Stephen E. L. Howell

The use of operational ice charts for evaluating passive microwave ice concentration data

Abstract More than 1380 regional Canadian weekly sea‐ice charts for four Canadian regions and 839 hemispheric U.S. weekly sea‐ice charts from 1979 to 1996 are compared with passive microwave sea‐ice concentration estimates using the National Aeronautics and Space Administration (NASA) Team algorithm. Compared with the Canadian regional ice charts, the NASA Team algorithm underestimates the total ice‐covered area by 20.4% to 33.5% during ice melt in the summer and by 7.6% to 43.5% during ice growth in the late fall. The wide range in performance occurs because some Canadian regions such as the western Canadian Arctic are only partly in the marginal sea‐ice zone while the Hudson Bay and the East Coast regions are entirely in the marginal sea‐ice zone, being completely ice free in summer. Compared with the U.S. National Ice Center hemispheric chart series, the average underestimation is 18.6% in summer. During other times of the year the differences are less than 7.8%. The magnitudes of the underestimation during ice melt and ice growth periods with respect to the Canadian regional charts are higher than found in other studies in the literature. The difference in performance of the NASA Team algorithm when compared with the Canadian regional ice charts and hemispheric charts is attributed to: 1) passive microwave data are used in preparing the hemispheric charts so comparison with the hemispheric charts is not independent; 2) the Canadian regional charts are in the marginal ice zone where sea‐ice melt or growth conditions occur over a large percentage of the region and last for several weeks to a month; and 3) the tie points used to calibrate the NASA Team algorithm are tuned to perform best over the hemisphere rather than over individual regions. If the Canadian regional ice charts can be accepted as correct, then these results suggest exercising caution in using passive microwave ice concentration data in the marginal ice zone where ice melt and ice growth conditions are a major component of the sea‐ice regime.

Changing sea ice conditions and marine transportation activity in Canadian Arctic waters between 1990 and 2012

Declining sea ice area in the Canadian Arctic has gained significant attention with respect to the prospect of increased shipping activities. To investigate relationships between recent declines in sea ice area with Arctic maritime activity, trend and correlation analysis was performed on sea ice area data for total, first-year ice (FYI), and multi-year ice (MYI), and on a comprehensive shipping dataset of observed vessel transits through the Vessel Traffic Reporting Arctic Canada Traffic Zone (NORDREG zone) from 1990 to 2012. Links to surface air temperature (SAT) and the satellite derived melt season length were also investigated. Between 1990 and 2012, statistically significant increases in vessel traffic were observed within the NORDREG zone on monthly and annual time-scales coincident with declines in sea ice area (FYI, MYI, and total ice) during the shipping season and on a monthly basis. Similarly, the NORDREG zone is experiencing increased shoulder season shipping activity, alongside an increasing melt season length and warming surface air temperatures (SAT). Despite these trends, only weak correlations between the variables were identified, although a step increase in shipping activity is apparent following the former summer sea ice extent minimum in 2007. Other non-environmental factors have also likely contributed to the observed increase in Arctic shipping activity within the Canadian Arctic, such as tourism demand, community re-supply needs, and resource exploration trends.

Surface-Based Polarimetric C-Band Scatterometer for Field Measurements of Sea Ice

A portable surface-based polarimetric C-band scatterometer for field deployment over sea ice is presented. The scatterometer system, its calibration, signal processing, and near-field correction are described. The near-field correction is shown to be effective for both linear polarized and polarimetric backscatter. Field methods for the scatterometer are described. Sample linear polarized and polarimetric backscatter results are presented for snow-covered first-year sea ice (FYI), multiyear hummock ice, and rough melt pond water on FYI. The magnitude of backscatter signature variability due to system effects is presented, providing the necessary basis for quantitative analysis of field data.

Regional variability of a projected sea ice‐free Arctic during the summer months

Climate projections of sea ice retreat under anthropogenic climate change at the regional scale and in summer months other than September have largely not been evaluated. Information at this level of detail is vital for future planning of safe Arctic marine activities. Here the timing of when Arctic waters will be reliably ice free across Arctic regions from June to October is presented. It is shown that during this century regions along the Northern Sea Route and Arctic Bridge will be more reliably ice free than regions along the Northwest Passage and the Transpolar Sea Route, which will retain substantial sea ice cover past midcentury. Moreover, ice-free conditions in the Arctic will likely be confined to September for several decades to come in many regions. Projections using a selection of models that accounts for agreement of models in each region and calendar month with observations yield similar conclusions.

Ice thickness in the Northwest Passage

Recently, the feasibility of commercial shipping in the ice-prone Northwest Passage (NWP) has attracted a lot of attention. However, very little ice thickness information actually exists. We present results of the first ever airborne electromagnetic ice thickness surveys over the NWP carried out in April and May 2011 and 2015 over first-year and multiyear ice. These show modal thicknesses between 1.8 and 2.0 m in all regions. Mean thicknesses over 3 m and thick, deformed ice were observed over some multiyear ice regimes shown to originate from the Arctic Ocean. Thick ice features more than 100 m wide and thicker than 4 m occurred frequently. Results indicate that even in todays climate, ice conditions must still be considered severe. These results have important implications for the prediction of ice breakup and summer ice conditions, and the assessment of sea ice hazards during the summer shipping season.

