C41B-0196 0800h
Past sea ice Variability Inferred From ice Cores: Development of a Transfer Function for the Canadian Arctic
Recent studies have revealed that Arctic sea-ice cover is rapidly decreasing. Most studies were based on satellite observations spanning the last 30 years, while operational charts and historical observations, of less reliable quality and precision, have increased the record back to a hundred years. The observed changes have been shown to be non-homogenous in time and space, due to complex interactions between surface air temperature and the atmospheric and ocean circulation. It is therefore primordial to better understand these interactions in order to predict the response of the sea ice cover to future greenhouse warming. In this respect, there is a critical need for longer records of sea ice variability which would allow the discrimination of the anthropogenic influence on sea ice conditions from longer-term, natural variability. As a premise to the development of a 1000 year-long sea ice proxy record, this work explores the possible relationships between recent changes (1969-2002) in sea ice conditions in the Canadian Arctic and the glaciochemistry of ice cores from three Canadian sites: Agassiz ice cap (Ellesmere Island), Devon ice cap, (Devon Island) and Penny ice cap (Baffin Island).
C41B-0197 0800h
The 2003-04 Sea Ice Season at Barrow as Seen by Land-Based Radar
The objective was to determine the position of the landfast sea ice edge and its morphology throughout the 2003-04 sea ice season near Barrow using a land-based 10 kW, X-band (3 cm) marine radar mounted on a building near the beach at the Ukpeagvik Inupiat Corporation Naval Artcic Research Laboratory (UIC-NARL). This data would then be used to help understand how landfast ice attaches to coast and what causes it to breakaway. An improved understanding together with real-time data available on the internet will provide more information for the safety of whalers, Native people, and the development of nearshore operations. X-band radar does not discriminate well between level ice and open water, since both areas are characterized by very low back-scatter, but it detects ice floes and rough ice that contain surfaces facing towards the radar. It is very effective at monitoring nearshore sea ice motion in a time series of data. Landfast ice can change very rapidly with break-offs and ice shoves occurring in a matter of hours and nearshore pack ice motion can reverse several times in one day. We were able to monitor these changes by the animation of radar images spaced at five-minute intervals. A very similar land-based radar was installed on the beach approximately 500m further towards Barrow between 1973 and 1979 (Shapiro and Metzner, 1991). In a comparison with this earlier study, the 2003-04 season was noted to be much more dynamic and there was less observed pack ice. In conjunction with field measurements, it was also observed that a stable fast ice edge does not necessarily correlate with a grounded ridge, contrary to the World Meteorological Organization's definition of fast ice.
C41B-0198 0800h
Winds and Ice Motion in Nares Strait and Smith Sound From a Regional Mesoscale Model and Satellite Observations
Flow of sea-ice through the Canadian Archipelago is a potentially important element of Arctic-Subarctic flux balances that is not well quantified observationally and not well represented in current circulation models. As part of an ongoing observational program, concurrent estimates of wind forcing from a multiply-nested mesoscale model and ice motion from AMSR/E satellite data have been made in the Nares Strait and Smith Sound channels west of Greenland. Preliminary results from the meteorological modeling and satellite image analysis provide circumstantial support for the accuracy of the model wind fields and suggestive indications of wind-forcing of ice motion through Nares Strait. Daily forecast simulations during August 2003 through July 2004 are used to estimate the monthly mean wind fields in the Strait and to examine the characteristics and evolution of specific events. Seasonal variations of monthly mean model surface wind fields are consistent with anticipated large-scale circulation patterns around Greenland, with southerly flow in summer and northerly flow in winter. Winds within Nares Strait are topographically controlled and largely channeled through the Strait. Winter winds are stronger and aligned more closely with the channel, while summer winds show outflows from Greenland that extend across the channel to Ellesmere Island. During the winter, the sea-ice moves through this region almost continually and sea ice floes can move at over 60 km per day. An apparently wind-driven flow reversal occurred in early February 2004.
