• Atkinson, D. E., 2005: Observed storminess patterns and trends in the circum-Arctic coastal regime. Geo-Mar. Lett., 25, 98109.

  • Barber, D. G., and J. M. Hanesiak, 2004: Meteorological forcing of sea ice concentrations in the southern Beaufort Sea over the period 1979 to 2000. J. Geophys. Res., 109, C06014, doi:10.1029/2003JC002027.

    • Search Google Scholar
    • Export Citation
  • Bell, G. D., and L. F. Bosart, 1988: Appalachian cold-air damming. Mon. Wea. Rev., 116, 137161.

  • Cavalieri, D. J., C. L. Parkinson, and K. Y. Vinnikov, 2003: 30-year satellite record reveals contrasting Arctic and Antarctic decadal sea ice variability. Geophys. Res. Lett., 30, 1970, doi:10.1029/2003GL018031.

    • Search Google Scholar
    • Export Citation
  • Christensen, J., and Coauthors, 2007: Regional climate projections. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 847–940.

    • Search Google Scholar
    • Export Citation
  • Colucci, S. J., and J. C. Davenport, 1987: Rapid surface anticyclogenesis: Synoptic climatology and attendant large-scale circulation changes. Mon. Wea. Rev., 115, 822836.

    • Search Google Scholar
    • Export Citation
  • Comiso, J. C., 2003: Warming trends in the Arctic from clear sky satellite observations. J. Climate, 16, 34983510.

  • Danard, M., A. Munro, and T. Murty, 2003: Storm surge hazard in Canada. Nat. Hazards, 28, 407431.

  • Drobot, S. D., and J. A. Maslanik, 2003: Interannual variability in summer Beaufort Sea ice conditions: Relationship to winter and summer surface and atmospheric variability. J. Geophys. Res., 108, 3233, doi:10.1029/2002JC001537.

    • Search Google Scholar
    • Export Citation
  • Hudak, D., and J. Young, 2002: Storm climatology of the southern Beaufort Sea. Atmos.–Ocean, 40, 145158.

  • Jones, B. M., C. D. Arp, M. T. Jorgenson, K. M. Hinkel, J. A. Schmutz, and P. L. Flint, 2009: Increase in the rate and uniformity of coastline erosion in Arctic Alaska. Geophys. Res. Lett., 36, L03503, doi:10.1029/2008GL036205.

    • Search Google Scholar
    • Export Citation
  • Jorgenson, M. T., Y. L. Shur, and E. R. Pullman, 2006: Abrupt increase in permafrost degradation in Arctic Alaska. Geophys. Res. Lett., 33, L02503, doi:10.1029/2005GL024960.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471.

  • Keegan, T., 1958: Arctic synoptic activity in winter. J. Meteor., 15, 513521.

  • Lynch, A., J. A. Curry, R. D. Brunner, and J. A. Maslanik, 2004: Toward an integrated assessment of the impacts of extreme wind events on Barrow, Alaska. Bull. Amer. Meteor. Soc., 85, 209221.

    • Search Google Scholar
    • Export Citation
  • Manson, G., and S. Solomon, 2007: Past and future forcing of Beaufort Sea coastal change. Atmos.–Ocean, 45, 107122.

  • McCabe, G. J., M. P. Clark, and M. C. Serreze, 2001: Trends in Northern Hemisphere surface cyclone frequency and intensity. J. Climate, 14, 27632768.

    • Search Google Scholar
    • Export Citation
  • Mesinger, F., and Coauthors, 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343360.

  • Screen, J. A., and I. Simmonds, 2010a: The central role of diminishing sea ice in recent Arctic temperature amplification. Nature, 464, 13341337.

    • Search Google Scholar
    • Export Citation
  • Screen, J. A., and I. Simmonds, 2010b: Increasing fall–winter energy loss from the Arctic Ocean and its role in Arctic temperature amplification. Geophys. Res. Lett., 37, L16707, doi:10.1029/2010GL044136.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., 1995: Climatological aspects of cyclone development and decay in the Arctic. Atmos.–Ocean, 33, 123.

