Sea Ice Group at the Geophysical Institute logo1
logo2
logo 3

2010 Sea Ice Break-Up Forecast
for Barrow, Alaska
The ice is gone. Ice broke out at NARL on 4 July 2010. There was little evidence of stabilizing pressure ridges, which is unusual compared to previous decades but similar to 2003, 2004, and 2007. This forecast does not apply to these ice conditions. For reference, the forecasted date of break-up was 5 July 2010. See this year's observations and break-up in recent years.

This is a summary
annotated map
Location of Barrow
of the break-up forecast for landfast ice at the Chukchi Sea coast at Barrow, Alaska. In typical years, the timing of break-up is associated with the amount of incoming solar energy (i.e., essentially, how sunny or cloudy it is). In order to predict break-up, we use a combination of data, covering

Latest Forecast

Break-up Forecast

breakup forecast plot
This year's melt progress compared with previous years. Lines end on the day of observed break-up.
Last forecast for break-up: on July 5
Break-up can be forecast no more than 16 days in advance.
In years with significantly grounded pressure ridges, level and rubbled landfast sea ice close to the shore will break up as a result of input of solar energy that leads to structural weakening and disintegration into individual floes. In the figure above, the accumulated incoming shortwave flux is used as an indicator for the time this break-up occurs. The start date of June 5 is chosen as it is typical the beginning of meltpond formation. In 2002, 2003, 2004, and 2007, ice broke out in the absence of grounded pressure ridges, which cannot be predicted with this approach.
The figure above contains the expectation based on the 16-day extended weather forecast (dark red line, end point labeled 2010???), and, after June 5, this year's path already taken (bright red, end point labeled 2010).

Break-up Forecast History

forecasted date
History of forecasted break-up date. Red discs are forecasted break-up dates, black triangles indicate that break-up will happen after the indicated date.
A break-up date was forecasted since June 19 (red dots in figure above). The forecasted date was July 5 ±2 days until July 1, and consistently July 5 since. In the figure above, the green bars indicate the dates with measurements available, the purple bars indicate dates with estimates based on weather observations, and the black bars are dates based on the WRF weather forecast. Black triangles mean that break-up was predicted to take place outside the range of the forecast.

Weather Conditions and Weather Forecast

shortwave radiation data
Downwelling solar shortwave radiation.
For the break-up forecast, the most recent radiation measurements of the ARM site are used. These data are available with approximately 3–4 days delay (black dots).
On days without radiation measurements we estimate the incoming shortwave radiation based on the hourly cloud coverage reported by NOAA at the Barrow Airport and our estimate of the clearsky radiation for the day (purple squares).
Solar radiation forecasts are from the WRF forecast model. Weather forecasts for individual days are not reliable beyond 3–7 days, and this forecast is no exception. Forecast data plotted in the figure above for days in the past are the 24-hour forecasts for the respective days (green triangles).
weather data
Barrow weather and forecast in June.
This is a summary of recent NOAA weather observations at the Barrow airport (PABR), and the output of the latest WRF forecast.

Observations 2010

  • 8 July, 2010: break-out of remaining ice South of Point Barrow
    With the exception of essentially one pressure ridge, landfast ice that remained between NARL and Point Barrow broke out in the early morning hours of July 8, during strong winds from SW. Evidence of this event is recorded in the coastal RADAR and visual satellite imagery. The last pressure ridge dislodged by noon on July 9.
    MODIS image of 8 July, 2010
    MODIS image of noon 8 July, 2010, courtesy MODIS Rapid Response Team.
  • 4 July, 2010: break-out at NARL, ice present between Point Barrow and NARL
    The coastal RADAR showed considerable landfast ice movement at NARL in the early morning hours of the 4th of July. This is the region of interest for the break-up forecast. In addition, the webcam overlooking the coast at downtown Barrow showed a large number of pieces of ice drifting by starting in the early morning hours of 4 July. Satellite images taken on the 5th of July show that the ice off the coast of NARL had disappeared. We are currently investigating when exactly the ice started to move at NARL, and whether there were any grounded pressure ridges involved. Our best estimate is at the moment that either break-up or break-out took place on the 4th of July. This is comparatively early (see below). The sealevel pressure reached a minimum at midnight, indicating potentially high sea level.
    MODIS image of 5 July, 2010
    False-color MODIS image of 5 July, 2010. Sea ice appears blue, clouds are white.
  • 25 June, 2010: break-out downtown Barrow, ice present between Point Barrow and Browerville
    The coastal RADAR and Barrow webcam showed a large chunk of landfast ice break out within one hour at 6 am on 25 June. While this section is in plain view of the webcam, it is South of the region of interest for the forecast. A break-out event, i.e. the removal of pressure ridges and near-shore ice at the same time, indicates that the ice was only marginally grounded, if grounded at all. The sealevel pressure reached a minimum during break-out, indicating potentially high sea level. This event is not related to the break-up forecast as the forecast considers ice North of Barrow, at NARL. This forecast is based on the assumption that the ice off the coast of NARL is grounded and that pressure ridges disappear after the near-shore ice starts to move.
    SAR of 27 June, 2010
    SAR image of 27 June, 2010. Open water is white, sea ice appears as dark gray.

Background

What is Break-Up

aerial photo of landfast ice
Photo of landfast ice between Barrow and Point Barrow. This forecast focuses on the ice North of the site of the Navy Arctic Research Laboratory (NARL).
We define
Chukchi Sea bathymetry at Barrow, Alaska
Coastal bathymetry (depth in m).
break-up as the first detected movement of landfast ice shoreward of grounded ridges within the 20 m-isobath off NARL, approximately 5 miles north of Barrow. Typically, movement is parallel to the coast, confined by grounded pressure ridges at the 20 m-isobath. However, we exclude ice affected by dust from town and a coastal road, i.e. the first approx. 100 m off shore. Typically, we detect movement from coastal RADAR and from Satellite imagery. In previous years, webcam images near NARL were available.

