C44A-01
Recent State of Arctic Sea Ice
We present the recent state of Arctic sea ice including observations from 2008 in a context of a multi-decadal perspective. A new record has been set in the reduction of Arctic perennial sea ice extent this winter. As of 1 March 2008, the extent of perennial sea ice was reduced by one million km2 compared to that at the same time last year as observed by the NASA SeaWinds scatterometer on the QuikSCAT satellite (QSCAT). This decrease of perennial ice continues the precipitous declining trend observed in this decade. Furthermore, the perennial sea ice pattern change was deduced by buoy-based estimates with 50 years of data from drifting buoys and measurement camps to track sea ice movement around the Arctic Ocean. The combination of the satellite and surface data records confirms that the reduction of winter perennial ice extent broke the record in 2008 compared to data over the last half century. In the winter, the loss of perennial ice extent was driven by winds that compressed the ice and transported it out of the Fram Strait and Nares Strait to warmer ocean waters at lower latitudes, where the ice melted very effectively. Another historical fact is that the boundary of perennial sea ice already crossed the North Pole (NP) in February 2008, leaving the area around the NP occupied by seasonal sea ice. This is the first time, not only from the satellite data record but also in the history of sea ice charting at the National Ice Center since the 1970's, that observations indicate the seasonal ice migration into the NP area so early in winter. In the Bering Sea by 12 March 2008, the sea ice edge reached to an extent that coincided with the continental shelf break, indicating bathymetric effects on the distribution of water masses along the Aleutian North Slope, Bering Slope, Anadyr, and Kamchatka Currents that governed the pattern of sea ice formation in this region. Moreover, QSCAT observations showed that, in the 2008 winter, seasonal ice occupied the Northern Sea Route, and most of two routes of the Northwest Passage, north and south of Victoria Island, which facilitated ice retreat and the opening of waterways this summer. Most importantly, the shift from a perennial to a seasonal ice covered Arctic Ocean significantly decreases the overall surface albedo resulting in enhanced solar heat absorption in spring and summer, which further decreases the Arctic ice pack through the ice albedo feedback mechanism. In early September 2008, a major melt event occurred over a large region extending from the Beaufort Sea across the Kara Sea toward the Laptev Sea, with active melt areas encroaching in the NP vicinity. This melt event was caused by an advection of warm air from the south, which melted and pushed sea ice away at the same time. At that time, the total extent of Arctic sea ice was about 0.5 million km2 (size of Spain) larger than that at the same time last year.
C44A-02 INVITED
Albedo changes of the Arctic sea ice cover
The summer extent of the Arctic sea ice cover has decreased in recent decades and there have been alterations in the timing and duration of the summer melt season. This has resulted in changes in the evolution of albedo of the Arctic sea ice cover, and consequently in the partitioning of solar energy. These changes are examined on a pan-Arctic scale on a 25 x 25 km Equal Area Scalable Earth Grid for the years 1979 – 2007. Daily values of incident solar irradiance are obtained from ERA-40 reanalysis products and ice concentrations are determined from passive microwave satellite data. The albedo of the ice is modeled by a five-phase process that includes dry snow, melting snow, melt pond formation, melt pond evolution, and freezeup. The timing of these phases is governed by the onset dates of summer melt and fall freezeup, which are determined from satellite observations. Results indicate a general trend of increasing solar heat input to the Arctic ice-ocean system due to reductions in ice concentration and longer melt seasons. This trend may accelerate the loss of sea ice through the ice-albedo feedback. The evolution of albedo, and hence the total solar heating of the ocean, is more sensitive to the date of melt onset than the date of fall freezeup.
C44A-03 INVITED
Mechanisms of Upper Ocean Warming in the Arctic and the Effect on Sea Ice Melt
The surface layers of the Arctic Ocean are experiencing an unprecedented summertime warming, with temperatures up to about 5 degrees Celsius above the historical mean. Using a numerical model, we examine three distinct heat budgets related to this warming. First, we consider the causes of ocean surface warming over the upper 60 m, separating the forcing into net surface heating, oceanic lateral heat flux convergence, and a residual that represents vertical exchange at 60 m depth. We find that lateral heat flux convergence is mainly important near northwestern Alaska, where Pacific/Chukchi water pumps heat into the upper layers. In recent years with less sea ice cover, this effect is somewhat larger and more extensive. Nonetheless, surface heating is a far more extensive effect and accounted for about 3/4 of the net warming of the open water area in the summer of 2007. Second, we consider the relative roles of the ocean and the atmosphere in melting sea ice from the bottom/sides versus the top surfaces, and compare our results with recent data taken by ice mass balance buoys. Third, we partition the net atmospheric heating of the ocean during summer into the part used to melt ice and the part used to warm the upper ocean. The latter term is growing in recent years as sea ice retreats.
