Land-fast ice project/experiments
Participants:
Bruno Tremblay (leading PI, bruno.tremblay@mcgill.ca)
Andy Mahoney (mahoney@gi.alaska.edu)
Vera Fofonova (Vera.Fofonova@awi.de)
Valeria Selyuzhenok (lera.selyuzhenok@awi.de)
Polona Itkin (Polona.Itkin@awi.de)
Thomas Krumpen (tkrumpen@awi.de)
Einar Olason (einar.ola@gmail.com)
Mathieu Plante (mathieu.plante@mail.mcgill.ca)
Jean-Francois Lemieux (jean-francois.lemieux@ec.gc.ca)
Andrey Proshutinsky (aproshutinsky@whoi.edu)
Introduction
“Land-fast (hereafter, landfast) ice is sea ice which forms and remains fixed along a coast, where it is attached either to the shore, or held between shoals or grounded icebergs. Landfast ice fundamentally modifies the momentum exchange between atmosphere and ocean, as compared to pack ice. It thus affects the heat and freshwater exchange between air and ocean and impacts on the location of ocean upwelling and downwelling zones. Further, the landfast ice edge is essential for numerous Arctic mammals and Inupiat who depend on them for their subsistence.” (from: Arctic Landfast Sea Ice by Christof S. Konig, 2007: http://nsidc.org/data/docs/noaa/g02195fastice/arctic_landfast_sea_ice_thesis.pdf, David Holland – adviser).)
Landfast Ice 1953-1998 data set is available at NSIDC (http://nsidc.org/data/g02195.html) and it contains monthly mean concentrations of landfast ice for the Arctic Ocean from both the Arctic and Antarctic Research Institute (AARI) and the Canadian Ice Services (CIS) sources. Recently published paper “Landfast sea ice extent in the Chukchi and Beaufort Seas: The annual cycle and decadal variability” by Andy Mahoney et al. (2014) about the Chukchi and Beaufort Sea landfast ice is available at http://www.sciencedirect.com/science/article/pii/S0165232X14000585 (in press).
Science Questions / Action Items
1. What is responsible for landfast ice onset, breakup and extent variability in each of the peripheral seas of the Arctic (i.e. those aspects that were not studied by Mahoney et al, Howell et al, Yu et al, etc, Divine et al.)? The Laptev, East Siberian seas and CAA seems to be most understudied regions for landfast ice variability (but note that there is a lot of Russian studies of landfast ice and there are many different methods developed by the Arctic and Antarctic Research Institute (AARI) scientists since the 1940s. This information is available in the AARI publications such as “Trudy AANII” and “Problems of the Arctic and Antarctic” [note: University of Alaska Rasmuson Library has these issues starting from the 1940s: search library catalog for: Arkticheskiĭ i antarkticheskiĭ nauchno-issledovatelʹskiĭ institut)]. The papers can be analyzed by Russian scientists involved in this project and major results about landfast ice of the Siberian Seas can be summarized and reported to the scientific community. These summaries have to include: a) properties and variability of landfast ice for different seas, b) methods of landfast ice predictions. In the CAA the work of Howell et al is similar to that of Yu et al.; i.e. they look at trends in onset/breakup and total area covered by landfast ice region by region, but they have not looked at attribution in details, i.e. the causes of variability.
2. Download/analyze visual images (1m resolution) from the JPL site to identify signs of grounding (ridging) in the East Siberian and Laptev seas.
3. Analyze higher resolution AARI bathymetry data for the Laptev Sea? ESS as well?
4. Develop/validate models of landfast ice
Goals
1. Understand the mechanisms responsible for the presence (e.g. onset, break-off, extent) and intra-inter-annual variability of landfast sea ice in all peripheral seas of the Arctic Ocean – including the Canadian Arctic Archipelago, Beaufort Sea (Alaskan coastline), East Siberian Sea, Laptev Sea, Kara Sea, Lincoln Sea and Greenland Sea (do we want to include those last two items?)
2. When we have a good understanding of the physical mechanisms and external forcing responsible for landfast sea ice presence and for intra-inter-annual variability of landfast sea ice, we would like to develop a sea ice model that will incorporate the correct physics, spatial resolution, parametrization to properly simulate landfast sea ice + variability.
Methods
For goal 1: Analysis of the NSIDC, NIC and AARI landfast sea ice data set; bathymetry (IBCAO, AARI?, ArcticNet for the CAA too sparse?); atmosphere reanalyses (NCEP, NARR, ERA, MERRA). Document variability in landfast sea ice extent from weekly to yearly, to interdecadal time scales. Assess whether bottom topography, air temperatures, winds, runoff or a combination of the three is responsible for the observed inter-intra-annual variability.
