Dr Gary Brassington¹, Pavel Sakov¹, Andy Hogg², Andrew Kiss², Stewart Allen¹, Helen Beggs¹, Russell Fiedler³, Duan Beckett¹, Jan Lieser¹, Petra Heil⁴, Mikhail Entel¹, Ahmad Khan¹

¹Bureau Of Meteorology, Melbourne, Australia, ²Australian National University, Canberra, Australia, ³CSIRO, Hobart, Australia, ⁴Australian Antarctic Division, Hobart, Australia

Sea-ice forecasting services are not currently available for the Antarctic/Southern Ocean from the Bureau of Meteorology. The absence of these services results in a higher level of risk to Australia’s maritime operations in the polar region.

A sea-ice forecast demonstration project was undertaken to  implement a global ocean sea-ice forecast system as a research service on the national computing infrastructure. The forecast system was based on a global ocean sea-ice model developed under the Consortium for Ocean Sea-Ice Modelling in Australia (COSIMA). The specific model referred to as ACCESS-OM2-01 is a 1/10 degree implementation of the GFDL MOM5 coupled to the CICE5 with 5 thickness categories. The initial conditions were obtained by a hybrid-EnKF using 48 dynamic members and 144 stationary low mode members using ENKF-C. Apart from observations assimilated by the operational global ocean forecasting system OceanMAPS the system also assimilates L2P sea-ice concentration from AMSR2.

Several upgrades were implemented to an earlier prototype system to prepare the forecast system: (1) optimising computational performance; (2) implementing the Bureau’s ACCESS-G3 atmospheric forcing; (3) implement the Bureau’s real-time observational datasets; and (4) upgrade the hybrid-EnKF system. All upgrades were successfully implemented resulting in a significant gain in performance relative to the prototype system.

The system completed a hindcast (Oct 2021 to May 2022) permitting its evaluation and the development of a set of sea-ice products. The performance overall is robust. The forecasting skill is on par with, or slightly beating, the persisted analysis. The limiting factor is likely to be the warm summer SST bias of about 0.2 K in polar regions (also seen in the OFAM3 based OceanMAPS) which in turn resulted in fresh SSS bias of 0.1-0.15 PSU. Investigations are being conducted to enhance the performance and reduce model biases ahead of its implementation for the coming shipping season.

Biography:

Dr Gary Brassington is one of Australia’s leading researchers in ocean forecasting. He is the Bureau of Meteorology’s Senior Principal Research Scientist for Ocean Modelling and leads award winning teams that research, develop and deploy operational ocean forecast systems including the Ocean Model, Analysis and Prediction System (OceanMAPS). OceanMAPS is developed and sustained through a government partnership of the Bureau of Meteorology, CSIRO and Australian Defence Force. Dr Brassington is the national representative for the OceanPredict science team since 2006 and the inaugural chair of the Expert Team on Operational Ocean Forecast Systems.