  • Serreze, M. C., and A. P. Barrett, 2008: The summer cyclone maximum over the central Arctic Ocean. J. Climate, 21, 10481065.

  • Serreze, M. C., J. E. Box, R. G. Barry, and J. E. Walsh, 1993: Characteristics of Arctic synoptic activity, 1952–1989. Meteor. Atmos. Phys., 51, 147164.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., and Coauthors, 2003: A record minimum Arctic sea ice extent and area in 2002. Geophys. Res. Lett., 30, 1110, doi:10.1029/2002GL016406.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., M. M. Holland, and J. Stroeve, 2007: Perspectives on the Arctic’s shrinking sea-ice cover. Science, 315, 15331536.

  • Simmonds, I., and K. Keay, 2009: Extraordinary September Arctic sea ice reductions and their relationships with storm behavior over 1979–2008. Geophys. Res. Lett., 36, L19715, doi:10.1029/2009GL039810.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., and E.-P. Lim, 2009: Biases in the calculation of Southern Hemisphere mean baroclinic eddy growth rate. Geophys. Res. Lett., 36, L01707, doi:10.1029/2008GL036320.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., C. Burke, and K. Keay, 2008: Arctic climate change as manifest in cyclone behavior. J. Climate, 21, 57775796.

  • Solomon, S. M., 2005: Spatial and temporal variability of shoreline change in the Beaufort–Mackenzie region, Northwest Territories, Canada. Geo-Mar. Lett., 25, 127137.

    • Search Google Scholar
    • Export Citation
  • Stroeve, J. C., M. M. Holland, W. Meier, T. Scambos, and M. Serreze, 2007: Arctic sea ice decline: Faster than forecast. Geophys. Res. Lett., 34, L09501, doi:10.1029/2007GL029703.

    • Search Google Scholar
    • Export Citation
  • Whiteman, C. D., and J. C. Doran, 1993: The relationship between overlying synoptic-scale flows and winds within a valley. J. Appl. Meteor., 32, 16691682.

    • Search Google Scholar
    • Export Citation
  • Zhang, X., J. E. Walsh, J. Zhang, U. Bhatt, and M. Ikeda, 2004: Climatology and interannual variability of Arctic cyclone activity, 1948–2002. J. Climate, 17, 23002317.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Map of the Beaufort coast showing the location of the Tuktoyaktuk station and the surface elevations of the local topography.

  • View in gallery

    Box plots of the hourly wind speed (m s−1) distribution by month. The top and bottom of the box indicate the 75th and 25th percentile wind speeds, respectively. The line inside the box indicates the median wind speed. The length of the dashed lines (whiskers) is 1.5 times the vertical width of the box (the IQR). All observations outside the whiskers (crosses) are considered to be outliers. On the x axis, months 1–12 correspond to January–December, respectively.

  • View in gallery

    Box plots of the observed wind speed (m s−1) at Tuktoyaktuk by direction for the months of (a) July, (b) August, and (c) September. The wind speed is sorted by meteorological wind direction into bins of 30° and a box plot is created for each bin. The top and bottom of the box indicate the 75th and 25th percentile wind speeds, respectively. The line inside the box indicates the median wind speed. The length of the dashed lines (whiskers) is 1.5 times the IQR. All observations outside the whiskers (red crosses) are considered to be outliers.

  • View in gallery

    Wind roses of (a)–(c) observed and (d)–(f) 925-hPa geostrophic winds at Tuktoyaktuk (m s−1) for July, August, and September.

  • View in gallery

    The empirical histogram describing the 2D probability distribution of the 925-hPa geostrophic and observed wind directions at the same observation time for (a) July, (b) August, and (c) September. The black line indicates where the 925-hPa geostrophic and surface winds are in the same direction. The histogram is normalized so that the units are the percentage of total observations.

