Data Assimilation for Air Quality—Fusion of Chemical Transport Model and Observations—

Abstract

Data assimilation was initially developed for numerical weather prediction (NWP) to obtain the appropriate initial and lateral conditions for the model simulations (forecasts) through a statistical combination of observation and short-range forecasts (Kalnay, 1999). Recently, progress in the observation and sophistication of the numerical model encourage data assimilation techniques to combine the Chemical Transport Model (CTM) and observations. The aim of this fusion is to be able to label events into the following four major objectives: (1) to obtain a more accurate state for forecasts (initial value problem), (2) to optimize parameters and forcing (emission) in the model (inverse modeling), (3) to generate a uniform, continuous and best-estimated data set (reanalysis), and (4) to investigate the potential impact of the observing system (Observation System Simulation Experiment; OSSE). This paper provides examples of the fusion along with the objectives.