C41B-0199 0800h
Arctic Sea Ice
Of all the recent observed changes in the Arctic environment, the reduction of sea ice cover stands out most prominantly. Several independent analysis have established a trend in Arctic ice extent of -3% per decade from the late 1970s to the late 1990s, with a more pronounced trend in summer. The overall downward trend in ice cover is characterized by strong interannual variability, with a low September ice extent in one year typically followed by recovery the next September. Having two extreme minimum years, such as what was observed in 2002 and 2003 is unusual. 2004 marks the third year in a row of substantially below normal sea ice cover in the Arctic. Early summer 2004 appeared unusual in terms of ice extent, with May a record low for the satellite period (1979-present) and June also exhibiting below normal ice extent. August 2004 extent is below that of 2003 and large reductions in ice cover are observed once again off the coasts of Siberia and Alaska and the Greenland Sea. Neither the 2002 or 2003 anomaly appeared to be strongly linked to the positive phase of the Arctic Oscillation (AO) during the preceding winter. Similarly, the AO was negative during winter 2003/2004. In the previous AO framework of Rigor et al (2002), a positive winter AO implied preconditioning of the ice cover to extensive summer decay. In this hypothesis, the AO does not explain all aspects of the recent decline in Arctic ice cover, such as the extreme minima of 2002, 2003 and 2004. New analysis by Rigor and Wallace (2004) suggest that the very positive AO state from 1989-1995 can explain the recent sea ice minima in terms of changes in the Arctic surface wind field associated with the previous high AO state. However, it is also reasonable to expect that a general decrease in ice thickness accompanying warming would manifest itself as greater sensitivity of the ice pack to wind forcings and albedo feedbacks. The decrease in multiyear ice and attendant changes in ice thickness distribution could in turn precondition the Arctic ice cover to further reductions in the subsequent summer(s) regardless if the summer temperatures were anomalously warm. The NSIDC Sea Ice Index (http://nsidc.org/data/seaice_index/) can be used to view trends and anomalies from 1979 on.
C41B-0200 0800h
Tracking the Arctic's Shrinking Ice Cover: Another Extreme Minimum in 2004.
Of all the recent observed changes in the Arctic environment, the reduction of sea ice cover stands out most prominantly. Several independent analysis have established a trend in Arctic ice extent of -3% per decade from the late 1970s to the late 1990s, with a more pronounced trend in summer. The overall downward trend is characterized by strong interannual variability, with a low September ice extent in one year typically followed by recovery the next September. Having two extreme minimum years, such as what was observed in 2002 and 2003 is unusual. 2004 marks the third year in a row of substantially below normal sea ice cover in the Arctic. Early summer 2004 appeared unusual in terms of ice extent, with May a record low for the satellite period (1979-present) and June also exhibiting below normal ice extent. August 2004 extent is below that of 2003 and large reductions in ice cover are observed once again off the coasts of Siberia and Alaska and the Greenland Sea. Neither the 2002 or 2003 anomaly appeared to be strongly linked to the positive phase of the Arctic Oscillation (AO) during the preceding winter. Similarly, the AO was negative during winter 2003/2004. In the previous AO framework of Rigor et al (2002), a positive winter AO implied preconditioning of the ice cover to extensive summer decay. In this hypothesis, the AO does not explain all aspects of the recent decline in Arctic ice cover, such as the extreme minima of 2002, 2003 and 2004. New analysis by Rigor and Wallace (2004) suggest that the very positive AO state from 1989-1995 can explain the recent sea ice minima in terms of reductions in the overall age of ice driven by the previous high AO state. However, it is also reasonable to expect that a general decrease in ice thickness accompanying warming would manifest itself as greater sensitivity of the ice pack to wind forcings and albedo feedbacks. The decrease in multiyear ice and attendant changes in ice thickness distribution could in turn precondition the Arctic ice cover to further reductions in the subsequent summer(s) regardless if the summer temperatures were anomalously warm. The NSIDC Sea Ice Index (http://nsidc.org/data/seaice_index/) can be used to view trends and anomalies from 1979 on.
C41B-0201 0800h
Decoupling of ice production and ice extent in seasonally ice covered marginal seas
While the seasonally ice covered marginal seas are ice free in the summer, there may be substantial production and transport of ice during the winter. The predominant ice types in these seas are frazil/grease ice, pancake ice and thin sheet ice. Relative production of the different ice types has a dramatic effect on the amount of brine production and the local energy exchange rate between the ocean and the atmosphere. We have developed a model which utilizes daily observations of SMMR and SSM/I microwave radiometers to track the volume and areal ice concentrations of each ice type; allowing us to evaluate the contribution of each ice type to the ice mass, salt, and fresh water redistribution. From this framework, we have calculated the spatial distribution of the annual net salt and fresh water flux to the Bering Sea over a 15 year period. The results indicate there may be a de-coupling between sea ice coverage and sea ice production for the Bering Sea, with the ice production being much less variable than the sea ice coverage. These results are interesting because, to first order, ice production is not responsive to variability in atmospheric forcing from year to year. In this paper, we will discuss the relative importance of different causal mechanisms, and examine linkages and feedbacks between the sea ice, ocean and atmosphere.