  • View in gallery

    (a) The empirical CDF of the 975-hPa stability [θz; the vertical derivative of potential temperature with height (K m−1)] for JAS. The box plots show the distribution of the stability index for cases of nonnorthwesterly winds (Other; 45°–225°), northwesterly winds (NW; 290°–340°), and extreme northwesterly winds (NWext; northwesterly with speed >9 m s−1) for all hourly observations during the months of (b) July, (c) August, and (d) September.

  • View in gallery

    The 2D histogram of the 850-hPa winds from the NARR dataset and observed surface winds for (top) nonnorthwesterly (Other) and (middle) northwesterly (NW) winds for (a),(d) July, (b),(e) August, and (c),(f) September. The solid line in each plot is the best-fit line from the linear regression between the 850-hPa and observed surface winds; the dashed line indicates where the surface and 850-hPa wind speeds are equal. The correlation (R) is also shown for each case. (bottom) Correlation between the surface winds and the winds on several pressure levels for cases of northwesterly (dashed) and nonwesterly surface (solid) winds for (g) July, (h) August, and (i) September.

  • View in gallery

    Climatological SLP (solid contours) and 925-hPa temperature (dashed contours) for (a) July, (b) August, and (c) September. The gray circle marks the location of Tuktoyaktuk.

  • View in gallery

    (a)–(c) NARR mean 10-m wind anomaly and wind speed anomalies for extreme northwesterly wind events at Tuktoyaktuk for July, August, and September. The wind barbs are the anomalous 10-m wind and the contours are the anomalous 10-m wind speed. Both are in knots. (d)–(f) Composite mean SLP anomaly for extreme northwesterly wind events as in (a)–(c). For the SLP anomalies, negative contours are dashed and the 0 contour is highlighted. The contour interval is 1 hPa for both. The composites represent the mean of 25, 44, and 30 events with 323, 405, and 411 three-hourly observations in July, August, and September, respectively. The gray circle marks the location of Tuktoyaktuk.

  • View in gallery

    The extreme northwesterly wind event composite mean of 925-hPa wind (barbs; kt) and temperature (solid contour; K). Also shown is the temperature anomaly (dashed contour; beginning at 3 K). The gray circle marks the location of Tuktoyaktuk.

  • View in gallery

    Composite temperature and wind profile for all high northwesterly wind events at Tuktoyaktuk.

  • View in gallery

    (a),(b) The SLP (contour) and 10-m wind (barbs; kt) for (a) 1200 UTC 24 Sep and (b) 0000 UTC 25 Sep 1999. (c),(d) As in (a),(b), but for the 950-hPa wind (barb; kt) and potential temperature (shaded; K).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 60 55 3
PDF Downloads 47 45 3

Wind Regimes along the Beaufort Sea Coast Favorable for Strong Wind Events at Tuktoyaktuk

View More View Less
  • 1 Department of Atmospheric and Oceanic Sciences, McGill University, Montreal, Quebec, Canada
© Get Permissions
Full access

Abstract

The community of Tuktoyaktuk (Northwest Territories, Canada) along the Beaufort Sea experiences dramatic shoreline erosion during storm surge events that tend to occur during persistent northwesterly wind events in the late summer months (July–September) when the sea ice coverage of the Beaufort Sea reaches its annual minimum. This study compiles the climatology of hourly surface wind, low-level geostrophic wind, and static stability to investigate the physical mechanisms responsible for the high frequency of northwesterly winds observed at Tuktoyaktuk during the late summer. The results link the prevalence of westerly to northwesterly winds at the surface to the high frequency of northwesterly geostrophic winds and a tendency for low static stability. With an environment that favors strong northwesterly geostrophic wind and suggests lower static stability, the high frequency of strong northwesterlies observed at the surface appears to be associated with momentum mixing by turbulent eddies. A composite analysis indicates that persistently strong northwesterly winds are associated with anomalously low pressure northeast of Tuktoyaktuk and high pressure over the Bering Sea and eastern Siberia. The high pressure anomalies over the Bering Sea also extend well to the east along the northern edge of the Brooks Range. An apparent topographic modification of the sea level pressure (SLP) field by cold air trapped to the north of mountains produces the pressure gradient favorable for strong westerly to northwesterly geostrophic winds at Tuktoyaktuk. The results suggest that cold-air damming contributes to the wind regime at Tuktoyaktuk by altering the pressure gradient along the Beaufort coast.

