Impact of melt ponds on Arctic sea ice simulations from 1990 to 2007

Flocco, Daniela; Schröder, David; Feltham, D. L.; Hunke, Elizabeth

doi:10.1029/2012jc008195

Cited by 106 publications

(141 citation statements)

References 56 publications

Supporting

Mentioning

136

Contrasting

Order By: Relevance

“…With a typical sea ice area of about 5.5 million km 2 in September, a refreezing pond coverage of 20% [Flocco et al, 2012] and using our model estimate of a 22 cm reduction in basal growth when a trapped pond is present, we calculate that ignoring trapped ponds in the growth calculation gives a corresponding overestimate of ice growth of 265 km 3 in about 2 months. This is approximately 25% of the average sea ice volume change in the same period estimated by PIOMAS [Zhang and Rothrock, 2003].…”

Section: Resultsmentioning

confidence: 99%

“…A first attempt at modeling the refreezing of melt ponds was described in Taylor and Feltham [2004], although this ignored the role of pond salinity. A simpler version of this approach, essentially treating the ice lid growth as a classic Stefan phase change problem, was incorporated into a climate sea ice model by Flocco et al [2010Flocco et al [ , 2012. Based on these works, Hunke et al [2010] developed another melt pond routine based on similar principles.…”

Section: Introductionmentioning

confidence: 99%

“…Previous modeling studies of melt ponds have focused on their formation and summertime evolution [Taylor and Feltham, 2004;Flocco et al, 2010Flocco et al, , 2012L€ uthje et al, 2006;Skylingstad, 2009]. In autumn, the ponds refreeze at their upper surface to form an ice lid, which insulates them from the atmosphere and traps pond water between the sea ice and the ice lid.…”

Section: Introductionmentioning

confidence: 99%

“…Based on these works, Hunke et al [2010] developed another melt pond routine based on similar principles. The model presented by Flocco et al [2010Flocco et al [ , 2012 distinguishes between active (no ice lid) and nonactive ponds that are covered by an ice lid.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

The refreezing of melt ponds on Arctic sea ice

et al. 2015

Self Cite

View full text Add to dashboard Cite

The presence of melt ponds on the surface of Arctic sea ice significantly reduces its albedo, inducing a positive feedback leading to sea ice thinning. While the role of melt ponds in enhancing the summer melt of sea ice is well known, their impact on suppressing winter freezing of sea ice has, hitherto, received less attention. Melt ponds freeze by forming an ice lid at the upper surface, which insulates them from the atmosphere and traps pond water between the underlying sea ice and the ice lid. The pond water is a store of latent heat, which is released during refreezing. Until a pond freezes completely, there can be minimal ice growth at the base of the underlying sea ice. In this work, we present a model of the refreezing of a melt pond that includes the heat and salt balances in the ice lid, trapped pond, and underlying sea ice. The model uses a two-stream radiation model to account for radiative scattering at phase boundaries. Simulations and related sensitivity studies suggest that trapped pond water may survive for over a month. We focus on the role that pond salinity has on delaying the refreezing process and retarding basal sea ice growth. We estimate that for a typical sea ice pond coverage in autumn, excluding the impact of trapped ponds in models overestimates ice growth by up to 265 million km 3 , an overestimate of 26%.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

The refreezing of melt ponds on Arctic sea ice

et al. 2015

Self Cite

View full text Add to dashboard Cite

show abstract

“…Our reference simulation includes a prognostic melt pond model (Flocco et al, 2010(Flocco et al, & 2012 and the elastic anisotropic plastic rheology (Wilchinski and Feltham, 2006;Tsamados et al, 2013;Heorton et al, 2018). Otherwise, default CICE settings are chosen: 7 vertical ice layers, 1 snow layer, linear remapping ITD approximation (Libscomp and Hunke, 2004), Bitz and Libscomp (1999) thermodynamics, Maykut and Untersteiner (1971) conductivity, Rothrock (1975) ridging scheme with a Cf value of 12 (empirical parameter that accounts for frictional energy dissipation) and the Delta-Eddington radiation scheme 35 (Briegleb and Light, 2007 Comparing the CICE simulation with CS2 reveals that CICE default underestimates the mean monthly sea ice thickness by about 0.8 m (see Fig.…”

