Assessing the cloud radiative bias at Macquarie Island in the ACCESS-AM2 model

A significant overestimation of downwelling surface shortwave radiation (+9.5±33.5 W m −2) and underestimation of longwave radiation (-2.3±13.5 W m −2) in the ACCESS-AM2 climate model was identified at Macquarie Island, with biases more pronounced during summer (+25.0±48.0 W m −2). The study utilized surface-based observations and lidar simulations to reveal that these radiation biases are linked to underestimated cloud fraction and low-level cloud occurrence, suggesting modeled cloud phase may also play a role. Additionally, the CERES product showed similar biases, indicating both require further development to improve accuracy in simulating Southern Ocean radiation. This long-term, seasonal evaluation highlights critical gaps in climate modeling that impact sea surface temperature and atmospheric circulation.

Pei, Z., Fiddes, S. L., French, W. J. R., Alexander, S. P., Mallet, M. D., Kuma, P., & McDonald, A. (2023). Assessing the cloud radiative bias at Macquarie Island in the ACCESS-AM2 model. Atmospheric Chemistry and Physics, 23(23), 14691–14714. https://doi.org/10.5194/acp-23-14691-2023