Corresponding author address: David Small, Dept. of Atmospheric and Oceanic Sciences, McGill University, Rm. 945, Burnside Hall, 805 Sherbrooke St. West, Montreal, QC H3A 2K6, Canada. E-mail: david.small2@mail.mcgill.ca

Abstract

The community of Tuktoyaktuk (Northwest Territories, Canada) along the Beaufort Sea experiences dramatic shoreline erosion during storm surge events that tend to occur during persistent northwesterly wind events in the late summer months (July–September) when the sea ice coverage of the Beaufort Sea reaches its annual minimum. This study compiles the climatology of hourly surface wind, low-level geostrophic wind, and static stability to investigate the physical mechanisms responsible for the high frequency of northwesterly winds observed at Tuktoyaktuk during the late summer. The results link the prevalence of westerly to northwesterly winds at the surface to the high frequency of northwesterly geostrophic winds and a tendency for low static stability. With an environment that favors strong northwesterly geostrophic wind and suggests lower static stability, the high frequency of strong northwesterlies observed at the surface appears to be associated with momentum mixing by turbulent eddies. A composite analysis indicates that persistently strong northwesterly winds are associated with anomalously low pressure northeast of Tuktoyaktuk and high pressure over the Bering Sea and eastern Siberia. The high pressure anomalies over the Bering Sea also extend well to the east along the northern edge of the Brooks Range. An apparent topographic modification of the sea level pressure (SLP) field by cold air trapped to the north of mountains produces the pressure gradient favorable for strong westerly to northwesterly geostrophic winds at Tuktoyaktuk. The results suggest that cold-air damming contributes to the wind regime at Tuktoyaktuk by altering the pressure gradient along the Beaufort coast.

Corresponding author address: David Small, Dept. of Atmospheric and Oceanic Sciences, McGill University, Rm. 945, Burnside Hall, 805 Sherbrooke St. West, Montreal, QC H3A 2K6, Canada. E-mail: david.small2@mail.mcgill.ca

1. Introduction

The coastal community of Tuktoyaktuk along the Beaufort coast of the Northwest Territories in the western Canadian Arctic is located on a narrow, low-lying peninsula in the Kugmallit Bay (Fig. 1) with a history of serious coastal erosion (Danard et al. 2003; Solomon 2005). Most of the coastal erosion damage along the Beaufort coast is produced by wind-induced storm surge events that occur during periods of persistent northwesterly winds frequently observed along the Beaufort coast in late summer when the Beaufort Sea is generally ice free (Manson and Solomon 2007; Jones et al. 2009). Although storm surges can occur any time of the year, the most dramatic impacts at Tuktoyaktuk are observed during the summer and early fall (July–September) when the breakup of the sea ice coverage from the previous winter results in large areas of open water.

Fig. 1.
Fig. 1.

Map of the Beaufort coast showing the location of the Tuktoyaktuk station and the surface elevations of the local topography.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

