Validation of the on-line aerosol retrieval and error characterization algorithm from the OMI Near-UV observations during the DRAGON-NE Asia 2012 campaign

Friday, 19 December 2014
Ukkyo Jeong1, Changwoo Ahn2, Jhoon Kim1, Pawan K Bhartia3, Omar Torres3, Robert J D Spurr4, Xiong Liu5, Kelly Chance6 and Brent N Holben3, (1)Yonsei University, Seoul, South Korea, (2)Science Systems and Applications, Inc., Lanham, MD, United States, (3)NASA Goddard Space Flight Center, Greenbelt, MD, United States, (4)Rt Solutions Inc, Cambridge, MA, United States, (5)Harvard-Smithsonian Center for Astrophysics, Cambridge, MA, United States, (6)Harvard-Smithsonian, Cambridge, MA, United States
One of the representative advantages of using ultraviolet channel to retrieve aerosol optical property is that the results are less affected by the uncertainty of surface reflectance database. The retrieved aerosol products have relatively uniform quality at both land and ocean except the ice-snow surface. The near UV technique of aerosol remote sensing has additional merit that it has long period database since TOMS (Total Ozone Mapping Spectrometer) including aerosol absorption properties. Thus the retrieved product using the near UV technique using TOMS and OMI (Ozone Monitoring Instrument) measurement is quite appropriate for climatological research. For such purposes, assessment of accuracy of the retrieved product is essential to evaluate the radiative forcing of the aerosols. In this study, the error characterizations of the near UV technique using OMI measurements have been performed with the optimal estimation method during the DRAGON-NE Asia 2012 campaign. In order to avoid the interpolation error, we developed the on-line retrieval scheme based on the traditional near UV method. The retrieval noise and smoothing error of retrieved AOT (Aerosol Optical Thickness) were compared with the biases between 380 nm AOT from AERONET and retrieved 388 nm AOT. They showed positive correlations which infer the possibility of the estimated errors using the optimal estimation method to be used to evaluate the error of retrieved products. Forward model parameter errors were analyzed separately which depends on the quality of the used database, thus can be reduced by improving the database.