C44A-04
Vertical structure of recent Arctic warming from observed data and reanalysis products
Spatial distribution of recent arctic air temperature changes is complex however knowledge of its vertical structure and geographical distribution may shed light on physical mechanisms driving this warming. Sparseness of high-latitude observational network has implications for the quality of available datasets. In this study, the spatiotemporal patterns of the recent air temperature trends are evaluated using three reanalysis datasets and compared against rawinsonde data. Our analysis focused on 1979-2008 reveals large differences between the datasets due to changes in the data assimilation procedure used in reanalysis. This casts certain doubts on the robustness of recently documented Arctic trends. Both NCEP products as well as NARR show consistent results regarding recent Arctic temperature trends. ERA-40 tends to be better in terms of root mean square error although it has certain jumps in the behavior of tropospheric temperature. A change of sign in the winter temperature trend (from negative to positive) since 1980th-90th is documented in the upper troposphere/lower stratosphere with a maximum over the Canadian Arctic. This result seems to be robust and is reproduced by NCEP reanalysis products and NARR. This change is associated with weakening of the stratospheric polar vertex and shift of its maximum toward Siberian coast.
C44A-05
Arctic Ocean, Spring 2008: What's left after the minimum ice extent of 2007?
On the heels of a record minimum ice extent in September 2007, unique changes were observed in the central Arctic Ocean and Beaufort Sea in spring 2008 as part of collaboration amongst the North Pole Environmental Observatory (NPEO), Beaufort Gyre Exploration Project (BGEP), Ice Tethered Profiler (ITP) and Switchyard programs. NPEO includes an automated drifting station that samples air-ice-ocean conditions while drifting from the Pole toward Fram Strait plus a deep ocean mooring near the Pole. NPEO and the Switchyard program also include repeated airborne hydrographic surveys that track changes along sections radiating from the Pole and north from Ellesmere Island. For IPY, the airborne survey was expanded to include the Beaufort Sea sections of BGEP, with the spring samples giving a seasonal counterpoint to the BGEP summer ship-based surveys. Most of the airborne stations were done with a ski-equipped aircraft landed on the sea ice to take water samples and CTD, oxygen and nutrient profiles. The most remote stations were made using new Air eXpendable CTDs (AXCTD) dropped in leads from an over-flying aircraft. The most remarkable difference from past years was the dominance of first-year ice in most areas, typically a little less than 2 m thick with relatively little snow cover. The Web cameras at the NPEO automated drifting station suggest that the lack of snow may have reduced ice melt. When the snow melted, water appeared to drain through the ice quickly, leaving the ice surface with few melt ponds and a relatively high-albedo appearance. This process, if it was widespread, may have helped more first year ice to survive the summer of 2008 than expected. In 2008, as in 2007, the NPEO ice mass balance buoy showed mostly top melt in spite of colder than normal air temperatures. Similar buoys in the Beaufort Sea showed mostly bottom melt. Near the Pole, upper ocean salinities and Atlantic Water temperatures increased in the early 1990s as the pattern of water circulation and ice drift shifted cyclonically. From 2000 to 2005, NPEO saw conditions revert nearly to the pre-1990 state. In 2008, North Pole upper ocean salinity and Atlantic water temperature anomalies increased again in a pattern more reminiscent of the 1990s. Throughout the Beaufort Sea, the upper ocean was exceptionally fresh, with the surface salinity observed to be about 25 at the east margin of our survey, near the middle of the eastern half Canadian Basin. Water chemistry indicates that above 190 meters depth, runoff dominated ice melt as a source of the freshwater. The melt contribution was negative in many samples, consistent with net annual ice production in spite of the spectacular melt of 2007. The surface layer salinity was also reduced 1-2 psu relative to climatology (and 2007) along the 90° W Switchyard section between the Pole and Ellesmere Is., suggesting a movement of freshened water eastward along the Archipelago toward Fram Strait.
