A11I-3105:
Comparison of AOD Data from VIIRS and MODIS over Ocean Using a Data Assimilation System
Monday, 15 December 2014
Edward J Hyer, US Naval Research Laboratory, Marine Meteorology, Monterey, CA, United States, Peng Lynch, Naval Research Lab, Monterey, CA, United States; Computer Science Corporation, Monterey, CA, United States and Min M Oo, University of Wisconsin - Madison, Madison, WI, United States
Abstract:
The NPP VIIRS aerosol products produced by NOAA are being prepared for operational use in the Navy Aerosol Analysis and Prediction System (NAAPS). To utilize the data in this system, it is necessary to examine the data carefully to discern its error characteristics and identify artifacts that can be removed by filtering. After filtering, a prognostic error model must be constructed to estimate the uncertainty in each observation. Then the data must be aggregated, and representativeness error estimated, to match the observations to the scale of feature that can be represented in the aerosol model. Finally, observations from multiple sources must be combined to optimize observation impact. This talk describes these steps for the VIIRS aerosol EDR data over ocean, including construction of the prognostic error model and evaluation of model skill, comparison of VIIRS and MODIS data to identify artifacts and evaluate data quality, and evaluation of NAAPS analysis skill using assimilated MODIS, VIIRS, and a combination of both sensors. Methods for combining observations from different sources, and the implications of these choices for data assimilation, are also discussed.