Global warming has led to dramatic changes in the environment of the Arctic, with the most rapid increases in temperature (Comiso 2003) and decreases in sea ice (Barber and Hanesiak 2004) occurring over the western Canadian Arctic, Beaufort Sea, and Mackenzie River basin. Increasing temperatures and changes in circulation have led to dramatic decreases in the sea ice extent, which eventually reached record minima in 2002 and again in 2007 (Cavalieri et al. 2003; Serreze et al. 2003, 2007; Stroeve et al. 2007). The increase in open water and degradation of the permafrost (Jorgenson et al. 2006) from the rapid warming have in turn led to large increases in the rate of coastal erosion in many regions of the Arctic, including the Beaufort coast, where the rates are already among the highest the world (Solomon 2005; Jones et al. 2009). The latest Intergovernmental Panel on Climate Change (IPCC) report (Christensen et al. 2007) indicates that the increase in temperature and decline in sea ice coverage over the Arctic is “very likely” to continue over the course of the twenty-first century. This has led some authors to suggest that coastal erosion will also increase in the Arctic (Drobot and Maslanik 2003; Simmonds et al. 2008), including along the Beaufort coast (Manson and Solomon 2007; Jones et al. 2009). The IPCC models, however, are less certain about wind and circulation changes in the Arctic than about changes in temperature over the next century (Christensen et al. 2007). This uncertainty highlights the importance of understanding the wind regimes of the Arctic, especially those that affect the fastest changing regions such as the Beaufort coast in the western Canadian Arctic. This study is motivated by the challenge of understanding the physical processes that contribute to the wind regime along the Beaufort coast, a critical component of the complex and fast-changing Arctic climate system.

Previous studies have found that the wind regime near Tuktoyaktuk is bimodal during the late summer with high frequencies of westerly (northwesterly) and southeasterly winds (Solomon 2005; Manson and Solomon 2007; Atkinson 2005; Hudak and Young 2002). The prevalence of westerly winds along the Beaufort coast has often been linked to the passage of powerful storms (Hudak and Young 2002; Atkinson 2005; Manson and Solomon 2007). Although cyclone activity in the Arctic reaches its maximum in summer, the Beaufort Sea is a region with relatively few cyclones (Keegan 1958; Serreze et al. 1993; Serreze 1995; McCabe et al. 2001; Lynch et al. 2004; Zhang et al. 2004; Simmonds et al. 2008) and among the highest frequency of anticyclones in the Arctic (Colucci and Davenport 1987; Serreze et al. 1993). The high frequency of anticyclones has been attributed to a persistent ridge that is often located over the Siberia–Chukchi–Beaufort region during fall and winter (Serreze et al. 1993; Serreze and Barrett 2008). High wind events at Barrow, a community located at the western edge of the Beaufort Sea, are often caused by synoptic conditions other than strong cyclones, including the presence of a persistent ridge over the Beaufort–Chukchi region (Lynch et al. 2004). The potential impact of a persistent ridge over the Bering Sea on the wind regime at Tuktoyaktuk in the late summer months has not been investigated.

In this study, we investigate the mechanism favoring prevalent northwesterly winds at Tuktoyaktuk by addressing two primary objectives. The first objective is to document the surface wind climatology at Tuktoyaktuk for the months of July, August, and September (JAS) to identify the presence of a preferred northwesterly wind regime and to investigate the possible contribution of momentum mixing. The second goal is to indentify the circulation patterns conducive for persistent northwesterly winds at Tuktoyaktuk and discuss the possible role of topography.

2. Data

Several sources of data are utilized in this study. The hourly surface wind observations from the meteorological station at the Tuktoyaktuk airport [World Meteorological Organization (WMO) station 71954; 69.4°N, 133.0°W] for the period 1971 to 2006 are used to compile the climatology of the wind observations. For most of the available record, observations from 0500 to 1100 UTC are not reported at Tuktoyaktuk, but the wind data are otherwise over 94% complete. The surface wind direction is reported with a precision of 10° resulting in 36 possible wind directions. The observed surface wind data are used where possible because the observations are believed to be more reliable than either the National Centers for Environmental Prediction (NCEP)–National Center for Atmospheric Research (NCAR) reanalysis (Kalnay et al. 1996) or North American Regional Reanalysis (NARR; Mesinger et al. 2006) in the Arctic where the data assimilation model is less constrained by observations owing to the sparseness of observations.