For goal 2: Develop a model that includes the necessary ingredients to properly simulate landfast sea ice variability in all regions of the Arctic.
What we know about landfast ice and its variability
The Kara Sea region:
Formation: Landfast ice forms between the coast and offshore islands that provide anchors for ice arches to form between a series of Islands running offshore from Dickson to Severnaya Zemlya.
Variability: Two different distinct regimes exist with stable landfast ice between the coastline and offshore islands (see reference below).
Physical processes: Primarily arch formation between islands. Anchoring of ridges and tensile strength play a secondary role.
Necessary model ingredients: the rheology should include cohesion (uni-axial tensile strength) for the arches to form, and reproduce the observations. The model should also have a very high resolution to resolve the small offshore islands in the Kara Sea for arching to occur.
Reference: Divine et al, Einar Olason PhD Thesis. Contact Einar for details
The Laptev Sea region:
Formation: landfast ice forms each year linking roughly the mouth of the Lena river delta and the New Siberian Islands.
Variability: very small.
Physical processes: Possibly a combination of anchoring of ridges, arch formation between ridges and islands and tensile/shear strength of sea ice. Is the lack of landfast ice variability due to ice anchoring in shallower area and around islands (Stolbovoy and Belkovsky Islands) offshore? Or is it due to shear and tensile ice strength along the coastline and New Siberian Islands? Or a combination of both? Reimnitz et al argue that no ridging or anchoring on the ocean is present in the Laptev Sea based one field campaign. Is this result general, applicable to all years?
Necessary model ingredients: See “Alaskan coastline” section below for ingredients. The model may also need uni- and bi-axial tensile strength. Note that coastal sea ice definitely has non-negligible bi-axial tensile strength.
Reference: Polona worked on this. Vera Fofonova as well? Others reference (see introduction note about literature in Russian from AARI)?
The East Siberian Sea region:
Formation: Landfast sea ice forms between the New Siberian Islands and the Cape just north of Pevek. There is always landfast sea ice between the coast and the 20 m (IBCAO) isobath and there is sometimes (50% of the times) landfast sea ice offshore of the 20 m isobath. The maximum extent of landfast ice is approximately a straight line between the New Siberian Islands and Pevek.
Variability: variability in landfast ice extent in the ESS is correlated with ofshore winds and surface air temperature (r^2 ~= 0.6) AND the landfast ice edge also follows lines of constant depth (isobaths). This suggest that both atmospheric forcing and bottom topography (grounding on the ocean floor) is
responsible for the observed landfast ice extent and variability.
Physical processes: Probably a combination of anchoring of ridges, arch formation, and tensile strength.
Necessary model ingredients: See Alaskan coastline section for ingredient. The model may also need uni- and bi-axial tensile strength.
References: (?)
The Alaskan shallow waters region:
Formation: Landfast ice forms along the Alaskan coastline and extends offshore up to the 20-30-40 m isobaths where ridges anchors the ice on the ocean floor in the stamukhi region.
Variability: very small.
Physical processes: Probably primarily anchoring of pressure ridges. The way the ice follows the isobaths suggests that bottom topography is the key ingredient for stable landfast ice. Air temperature in this region correlates with onset and breakup date of the landfast ice. This has not been looked at in other peripheral seas of the Arctic but is likely to be important there too. Note that offshore winds are very seldom and weak in this region – note also that the keel strength on the ocean is much larger than surface ocean and wind stress integrated over the wind fetch (or landfast ice width) which suggest that breakup imply melting of the ice keels for the fast ice break up to occur (see reference below).
Necessary model ingredients: A sea ice model should have an ice thickness distribution to calculate the probability that the thickest ice present in a grid cell will anchor on the ocean floor + a parameterization for the friction (form and surface drag) between the ice keels and the ocean floor. The model resolution should also be relatively high to resolve the bathymetry.
Reference: Mahoney et al (2007a-b), see also recent paper by Mahoney et al. (2014) above.
The Canadian Archipelago region:
Formation: Landfast ice is present in the CAA pretty much everywhere.
Variability: variability in landfast ice is present in the Lancaster Sound, Amundsen Gulf, Prince Regent Inlet and Foxe Basin. Factor influencing the variability in landfast ice extent is under investigation.
Physical processes: Probably arching and tensile strength since topographical effects do not appear to be important. Ice Stress Buoy data suggest that anisotropy in sea ice thermal stresses, associated with c-axis alignment and surface ocean currents, is a key player affecting the shape of the landfast ice edge in Lancaster Sound. Thermal Tensile stress dominates the stress in this record and dynamic stress are less important.
Necessary model ingredients: the ice rheology should include (uni- and) bi-axial tensile strength and should also include both dynamic and thermal (anisotropic) stresses.
Reference: Howell et al, Hata et al, 2013. Other?
