6533b7d0fe1ef96bd125b983

RESEARCH PRODUCT

The UKC3 regional coupled environmental prediction system

John SiddornAndrew SaulterJames R. ClarkJoachim FallmannJoachim FallmannJennifer A. GrahamJennifer A. GrahamJuan Manuel Castillo SanchezAlberto Martínez-de La TorreTamzin PalmerAlex ArnoldHuw LewisAdrian LockJohn M. EdwardsLucy BrichenoMike Bush

subject

010504 meteorology & atmospheric sciencesMeteorology010505 oceanographylcsh:QE1-996.5Forecast skillContext (language use)Unified Model01 natural sciencesWind speedAtmospherelcsh:GeologyCoupling (physics)Meteorology and ClimatologySurface waveRange (statistics)Environmental sciencePhysics::Atmospheric and Oceanic Physics0105 earth and related environmental sciences

description

Abstract. This paper describes an updated configuration of the regional coupled research system, termed UKC3, developed and evaluated under the UK Environmental Prediction collaboration. This represents a further step towards a vision of simulating the numerous interactions and feedbacks between different physical and biogeochemical components of the environment across sky, sea and land using more integrated regional coupled prediction systems at km-scale resolution. The UKC3 coupled system incorporates models of the atmosphere (Met Office Unified Model), land surface with river routing (JULES), shelf-sea ocean (NEMO) and ocean surface waves (WAVEWATCH III), coupled together using OASIS3-MCT libraries. The major update introduced since the UKC2 configuration is an explicit representation of wave processes in the ocean and their feedbacks through wave-to-ocean coupling. Ocean model results demonstrate that wave coupling, in particular representing the wave modified surface drag, has a small but positive improvement on the agreement between simulated sea surface temperatures and in situ observations, relative to simulations without wave feedbacks. Other incremental developments to the coupled modelling capability introduced since the UKC2 configuration are also detailed. Coupled regional prediction systems are of interest for applications across a range of timescales, from hours to decades ahead. The first results of simulations run over extended periods, covering four experiments each of order one month in duration are therefore analysed and discussed in the context of further characterising the potential benefits of coupled prediction on forecast skill, and on the stability of such systems over longer time periods. Results across atmosphere, ocean and wave components are shown to be of at least comparable skill to the equivalent uncoupled control simulations, with notable improvements demonstrated in surface temperature and wave state predictions in some near-coastal regions, and in wind speeds over the sea.

https://doi.org/10.5194/gmd-2018-245