Evaluation of boundary-layer cloud forecasts over the Southern Ocean in a limited-area numerical weather prediction system using in situ, space-borne and ground-based observations

Yi Huang, Charmaine N Franklin, Steven T Siems, Michael J Manton, Thomas Chubb, Adrian Lock, Simon Alexander, Andrew Klekociuk

Research output: Contribution to journalArticleResearchpeer-review

25 Citations (Scopus)


Near-synchronized in situ, space-borne (A-Train) and ground-based lidar observations are employed to evaluate the boundary-layer clouds (BLCs) over Tasmania and the adjacent Southern Ocean (SO) simulated by the limited-area version of Australian Community Climate and Earth System Simulator (ACCESS-C). Two winter cases featuring BLCs associated with a post-frontal environment and the leading side of a high-pressure ridge are studied. Previous studies showed that these synoptic conditions contribute to the largest reflected short-wave radiation biases simulated over the SO. Results of the simulations suggest that the ACCESS-C model demonstrates an appreciable level of skill in simulating the macrophysical properties of the BLCs over the SO, generally consistent with the in situ and remote-sensing observations. However, some notable challenges remain: the area cloud fraction of the marine BLCs is consistently underpredicted; the fine-scale structure of the marine cumuli is poorly represented in the 4 km grid-length simulations; the capping inversion over the marine boundary layer is generally too high, associated with the marine BLCs being predicted at the wrong altitude and temperature ranges; the liquid water content (LWC) of the BLCs is underestimated; and the model representation of drizzle production can be too efficient. Sensitivity studies are also conducted to test a newly developed autoconversion microphysics scheme and shear-dominated planetary boundary-level (PBL) scheme. These parametrizations show notable improvement in cloud prediction for CASE B (i.e. better area cloud fraction and better average and maximum values of LWC). However, none of these tests is able to improve the simulated marine PBL structure. Overall, the simulated cloud biases are jointly influenced by physical parametrizations, poor representations of large-scale advection, surface fluxes and subsidence. More substantial observations are needed to improve our understanding of the origins and development of these biases and the relative contribution of these errors to the radiation budget over the SO.
Original languageEnglish
Pages (from-to)2259-2276
Number of pages18
JournalQuarterly Journal of the Royal Meteorological Society
Issue number691
Publication statusPublished - 2015


  • Boundary-layer clouds
  • Simulations
  • Southern Ocean

Cite this