Geopotential height data used to calculate the geostrophic wind at Tuktoyaktuk are from the NCEP–NCAR global reanalysis at a 2.5° horizontal resolution. The geostrophic wind at 925 hPa was calculated from the geopotential height and then linearly interpolated to the Tuktoyaktuk station. The NCEP–NCAR global reanalysis is used to calculate the geostrophic wind because it contains data for the entire period when surface wind observations are available at Tuktoyaktuk (1971–2006). Temperature, wind, and sea level pressure (SLP) data used to create composites of persistent wind events are from the NARR data on a 32-km Lambert conformal grid for the period 1979 to 2006. NARR is used for the composites because the model assimilates IR surface brightness temperature and precipitation and provides a finer horizontal and vertical resolution, including a 25-hPa level spacing in the boundary layer.

3. Methods and results

A map is presented in Fig. 1 that shows the locations of Tuktoyaktuk and several geographic features that are mentioned this study, including the Barents, Chukchi, and Beaufort Seas, the Brooks Range, and the Gulf of Alaska. The map shows that the topography of the Beaufort coast is dominated by the confluence of the east–west-oriented Brooks Range and the northern extent of the Rocky Mountains. Tuktoyaktuk is located several hundred kilometers to the west of where the Brooks Range turns north from the interior of Alaska and becomes oriented in the northwest–southeast direction parallel to the Beaufort coast near the Canadian border.

a. Hourly wind climatology for Tuktoyaktuk

The seasonal cycle of hourly wind speed at the Tuktoyaktuk station is depicted in Fig. 2. Box-and-whisker plots (box plots) of the hourly wind speeds from each month are used to show the seasonal cycle instead of the monthly mean to allow for the non-Gaussian distribution of the hourly wind speeds. Box-and-whisker plots allow the shape and location (i.e., central tendency) of the distribution to be visualized without making any assumptions about the statistical distribution of the data. On a box-and-whisker plot of observed data, the top and bottom of the box indicate the 75th and 25th percentile values respectively. The location of the box therefore indicates where the middle 50% of the data lie, with the width of box representing the spread in the data. The line inside the box marks the location of the median (50th percentile) value and allows the skewness or asymmetry of the underlying distribution around the median to be visualized. Whiskers extend to the most extreme values within 1.5 times the interquartile range (IQR) from the ends of the box. The IQR is defined as the separation between the 25th and 75th percentile values and is often used as a measure of the spread in the distribution. Observations with values beyond the ends of the whiskers are assumed to be outliers and are indicated by crosses.

Fig. 2.
Fig. 2.

Box plots of the hourly wind speed (m s−1) distribution by month. The top and bottom of the box indicate the 75th and 25th percentile wind speeds, respectively. The line inside the box indicates the median wind speed. The length of the dashed lines (whiskers) is 1.5 times the vertical width of the box (the IQR). All observations outside the whiskers (crosses) are considered to be outliers. On the x axis, months 1–12 correspond to January–December, respectively.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

The box-and-whisker plots in Fig. 2 indicate that the monthly distribution of hourly wind speeds is slightly asymmetric, with the shape and location of the distribution depending on the season. The median wind speed is larger in the warm season (May–September) than in the cold season (October–April), but the distribution is more positively skewed (i.e., toward the right tail or the larger values) in the winter months than during the summer. The IQR (the vertical width of the box) is also smaller during the warm season months of March–July and is narrowest in June, indicating that the wind speed is less variable in the summer than the winter. The largest spread of observed wind speed is observed in the cold months of December–February. The box-and-whisker plots demonstrate that although the median wind speed is slightly larger in the warm season, the distribution of the wind speed during the cold season is wider (more variable) and more heavily skewed toward larger wind speeds. The box-and-whisker plots demonstrate that larger surface wind speeds are observed more frequently at Tuktoyaktuk during the cold season than during the summer. In our period of interest, July–September when the Beaufort Sea is generally ice free, the median wind speed is slightly larger than in winter and more variable and skewed toward higher values than earlier in the summer.

The previous discussion suggests that the shape of the wind speed distribution at Tuktoyaktuk, including the extreme values, depends on the season, but the potential dependence of the wind speed on direction has not been examined. The following analysis of the wind speed and direction focuses on July, August, and September because this is the period when the sea ice coverage declines and the storm surge risk at Tuktoyaktuk is greatest. The observed hourly wind speeds for July, August, and September were separated into bins of 30° of meteorological wind direction (measured clockwise in degrees from north), and a box-and-whisker plot is constructed for the wind speeds in each bin.

