Aerosol in the online coupled global chemical weather model CMA-GFS/CUACE CW1.0 -comparison with observations and ICAP ensembles

This study developed the Global chemical weather model CMA-GFS/CUACE CW1.0 (GCUACE), by fully integrating the CMA Unified Atmospheric Environment (CUACE) module with the CMA Global Forecast System. The model features a horizontal resolution of 0.25° and extends vertically to 0.1 hPa. Its daily aerosol forecasts from June 2016 to May 2017 were evaluated against 550 nm aerosol optical depth (AOD) data from the global AERONET ground-based network, MODIS satellite retrievals, and the International Cooperation for Aerosol Prediction (ICAP) multi-model ensemble product (ICAP-MME). Evaluation showed that GCUACE's multi-site annual mean total AOD (0.299) was closer to AERONET observations (0.30), though with a marginally larger standard deviation than ICAP-MME. In quantitative skill rankings among multiple international models, GCUACE achieved mid-range performance for both total and coarse-mode AOD forecasts at 6-hour and 72-hour. Spatially, GCUACE successfully reproduced major global aerosol patterns, matching satellite and ensemble data in highlighting high-value regions like East Asia, South Asia, and dust-affected Africa and the Middle East. The model demonstrated significant regional variation in performance: it excelled at forecasting East Asian dust storms but underestimated biomass burning aerosol over the African Sahel and the Amazon, as well as long-range transported Saharan dust to the Caribbean. Over oceans, it exhibited a north-high, south-low bias. A key strength was its robust temporal stability, with forecast correlation declining more slowly over time compared to ICAP-MME, indicating reliable performance. Furthermore, its skill was consistently higher for coarse-mode aerosols than for total AOD, underscoring a specific capability in dust prediction.