Development of CMAQ for East Asia CO2 data assimilation under an EnKF framework: a first result

Abstract

Under an Ensemble Kalman Filter (EnKF) framework, Regional Atmospheric Modeling System and Models-3 Community Multi-scale Air Quality (RAMS–CMAQ) modeling system is developed to be a CO2 data assimilation system EnKF–CMAQ, and the EnKF–CMAQ system is then applied to East Asia for validation with real continuous surface CO2 concentration observations available in the study domain instead of using an observation simulation system experiment. Experiments with an experimental period of January 23 to February 7, 2007 are conducted, and the experimental results of the EnKF–CMAQ system and the RAMS–CMAQ model are compared against continuous surface CO2 observations from assimilation sites and independent reference sites. Distributions of daily mean CO2 concentration increments show that the EnKF–CMAQ system confines the update of daily mean CO2 within areas nearby and downwind of the assimilation sites. Both the CO2 concentration ensemble spreads and background error covariances show flow-dependent patterns. The results indicate the crucial role of wind transport in the CO2 data assimilation, which agrees with the previous studies. The average bias and the average root-mean-square error (RMSE) of daily mean CO2 concentration at the assimilation sites are reduced by 1.00 and 1.83 ppm, respectively, and those at the reference sites are reduced by 0.24 and 0.22 ppm, respectively. The results demonstrate the EnKF–CMAQ system is capable of assimilating the continuous surface CO2 concentration observations to improve the simulation accuracy of the atmospheric CO2 synoptic variation. Since growing CO2 observations over East Asia are being available nowadays, this work is our first step to generate consistent spatial and temporal atmospheric CO2 concentration fields over East Asia, particularly over China, using both in situ and satellite observations.