Arctic ice ocean heat and freshwater fluxes a

Arctic ice & ocean heat and freshwater fluxes: a new FAMOS coordinated experiment Sheldon Bacon National Oceanography Centre, Southampton, UK With thanks to Yevgeny Aksenov, plus Stephen Fawcett and the NOC NEMO team Ann Keen, Alex West and the UKMO Hadley team FAMOS Meeting, WHOI, 23 -25 October 2013

Motivation • Arctic surface fluxes are poorly constrained in reanalyses – heat & FW fluxes are not well known – heat & FW fluxes are important components of the climate system • Fluxes of mass or energy – only meaningfully defined as transfers between different components of a system – in our case, between atmosphere and ocean (including both liquid and solid parts)

Approach (1) • Arctic ice & ocean heat and FW fluxes: – require closed volume – being calculated now (monthly) for one full annual cycle – based on closed circuit around Arctic (fixed measurement locations plus coastlines) – Bering, Davis, Fram Straits; Barents Sea Opening (see map) – NB calculation from models can include storage terms, eg Qnet = Qsurf + Qstor

Arctic Ocean Boundary Array Tsubouchi et al. , JGR 2012

Approach (2) • Description of method & first application in Tsubouchi, Bacon et al. , JGR 2012 • NB: outcome is real surface fluxes independent of any arbitrary reference values • Note equivalence of ice & ocean boundary calculation with surface flux calculation • Fluxes transferrable metrics (can be calculated in any model with suitable geographical extent) • Model fluxes can be compared with exact observational equivalent (in principle)

Illustrations • NEMO model in 3 runs – 1/4˚ with 2 different surface forcing sets – 1/4˚ & 1/12˚ with the same surface forcing • Showing monthly heat flux: – surface forcing – ocean liquid component – ocean ice (latent) component – ocean ice (sensible) component

Surface heat flux

Ice sensible heat flux (boundary)

Ice latent heat flux (boundary)

Liquid sensible heat flux (boundary)

Liquid sensible heat flux (boundary, Had. GEM 2)

Procedures & Outcomes • A draft “recipe” is available • Assessment of model performance in terms of flux metrics – comparisons (i) between models, and (ii) with observed metrics – decomposition into components (leads to identification of causes of differences)