C41B-0202 0800h
A New Look at the Northern Hemisphere Sea Ice Concentration
It is widely recognized that the Arctic sea ice cover has been shrinking over the past 25 years. Our knowledge of hemispheric sea ice concentration and extent comes almost entirely from satellite passive microwave (PM) data collected since 1978. In this study, we use a new data set of Northern Hemisphere sea ice concentration, derived from weekly operational ice charts spanning more than three decades (1972-2003), to re-examine the regional variability and trends in sea ice area and extent. The ice charts from the U.S. National Ice Center have been converted to EASE-Grid format. Source data for the charts include visible and infrared satellite imagery, active radar imagery, PM data, aerial reconnaissance, ship and shore observations, buoys, model output, information from foreign ice services, and climatology. The PM data are used only when all other forms of data are not available. Thus we have a unique gridded data set that is largely independent of the popular PM products that are widely used by the sea ice community. We divided the Arctic and sub-Arctic seas into regions and created monthly time series of sea ice area and extent for each region. We also obtained the monthly NASA Team sea ice concentration products. We re-gridded these to the same EASE-Grid format as the charts, and computed time series of sea ice area and extent for the same regions. We present comparisons of the regional differences and trends seen in the two data sets. We explain the differences based on the source data used in the charts, and the emissivity of sea ice as detected by the PM instruments. Future work with the ice chart data set includes analysis of multiyear, first-year, and new ice concentrations in order to understand the recent record-low summer ice minima; duration of the ice season, as an indicator of climate change; and analysis of the modes of variability of the ice edge, in order to develop a predictive capability for sea ice extent.
C41B-0203 0800h
The role of melt onset timing in the recent extreme Arctic summer ice extent minima
The recent reduction in summer ice cover in recent years has been one of the starkest indications of change in the Arctic. The September 2002 minimum extent was the lowest since the beginning of the satellite record in 1972. September 2003 was the second lowest on record and nearly as low as 2002. 2004 marks an unprecedented third year in a row of anomalously low ice extents. This marks a striking change from previous extreme low years, which have typically been followed by a rebound to higher ice extents the next year. Previous studies have indicated that a positive phase in the previous winter's Arctic Oscillation (AO) can lead to a larger export of thicker multiyear ice out of the Arctic through the Fram Strait. This results in a thinner average ice cover that is preconditioned for more extensive melt the following summer. However, in 2002-2004 the winter AO has been negative. Another possible factor leading to the September minimum extent is the length of the melt season. Warmer spring and summer temperatures that can result in earlier melt have been noted in many regions of anomalously low summer extents. Here, we investigate links between the timing of the melt onset, based on passive microwave data, with the following September extent as a possible factor in the extreme minima of the past three years.
C41B-0204 0800h
Arctic and Antarctic Ice Pack Changes during the Past Decade from a High Resolution Global Coupled Sea Ice-Ocean Model
Changes over the past decade in the global ice pack are analyzed using a coupled ice-ocean model and observational data sets. The model consists of the latest versions of the Los Alamos Parallel Ocean Program (POP) and sea ice model (CICE) and is configured on a moderately high-resolution global grid (0.4$\deg$ and 40 vertical levels). A model simulation forced with high frequency daily NCEP/NCAR atmospheric fields was integrated for 23 years (1979-2002). Following a decade-long ice spin-up, the model's ability to reproduce observed ice extent, ice thickness and ice drift distributions is evaluated by statistical comparisons using satellite, upward looking sonar and ice drift buoy data. In particular, the realism of the ice mean state and variability on time scales from daily to interannual are examined. To better understand ocean-ice interaction processes, coupled model results are compared to stand alone integrations of the ice and ocean models. Mean ice states are examined during the positive/negative phases of the North Atlantic Oscillation and Arctic Oscillation in the last decade of the coupled simulation. Particularly ice export from the Fram and Bering Straits during these phases will be considered.
C41B-0205 0800h
A Low Frequency Radar for Direct Measurement of Sea Ice Thickness: Implications of Ice Surface Roughness
Sea ice thickness is thought to be a primary indicator of global greenhouse warming, yet it has proven to be one of the most difficult variables to measure particularly on meaningful synoptic and climatic scales, including from satellites. An instrument concept study and associated field experiment system development are underway to measure the sea ice thickness directly with VHF radar. This system precludes the use of very wide bandwidth as in radar sounder, but instead uses new instrument technology as a combined spatial- and frequency-domain interferometer. Thickness is derived from phase obtained from a combination of slightly different narrow-band frequencies and incidence angles. The use of VHF is required to overcome the lossy nature of sea ice and penetrate to many meters of thickness to detect the sea-ice ocean interface. The approach relies heavily on the larger contrast in dielectric constant between the sea-ice/ocean than the sea-ice/snow as well as the sea ice bottom roughness, that results in a stronger backscatter return from the sea-ice ocean interface than sea-ice snow interface. The detailed formulations of the theoretical basis of this concept are presented in accompanying paper. Key information on surface roughness characteristics at relatively fine-scales, particularly of ice undersides for which comparatively little is known, and sea ice medium composition (for example, brine inclusions and air bubbles) are needed to develop the scientific basis of these technologies, as well as to develop the measurement strategy of a spaceborne sensor. In this presentation, we present an overview of the radar concept, fine-scale bottom roughness measurements from different upward looking sonar data sets, and an approach for a meaningful spatial and temporal measurement strategy for a future spaceborne instrument.