Geoscience Reference
In-Depth Information
stationsduringthesummerof1975,weknewthatinthenormalcourseoftheannual
melt cycle, fresh water from the surface would form a relatively shallow seasonal
pycnocline(density gradient).Just belowthis would be a layer of water preserving
approximatelythesametemperatureandsalinitycharacteristicsasthepreviouswin-
ter's mixed layer. By calculating the change in salinity of the water column above
thisremnantlayer,wecouldestimatetheamountoffreshwateraddedoverthemelt
season. It was clear from that first SHEBA sample in early October that a remnant
mixedlayerexistedfromabout30-45mdepth,andthattheseasonalmixedlayerin
October(widespreadbasalfreezinghadnotyetstarted)wassofreshthattheremust
havebeenexcessivemeltduringthesummerof1997.
2
Inotherwords,theevidence
indicated that a strong ice-albedo feedback had kicked in, and as I lay in my bunk
aboardthe
Des Grosiellier
thatnightwhathadbeenupuntilthenatheoreticalexer-
cise nowseemedveryreal.
Theconceptofalbedofeedbackiseasilygrasped—basicallyiceishighlyreflec-
tiveofincomingshortwave(solar)radiationwhileopenseawaterabsorbsnearlyall
of it, so more open water melts ice which creates more open water, and so on—
yet in addition to seasonal changes in albedo of the sea ice itself (Perovich et al.
2002), there are many subtleties in the problem, including storage of sensible heat
in the upper ocean, the rate at which ice melts in contact with above freezing wa-
ter, protection of thin ice by collection of meltwater at the ice undersurface, and
mixing of sensible heat from below the ice-ocean boundarylayer (hereafter abbre-
viatedIOBL).
Fromanearlieranalysisofocean-to-iceheatfluxduringtheAIDJEXexperiment
in thesummerof 1975(MaykutandMcPhee 1995),we hadgainedanappreciation
for the impact of relatively small changesin ice concentrationon the ice mass bal-
ance.AIDJEXcomprisedanarrayoffourdriftstations,threeofwhichmadeatrian-
gle with sides roughly 100km long, surrounding a central station. Over the course
of the summer, we found that integrated ocean heat flux at the easternmost station
(Blue Fox) was about 200MJm
−
2
compared with about 150MJm
−
2
at the other
stations.Thedifference,equivalenttoroughly20cmoficemelt,nearlyallaccumu-
lated during a ten-day period beginning about August 10, nearly two months after
the summer solstice. A mosaic of aerial photographs covering the entire AIDJEX
array taken fortuitously on August 18, revealed that there was considerably more
openwaterinthevicinityofBlueFoxcomparedwiththeotherstations.Apparently
this “open window”(we estimated about25% open water) during a relatively brief
period in late summer was enough to increase ice melt by about a third compared
withthemorecompact(10%openwater)regionsto thewest.
In the decade following the 1997-1998 SHEBA deployment, most of the sci-
entific community and indeed the mass media have become aware of profound
changesoccurringin the Arctic. With regard to sea ice, the most striking symptom
2
Welaterquantified these ideas(McPhee etal.1998) and suggested thatduring the1997summer,
freshening equivalent to as much as 2.5m of ice melt had occurred in the seasonal mixed layer in
theSHEBAvicinity.Wecitedisotopeevidence (courtesy D.Kadko) thatwaterintheintermediate
remnant layer had been in contact with the surface earlier in the summer, and ruled out advection
of fresh runoff as theprimary source because of our distance from the continental margin.