Break-up Process

landfast ice sketch
Sketch of landfast ice at the Chukchi Sea coast at Barrow, Alaska.
Break-up proceeds in two stages: Initially, ice shoreward of grounded pressure ridges begins to move collectively along the coast, followed by a period of sporadic break-out events of individual grounded ridges. After the disappearance of the snow cover in the first half of June, landfast ice closest to the shore weakens structurally due to solar heating of dirty ice and under meltponds. Eventually, regular winds are able to push the ice along shore. With reduced snow cover, more solar energy is absorbed and the oceanic heat flux to the ice increases, eroding grounded pressure ridges from below.
sketch of melting
Impact of solar energy on ice melt.
Some years in the past (2002, 2003, 2004, 2007) have seen insufficient stabilization by grounded pressure ridges, leading in 2004 to the sudden disappearance of the entire coastal ice within 24 hours as early as 18 June. These cases cannot be predicted with the approach presented here. However, if grounded ridges are present, break-up should take place in the first half of July, with some grounded ridges possibly persisting into early August.

Break-up in Recent Years

break-up in previous years
Break-up in previous years.
In the recent past, two modes of break-up were observed: break-up of near-shore ice in the presence of stabilizing, grounded pressure ridges, and break-up without stabilizing pressure ridges (2002, 2003, 2004, 2007, indicated as big gray discs in the figure above). In years with stabilizing pressure ridges, near-shore ice started to move (red triangles) considerably earlier than grounded pressure ridges (blue squares). In these years, the expected date of ice break-up (black dots) coincides with the observed movement of near-shore ice. In 2002, grounded pressure ridges did not appear until several months after the near-shore ice formed, and they formed beyond the reaches of the near-shore ice. Hence, break-up of near-shore ice took place without the stabilizing effect of grounded pressure ridges. (See the 2009 break-up page for details on how break-up progressed during the past 10 years.)

Ice Conditions 2010

As usual,
Coastal Polynia
Coastal polynia
landfast ice off Barrow comprises of patches of deformed ice (rubble) and ice grown in place close to the shore, sheltered by pressure ridges at the 20 m-isobath (approx. 0.5 to 5 km off shore), and attached ice further toward the coastal polynia.
In mid May,
Thickness profile of smooth ice.
Thickness profile of smooth ice
smooth landfast ice had a deeper snow cover than last year (30 cm in 2010 vs. 20 cm in 2009). There was evidence of desalination in the upper 15 cm as early as mid May (approx. 1 to 2 weeks earlier than expected), possibly the result of warm ice due to deep snow and above-average temperature in April. However, due to early freeze-up, the ice is comparatively thick at 1.3 to 1.6 m.
Grounded pressure ridges seem to be in place between Barrow and North towards Point Barrow. However, South of Barrow, landfast ice is essentially smooth and only few if any grounded ridges are present.

Uncertainties

The break-up forecast applies to snow-covered landfast ice with a relatively clean snow surface. For this year, we expect near-shore landfast ice to be held in place by grounded pressure ridges, allowing it to weaken in place starting with the appearance of meltponds, before drifting out past grounded pressure ridges either to the North or to the South. However, the coastal road became snow-free already in mid May, resulting in dust deposited on the snow on sea ice clearly visible on the 100 to 200 m closest to the shore. Hence, we expect the ice adjacent to coastal infrastructure to develop meltponds and subsequently disintegrate notably earlier (about 2 weeks) than the ice this forecast applies to. In particular, large portions of ice overlooked by the Barrow SIZONet webcam are affected by dust.
The start date of integration is the assumed beginning of meltpond formation. Last year, we used June 5 as start date for heat flux integration. However, it appeared that a small improvement could be obtained by using June 7 as start date. Unless meltpond formation starts significantly earlier or later, we expect no adverse affect. The break-up forecast relies on a weather forecast that needs to produce accurate solar energy averaged over one to two weeks. Based on GFS ensemble runs, the uncertainty of cumulative solar shortwave radiation is equivalent to approx. ±2 days in break-up prediction.
The high ice temperatures could move breakup forward by a few days as less energy is required for melt. However, the expected higher run-off from melting snow may be compensating for this as it decreases of the salinity immediately underneath the ice and thereby reduces the convective heat flux to the ice later in June.

Publications

Conference contributions:
  • Chris Petrich, Hajo Eicken, Mette Kaufman, and Don Perovich (2009) Integrating input from local, indigenous ice experts, a coastal ice observatory and measurements of ice decay into a forecast of coastal ice break-up. Poster presented at the Arctic Observing Network (AON) PI Meeting, 30 November–2 December 2009 in Boulder, CO, and at the International Glaciological Society (IGS) International Symposium on Sea Ice in the Physical and Biogeochemical System, 31 May–4 June 2010, in Tromsø, Norway. (pdf)
  • Chris Petrich, Hajo Eicken, Jing Zhang, and Jeremy Krieger (2009) Forecasting coastal ice decay and break-up in northern Alaska. Presentation at the 2009 Fall Meeting of the American Geophysical Union (AGU) in San Francisco, CA, 14–18 December, 2009.

Acknowledgements

The 2010 sea ice break-up forecast is made possible through support of the NOAA Alaska Center for Climate Assessment and Policy (ACCAP). Data are acquired based on work supported by the National Science Foundation (NSF) under Grants No. OPP-0632398, 0712673, and 0856867. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

Figures provided on this page are courtesy Chris Petrich.