The box plots of the wind speed by direction (Fig. 3) indicate that the entire distribution of hourly wind speed is shifted toward higher wind speeds for meteorological wind directions between 270° and 330° (from westerly to northwesterly), the direction parallel to the upstream topography and coincident with the direction of the longest fetch over open water. The observed shift in the distribution toward higher wind speeds is much larger in August and September than in July. The box plots also indicate that southeasterly winds are slightly enhanced, but the shift in the distribution is much smaller than for northwesterly winds. For wind observations with meteorological directions other than northwesterly or southeasterly, the distribution of the wind speed is very similar and displays little dependence on wind direction. This clear dependence of the wind speed on direction is an important consideration for statistical modeling of wind speeds at Tuktoyaktuk. Because the distribution of the wind speed depends so strongly on the wind direction, the wind speed observations are clearly not independent, identically distributed random variables. Any statistical modeling of the wind speed must therefore explicitly take into account the directional dependence.

Fig. 3.
Fig. 3.

Box plots of the observed wind speed (m s−1) at Tuktoyaktuk by direction for the months of (a) July, (b) August, and (c) September. The wind speed is sorted by meteorological wind direction into bins of 30° and a box plot is created for each bin. The top and bottom of the box indicate the 75th and 25th percentile wind speeds, respectively. The line inside the box indicates the median wind speed. The length of the dashed lines (whiskers) is 1.5 times the IQR. All observations outside the whiskers (red crosses) are considered to be outliers.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

The entire distribution of the wind speed is preferentially shifted toward higher values for westerly and northwesterly winds (Fig. 3), suggesting a physical mechanism that produces larger wind speeds for northwesterly winds all across the distribution and not just in the right tail (extreme values). The direction of the observed shift in the distribution of the wind speed is approximately parallel to the northwest to southeast orientation of the Brooks Range and suggests a role for the upstream topography in the modification of the wind speed. The shift in the distribution also coincides with the direction of the longest fetch over open water, suggesting that lower friction over the open water surface might also be an important consideration. In the next section, the observed surface winds are directly compared to the geostrophic winds aloft to link the northwesterly wind regime at Tuktoyaktuk to identify the possible role of momentum mixing.

b. Possible physical mechanism for bimodality of winds at Tuktoyaktuk

The results presented in the previous section documented a systematic modification of the wind speed distribution at Tuktoyaktuk for northwesterly winds. Two possible contributors to the preferential wind regime are topography (as will be discussed in section 3c) and momentum mixing by turbulent eddies (Whiteman and Doran 1993). When the geostrophic wind aloft and the observed surface wind are in the same direction, the downward mixing of momentum to the surface by turbulent eddies is likely to be a significant factor in the observed wind regime. Otherwise, we would expect a turning of the surface winds with respect to the geostrophic wind because of Ekman veering. To identify the signature of momentum mixing, wind roses of the observed surface wind and the 925-hPa geostrophic wind were constructed for July, August, and September to compare the distribution of the observed and geostrophic wind directions. The 925-hPa level is chosen because it is below the top of the topography and above the surface. The observed surface wind direction is then directly compared to the 925-hPa geostrophic wind at the corresponding 6-h period in the NCEP–NCAR reanalysis data by constructing two-dimensional histograms that define the probability distribution of wind events in terms of the geostrophic and observed surface wind directions. We constructed the histograms by binning the surface wind and 925-hPa geostrophic wind directions at the same observation (analysis) time into the same 10° bins and counting the number of occurrences for each possible combination of wind directions. The probabilities are normalized by the total number of occurrences and represent the percentage of the total number of observations.

