DRAGON) Northeast Asia Campaign 2012. Mean lidar ratios at Seoul and Osaka were retrieved as 65.41±21.42 sr and 65.04±20.62 sr, respectively. Mean lidar ratios from Seoul and Osaka, which are metropolitan cities in East Asia, are almost same and comparable to the lidar ratio for pollution used in CALIOP algorithm. Dust aerosol is easily recognized from depolarization ratio measurements due to its non-spherical shape. Burton et al. [2013] reported that aerosol type classification for dust is the most accurate in CALIOP algorithm. For this reason, lidar ratios are retrieved mainly focused on dust aerosol and compared other studies and CALIOP Level 2 Product. The lidar ratio for dust conditions are estimated to be 51.7±13.7 sr using 4-year measurements of elastic-backscatter lidar and SKYNET sun/sky radiometer at Seoul National University of Seoul, Korea. During DRAGON 2012 NE Asia campaign, lidar ratio for dust event on 27-28 April 2012 in Seoul is retrieved as 48.02±9.38 sr from elastic-backscatter lidar and AERONET sun/sky radiometer. From a synergy of CALIOP and AERONET, lidar ratio for Saharan/Arabian dust is derived to be 47.45±16.52 sr. Lastly, using CALIOP and MODIS measurements together, dust lidar ratio is retrieved as 45.50±15.17 sr. Especially, the mean lidar ratio for Asian dust is estimated much larger as 53.04±18.30 sr. All the lidar ratios for dust aerosol retrieved in this study using AOD constrained method show larger values than currently used lidar ratio for dust in CALIOP algorithm (40 sr), which suggests that dust lidar ratio in CALIOP algorithm is underestimated and needs to be increased.