The performance of the Met Office Unified Model over the central Arctic Ocean

It is believed that the central Arctic Ocean is more sensitive to climate change than lower latitudes due to processes such as the ice-albedo feedback. It is important that both regional and global models can simulate this region accurately to predict future climate with confidence. At present surface flux parameterisation schemes tend to perform poorly over the icecap, leading to significant errors in the surface energy balance and thus in the prediction of future climate. This study focuses on the summer melt and early autumn freeze-up season, in which the dominant feature in the lower atmosphere is the almost permanent layer of low lying stratus cloud.

Measurements of turbulent fluxes and near-surface profiles of mean meteorological parameters, obtained over an ice floe in the central Arctic Ocean during the Arctic Ocean Experiment 2001, are used to evaluate the performance of the UK Met Office Unified Model (UM ) over the region. Firstly, the performance of the surface scheme in the UM, the Joint UK Land Environment Simulator (JULES) is assessed over sea-ice by testing the surface energy budget parameterisation scheme. JULES is run for the observation period, separately from the UM, initialised with field observations at each time step. The resulting surface turbulent fluxes are evaluated against fluxes measured using the eddy covariance method. Secondly, a single column model of the UM is used to assess how the simulated surface turbulent fluxes interact with the atmosphere, with particular attention to the accuracy of the cloud simulation. This study aims to identify areas of poor model performance, understand why models fail in this area and ultimately improve them.