Fragmentation and melting of the seasonal sea ice cover

Lead Research Organisation: University of Reading
Department Name: Meteorology


Recent years have seen a rapid reduction in the summer extent of Arctic sea ice, renewing interest in the physical processes responsible for summer sea ice retreat. Sea ice components of atmosphere-ocean coupled Global Climate Models contain a simplified representation of physical processes at the edge of the sea ice cover. In particular sea ice climate models treat the floe size as a constant whereas observations show there to be a distribution of floe sizes, with significant spatial and temporal variability, especially in the seasonal and marginal ice zones. Smaller floes have an enhanced lateral melt, so providing a feedback when ocean the ocean is warm, potentially accelerating summer sea ice retreat. This PhD project will examine the impact of the observed heterogeneity in floe size on the summer melt and retreat of sea ice.
The seasonal reduction of the sea ice cover is driven by a combination of mechanical floe break up, and melting at the floe edges and upper and lower surfaces. Mechanical floe breakup occurs due to brittle failure in response to heterogeneity in imposed wind stresses, confining stresses, and, especially near the ice edge, by failure in flexure in response to incoming ocean waves and tidal motions, e.g. Squire[2007]. As floes break up, the total perimeter, or edge length, of a given area of sea ice increases. The reduction in floe size affects the response of the ice cover to the incoming wave field and promotes lateral melting. The process of floe breakup is currently absent from climate models: new, physics-based model development will allow its contribution to ice-climate feedbacks, relative to surface melt, to be evaluated for the first time. This PhD project will:
(1)Examine factors controlling the floe size distribution in the seasonal and marginal ice zone.
Observations show that the floe size distribution(FSD)in the marginal ice zone follows a power law distribution [Toyota et al, 2011]. New evidence from ICESAT and forthcoming Marginal Ice Zone campaigns (Office of Naval Research funded Arctic campaign, 2014) will be examined. A model of FSD evolution will be developed including fragmentation and flocculation (freezing together of floes[extending Dumont et al, 2011]. At the sea ice edge in summer, the lack of floe refreezing, and rapid response to storm-driven incoming ocean waves,will allow us to diagnose the FSD from the gross variables included in existing climate models. This FSD model will be ported to the sea ice model CICE.
(2)Examine lateral melt in CICE model.
The CICE model will be used to examine the impact of FSD on lateral melt at floe perimeters. The model will be driven with summer melt season atmospheric forcing and used to examine solar heat distribution into surface melt of the ice (and snow),warming of the mixed layer, and lateral melt. As solar heat is absorbed in the mixed layer and as ice melt occurs, the mixed layer is expected to shallow enabling concentration of heat in the mixed layer. The sensitivity of ice mass balance (concentration and thickness) to the FSD,formulation of lateral melt, and forcing conditions will be explored.
(3)Examine the impact of lateral melting in a coupled climate model
The HadGEM3 climate model will be modified to include the new version of CICE, containing the FSD model, and used to examine Arctic sea ice retreat in the seasonal and marginal ice zones (e.g. Barent's Sea, Fram Strait etc) in historical simulations to identify any enhanced sea ice - climate feedback.
Dumont D,A Kohout, and L Bertino (2011), A wave-based model for the marginal ice zone including a floe breaking parameterization, J.Geophys.Res.,116, doi:10.1029/2010JC006682.
Squire, VA, Of ocean waves and sea-ice revisited,Cold Reg.Sci.&Tech.,49,110-133, 2007.
Toyota T,C Haas, and T Tamura (2011),Size distribution and shape properties of relatively small sea-ice floes in the Antarctic marginal ice zone in later winter,Deep Sea Res. II, 58, 1182-1193.


10 25 50
Description The summer extent of Arctic sea ice has shown a rapidly decreasing trend in recent years, highlighting the need to better understand the mechanisms that drive seasonal retreat of Arctic sea ice. One area of interest is the size of floes (discrete areas of sea ice that make up the overall cover). Climate models assume these adopt a constant size, but observations show that they adopt a wide range of sizes. These observations show that a truncated power law generally produces a good fit to the floe size distribution (FSD). We have produced a research paper (Bateson et al., 2020) to investigate the impact of a power law based FSD on the Los Alamos sea ice model (CICE) coupled to a prognostic mixed layer model. The FSD model allows floe size to respond to several processes including lateral melting, wave induced break-up of floes and freeze-up of the sea ice cover.

This study has a number of key outcomes. Firstly, we have shown that imposing this FSD has a non-uniform impact on the sea ice cover. It enhances the rate of lateral melting within the Marginal Ice Zone (MIZ, defined here as regions with between 15 and 80 % sea ice cover) compared to the pack ice (regions with over 80% sea ice cover). Climate models generally perform poorly in modelling sea ice concentration, so this feature may be an important step towards correcting these biases. We also show that floe size parameters are a much larger source of uncertainty that the other parameters that contribute to lateral melt volume, justifying the focus on floe size as an area of model development. The study has also shown that the attenuation rate of waves under the sea ice cover is an important determinant of how strongly waves influence sea ice evolution, highlighting this as a key focus for further observations and modelling.

The study concludes that the presented model approach is a flexible tool for assessing the importance of a floe size distribution in the evolution of Arctic sea ice. Whilst parallel efforts to develop FSD models without the assumption of a power law do exist, these require the introduction of further parameters to constrain and are more computationally expensive. The power law FSD model hence remains suitable for applications where a simple but realistic floe size distribution model is required.
Exploitation Route My findings highlight key uncertainties within observations, which can help to clarify the important research questions that future data collection expeditions should answer. The output from the model approach developed in Bateson et al. (2020) can in future be compared against the results of ongoing expeditions such as MOSAiC ( alongside the output of other floe size distribution models.

My research into the floe size distribution is relevant to the NERC funded 'Towards a marginal Arctic sea ice cover' project. As I result, I have presented my research findings at meetings for this project and my results have informed some of the model choices made.

As the primary investigator on this award is supervised by researchers from three different institutions (UK Met Office, National Oceanography Centre Southampton and the University of Reading), the findings will help develop future research directions at all three institutions. In particular, NERC centres and the UK Met Office are moving towards the use of a new European sea ice model, SI3. The research produced from this award will hence help inform choices regarding the representation of floe size and wave-sea ice interactions within this new model. The SI3 model will have numerous applications including operational forecasting (used by shipping companies traversing the Arctic ocean) and within climate models.
Sectors Aerospace, Defence and Marine,Environment,Transport