Implications of fractured Arctic perennial ice cover on thermodynamic and dynamic sea ice processes

Decline of the Arctic summer minimum sea ice extent is characterized by large expanses of open water in the Siberian, Laptev, Chukchi, and Beaufort Seas, and introduces large fetch distances in the Arctic Ocean. Long waves can propagate deep into the pack ice, thereby causing flexural swell and failure of the sea ice. This process shifts the floe size diameter distribution smaller, increases floe surface area, and thereby affects sea ice dynamic and thermodynamic processes. The results of Radarsat-2 imagery analysis show that a flexural fracture event which occurred in the Beaufort Sea region on 6 September 2009 affected ∼40,000 km2. Open water fractional area in the area affected initially decreased from 3.7% to 2.7%, but later increased to ∼20% following wind-forced divergence of the ice pack. Energy available for lateral melting was assessed by estimating the change in energy entrainment from longwave and shortwave radiation in the mixed-layer of the ocean following flexural fracture. 11.54 MJ m−2 of additional energy for lateral melting of ice floes was identified in affected areas. The impact of this process in future Arctic sea ice melt seasons was assessed using estimations of earlier occurrences of fracture during the melt season, and is discussed in context with ocean heat fluxes, atmospheric mixing of the ocean mixed layer, and declining sea ice cover. We conclude that this process is an important positive feedback to Arctic sea ice loss, and timing of initiation is critical in how it affects sea ice thermodynamic and dynamic processes.

Sensitivity of AMSR-E Brightness Temperatures to the Seasonal Evolution of Lake Ice Thickness

The sensitivity of brightness temperature (TB) at 6.9, 10.7, and 18.7 GHz from Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) observations is investigated over five winter seasons (2002-2007) on Great Bear Lake and Great Slave Lake, Northwest Territories, Canada. The TB measurements are compared to ice thicknesses obtained with a previously validated thermodynamic lake ice model. Lake ice thickness is found to explain much of the increase of TB at 10.7 and 18.7 GHz. TB acquired at 18.7 GHz (V-pol) and 10.7 GHz (H-pol) shows the strongest relation with simulated lake ice thickness over the period of study (R2 > 0.90). A comparison of the seasonal evolution of TB for a cold winter (2003-2004) and a warm winter (2005-2006) reveals that the relationship between TB and ice growth is stronger in the cold winter (2003-2004). Overall, this letter shows the high sensitivity of TB to ice growth and, thus, the potential of AMSR-E mid-frequency channels to estimate ice thickness on large northern lakes.

Fusing AMSR-E and QuikSCAT Imagery for Improved Sea Ice Recognition

The benefits of augmenting Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E) image data with Quick Scatterometer (QuikSCAT) image data for supervised sea ice classification in the Western Arctic region are investigated. Experiments compared the performance of a maximum likelihood classifier when used with the AMSR-E-only data set against using the combined data. The preferred number of bands to use for classification was examined, as well as whether principal component analysis (PCA) can be used to reduce the dimensionality of the data. The reliability of training data over time was also investigated. Adding QuikSCAT often improves classifier accuracy in a statistically significant manner and never decreases it significantly when a sufficient number of bands are used. Combining these data sets is beneficial for sea ice mapping. Using all available bands is recommended, data fusion with PCA does not offer any benefit for these data, and training data from a specific date remains reliable within 30 days.

Local-level responses to sea ice change and cruise tourism in Arctic Canada's Northwest Passage

Abstract This article examines the interactions between sea ice change and patterns of cruise ship tourism through the Northwest Passage of Arctic Canada and how local communities are responding to this change. During the period 2006–2010, the Passage has emerged as the most popular expedition cruise area in the Canadian Arctic with an increase in planned cruises by 70%. This dramatic increase in cruise traffic has been, in part, facilitated by improved access as a result of decreases in sea ice. Since 1968, total sea ice area in northern route of the Northwest Passage has decreased by 11% per decade and total sea ice area in the southern of the Northwest Passage has decreased by 16% per decade. Integrating research from both social and geophysical science, this article presents an analysis of changing cruise tourism patterns through the Northwest Passage and analyses resident responses from Passage communities including Ulukhaktok, Gjoa Haven, and Pond Inlet. Discussion is focused on issues associated with infrastructure, security, protection of the marine environment, human safety, and search and rescue. This research is important to help prepare communities, policy makers, as well as the cruise sector itself, to be responsive to change in these remote locations.

The influence of declining sea ice on shipping activity in the Canadian Arctic

Significant attention has focused on the potential for increased shipping activity driven by recent observed declines in Arctic sea ice cover. In this study, we describe the first coupled spatial analysis between shipping activity and sea ice using observations in the Canadian Arctic over the 1990-2015 period. Shipping activity is measured using known ship locations enhanced with a least-cost path algorithm to generate ship tracks, and quantified by computing total distance travelled in km. Statistically significant increases in shipping activity are observed in the Hudson Strait (150-500 km travelled year-1), the Beaufort Sea (40-450 km travelled year-1), Baffin Bay (50-350 km travelled year-1), and regions in the southern route of the Northwest Passage (50-250 km travelled year-1). Increases in shipping activity are significantly correlated with reductions in sea ice concentration (Kendalls tau up to -0.6) in regions of the Beaufort Sea, Western Parry Channel, Western Baffin Bay, and Foxe Basin. Changes in multi-year ice dominant regions in the Canadian Arctic were found to be more influential on changes to shipping activity compared to seasonal sea ice regions.