C44A-06
Why global climate model predictions of Arctic warming are too conservative?
General circulation models (GCMs) that participated in the Intergovernmental Panel for Climate Change Fourth Assessment Report (IPCC-AR4) on average predict some 50% or more reduction of summer sea ice cover in the Arctic Ocean by the end of this century. Unfortunately the majority of those models have significant limitations in their representation of past and present variability in the Arctic. The inability of climate models to reproduce the recent warming and ice melt in the Arctic Ocean diminishes their accuracy of future climate predictions. Some of these limitations include: northward oceanic heat fluxes, distribution and variability of sea ice in the Arctic Ocean, and its export into the North Atlantic. The general tendency in those models is to transport warm Atlantic Water via the Barents Sea, with Fram Strait experiencing mostly outflow to the south. In reality, the West Spitsbergen Current flowing along the eastern part of Fram Strait delivers the majority of heat into the Arctic Ocean, while most of the heat entering the Barents Sea is released to the atmosphere before entering the central Arctic. More importantly, the heat advected by Summer Pacific Water through Bering Strait is distributed by local currents and eddies over the Chukchi Shelf and into the Beaufort Sea and is readily available for melting sea ice in the western Arctic, where most of the ice retreat has taken place. We argue that high resolution is required to realistically model the flow of Pacific Water and the associated heat advection to address GCM limitations and their ability to realistically simulate sea ice variability in the western Arctic. This means that the magnitude of oceanic heat input to the Arctic Ocean and its impact on the sea ice might be significantly under-represented both in space and time in global GCMs, which may help explain their conservative predictions of warming and ice melt there.
C44A-07 INVITED
The Hydrologic Cycle Response to Rapid Arctic Vegetation Change
Over the last fifty years, the Northern Hemisphere high latitude land areas have warmed at rates well in excess of what can be explained by the atmospheric rise in greenhouse gases alone. Changes in the albedo of the ocean and land, whether from the loss of Arctic Ocean sea ice, changes in land cover, or changes in winter precipitation patterns account for much of the amplified warming. Although the loss of sea ice is directly related to greenhouse gas warming and low-level winds, changes in the discharge of freshwater from Arctic river basins are also responsible. While changes in river discharge can be related to precipitation, snow and ice melt, and human modification of the landscape, natural vegetation changes due to warming may also be altering the land surface hydrologic cycle and contributing to changes in the flux of freshwater to the Arctic Ocean. Satellite imagery has shown that the Arctic is becoming greener, which not only affects the surface and lower-tropospheric energy budget, but also modifies the hydrologic cycle through altering the partitioning of transpiration and plant-soil evaporation. This leads to changes in precipitation recycling and runoff, which can ultimately affect the discharge of freshwater. To illustrate this mechanism, results of a land cover change and precipitation-recycling analysis using North American Regional Reanalysis data will be presented for the Mackenzie Basin in North America. Additionally, results from a dynamic global vegetation model will be presented to evaluate the potential consequences of continued extreme warming and land cover changes to the discharge of freshwater to the Arctic Ocean.
C44A-08
A Reconstructed 1784-2007 Time Series of Greenland Melt Extent
Total melt on the Greenland ice sheet has been rising over the past several decades. The melt extent observed in 2007 was the greatest on record according to several satellite-derived indices of Greenland melt. Observed melt extent across the Greenland ice sheet has been shown to be strongly related to summer station temperatures from locations along Greenland's coastal periphery, as well as to variations in the circulation of the atmosphere across the North Atlantic. We exploit these relationships with historical temperature and circulation observations to develop a 224-yr reconstructed history of annual Greenland melt extent from the late 18th century to 2007. This reconstruction allows us to put recent melt, particularly 2007, into a historical perspective and compare current melt to the well-known warm period in the early half of the 20th century. Our reconstruction indicates that the melt observed since the late 1990s is likely among the highest extents to have occurred since the late 18th century, although recent values are not statistically different from those common during the period 1923-1961, a time when summer temperatures along the southern coast of Greenland were similarly high as those experienced in recent years. The reconstruction indicates that if the current trend toward increasing melt extent continues, total melt across the Greenland ice sheet will exceed historic values of the past two and a quarter centuries.