Arctic sea ice melt in summer 2007: Sunlight, water, and ice NSIDC Sept 2007.

Arctic sea ice melt in summer 2007: Sunlight, water, and ice

NSIDCSept 2007

A long list of suspects

Examine the mass balance of the ice

• Warmer air temperature• Longer melt season• Winds• Export of perennial ice• Advection of ocean heat• Clouds• Enhanced solar heating• Ice albedo feedback

Sea ice mass balance

Simple, but powerful observation

• Simple concept• ice growth• surface melt• bottom melt

• Equipment• ablation stakes and thickness gauges• autonomous buoys

• Powerful tool – attributes change

Above

Below

Monitoring the mass balance

Observations in Beaufort and North Pole

http://www.crrel.usace.army.mil/sid/IMB

Summer ice melt

Interannual variability, latitude influence on surface melt

Summer ice melt

2007 at North Pole – typical melt

Summer ice melt

2007 Beaufort – typical surface melt

Summer ice melt

2007 Beaufort – huge increase in bottom melt

Beaufort Sea buoy

End of August – peak bottom melt

20-4040-60

60-80

80-100

150 km

August 27

Beaufort Sea buoy

End of August – peak bottom melt

20-4040-60

60-80

80-100

150 km

August 27

• Warmer air temperature• Longer melt season• Winds• Export of perennial ice• Advection of ocean heat• Clouds• Enhanced solar heating• Ice albedo feedback

Estimate of solar heating

Incident solar, ice concentration, and albedo -> heat input

Focn = Fr (1 – ) (1 – C)

Input:• 25 x 25 km equal area grid • Incident (Fr) from ERA-40, ECMWF• Ice concentration (C) - SSMI• Water albedo () = 0.07

Output:Solar input to the ocean (Focn)

Neglect ice for now

Mean annual input: 1979 – 2005

Ice edge evident, concentric rings from solar input

MJ m-2

3 MJ m-2 = 1 cm m-2 of ice melt

Percent anomaly in 2007

Heat input near buoy 400 to 500 % greater

through 22 Sept 800

700

600

500

400

300

200

100

0

-100

-200

-300

Percent

2007 vs “climatology”

Similar incident shortwave, but much less ice

Solar heat input

Total heat needed for bottom melting

Solar heat input

Twice as much solar heat even without ice transmitted

Conclusions

Was there a trigger?

• North Pole• business as usual

• Beaufort Sea• typical surface melt• huge bottom melt

• Solar incident comparable• Ice concentration much less• Solar input twice melt• Ice albedo feedback

…

Solar heat input

Peak ocean heat flux with ice convergence/divergence

Total solar heat input to ocean

Heat input near buoy from 600 to 800 MJ m-2

through 22 Sept 2007Units = MJ m-2

3 MJ m-2 ~ 1 cm thinning

Percent anomaly in 2007

General increase

Incident shortwave anomaly

Not much of a difference

• Incident shortwave from reanalysis • Pan – Arctic Ocean• 25 x 25 km equal area grid• Average 1979 – 2005

(2007 – average) % anomaly

%

27 year trend of annual ocean solar heat input

Units are percent per year

Increasing solar heat input in 89% of area