Section: Reference Simulation (Cice-default) 30mentioning

confidence: 99%

New insight from CryoSat-2 sea ice thickness for sea ice modelling

Feltham¹,

Schröder²,

Tsamados³

2018

Preprint

View full text Add to dashboard Cite

Abstract. Estimates of Arctic sea ice thickness are available from the CryoSat-2 (CS2) radar altimetry mission during ice growth seasons since 2010. We derive the sub-grid scale ice thickness distribution (ITD) with respect to 5 ice thickness categories used in a sea ice component (CICE) of climate simulations. This allows us to initialize the ITD in stand-alone 10 simulations with CICE and to verify the simulated cycle of ice thickness. We find that a default CICE simulation strongly underestimates ice thickness, despite reproducing the inter-annual variability of summer sea ice extent. We can identify the underestimation of winter ice growth as being responsible and show that increasing the ice conductive flux for lower temperatures (bubbly brine scheme) and accounting for the loss of drifting snow results in the simulated sea ice growth being more realistic. Sensitivity studies provide insight into the impact of initial and atmospheric conditions and, thus, on the role of 15 positive and negative feedback processes. During summer, atmospheric conditions are responsible for 50% of September sea ice thickness variability through the positive sea ice and melt pond albedo feedback. However, atmospheric winter conditions have little impact on winter ice growth due to the dominating negative conductive feedback process: the thinner the ice and snow in autumn, the stronger the ice growth in winter. We conclude that the fate of Arctic summer sea ice is largely controlled by atmospheric conditions during the melting season rather than by winter temperature. Our optimal model configuration does 20 not only improve the simulated sea ice thickness, but also summer sea ice concentration, melt pond fraction, and length of the melt season. It is the first time CS2 sea ice thickness data have been applied successfully to improve sea ice model physics.

show abstract

Surface and melt pond evolution on landfast first-year sea ice in the Canadian Arctic Archipelago

Landy

Ehn

Shields

et al. 2014

J. Geophys. Res. Oceans

110

View full text Add to dashboard Cite

The evolution of landfast sea ice melt pond coverage, surface topography, and mass balance was studied in the Canadian Arctic during May-June 2011 and 2012, using a terrestrial laser scanner, snow and sea ice sampling, and surface meteorological characterization. Initial melt pond formation was not limited to low-lying areas, rather ponds formed at almost all premelt elevations. The subsequent evolution of melt pond coverage varied considerably between the 2 years owing to four principle, temporally variable factors. First, the range in premelt topographic relief was 0.5 m greater in 2011 (rougher surface) than in 2012 (smoother surface), such that a seasonal maximum pond coverage of 60% and maximum hydraulic head of 204 mm were reached in 2011, versus 78% and 138 mm in 2012. A change in the meltwater balance (production minus drainage) caused the ponds to spread or recede over an area that was almost 90% larger in 2012 than in 2011. Second, modification of the premelt topography was observed during mid-June, due to preferential melting under certain drainage channels. Some of the lowest-lying premelt areas were subsequently elevated above these deepening channels and unexpectedly became drained later in the season. Third, ice interior temperatures remained 1-2 C colder later into June in 2012 than in 2011, even though the ice was 0.35 m thinner at melt onset, thereby delaying permeability increases in the ice that would allow vertical meltwater drainage to the ocean. Finally, surface melt was estimated to account for approximately 62% of the net radiative flux to the sea ice cover during the melt season.

show abstract

Impact of melt ponds on Arctic sea ice simulations from 1990 to 2007

Cited by 106 publications

References 56 publications

The refreezing of melt ponds on Arctic sea ice

The refreezing of melt ponds on Arctic sea ice

New insight from CryoSat-2 sea ice thickness for sea ice modelling

Surface and melt pond evolution on landfast first-year sea ice in the Canadian Arctic Archipelago

Contact Info

Product

Resources

About