The wind roses of observed winds indicate that the July surface winds are primarily northwesterly or northerly to easterly (from the water toward the land) and that the wind rarely blows from the land to the water (Fig. 4a). The relatively high frequency of flow from the cold water to the warm land brings stable air over the land and is consistent with the lower wind speeds observed in July compared to August or September. In August, the distribution of the surface wind exhibits weak bimodality with a higher frequency of northwesterly than easterly winds (Fig. 4b), though nearly all of the extreme winds in August (speeds greater than 10 m s−1) are northwesterly. The distribution of the wind direction in September (Fig. 4c) is also bimodal, but with a higher frequency of large easterly winds. The wind roses indicate a preference for northwesterly winds in each month, consistent with the histograms and box plots in Fig. 3. To identify the signature of momentum mixing and link the observed distribution of wind speeds to the topography, the observed surface wind is related to the overlying geostrophic wind and the stability of the atmospheric column.

Fig. 4.
Fig. 4.

Wind roses of (a)–(c) observed and (d)–(f) 925-hPa geostrophic winds at Tuktoyaktuk (m s−1) for July, August, and September.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

The geostrophic wind rose for July (Fig. 4d) indicates that most of the large 925-hPa geostrophic winds are westerly or northwesterly. In August (Fig. 4e), the geostrophic wind is bimodal, but with a much higher frequency of northwesterly geostrophic winds than southeasterly, especially those exceeding 10 m s−1. The September geostrophic wind (Fig. 4f) is also bimodal but with more large southeasterly winds than in August, similar to the distribution of the observed winds at the surface. The agreement between the geostrophic wind and the observed surface wind is greatest in September. The two-dimensional histograms of the observed and geostrophic wind directions are used to directly compare the observed and geostrophic wind directions.

In all three months, the histograms indicate that the nonzero probabilities tend to fall along the line corresponding to geostrophic and observed surface winds that are in the same direction (Figs. 5a–c). In July and September, the histograms suggest that the most frequently observed flow is southeasterly 925-hPa geostrophic and easterly surface winds with a lower frequency of northwesterly observed and geostrophic winds. The flow is significantly different in August, with the histogram indicating a much higher frequency of observations where the surface and 925-hPa geostrophic winds are both northwesterly than in July or September. This reflects the higher frequency of northwesterly geostrophic winds observed in August.

Fig. 5.
Fig. 5.

The empirical histogram describing the 2D probability distribution of the 925-hPa geostrophic and observed wind directions at the same observation time for (a) July, (b) August, and (c) September. The black line indicates where the 925-hPa geostrophic and surface winds are in the same direction. The histogram is normalized so that the units are the percentage of total observations.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

The high frequency of observed surface westerlies at Tuktoyaktuk appears to be associated with the high frequency of westerly to northwesterly geostrophic winds aloft. This is consistent with the suggestion that momentum mixing is an important factor in the enhancement of the northwesterly winds but does not preclude a role for other processes, especially reduced surface friction along the long northwesterly fetch over open water. For momentum mixing to be the dominant mechanism, we expect to observe neutral or unstable static stability and stronger mixing in the case of northwesterly surface winds. In the following paragraphs we document the differences in static stability and mixing during periods of westerly and nonwesterly surface winds.

We evaluated the static stability by calculating ∂θ/∂z(θz) at 975 hPa with centered finite differences and estimating the empirical cumulative distribution function (CDF) for each month from the NARR data. Here, θz is the derivative of the potential temperature with height in units of kelvins per meter. The vertical derivative is calculated with finite differences calculated between the 950- and 1000-hPa pressure surfaces. The NARR data are used because the fine resolution in the boundary layer is advantageous for estimating the stability in the layer below the top of the topography. The CDF of θz for July, August, and September (Fig. 6a) is evaluated to quantify the seasonal differences in the stability, which likely contribute significantly to the observed seasonality in the wind speed and direction. Box plots of θz during periods when the observed surface winds are northwesterly (290°–340°), northwesterly and extreme (northwesterly with a wind speed exceeding 9 m s−1), or nonnorthwesterly (45°–225°) were also constructed for each month to visualize the dependence of the observed surface winds on the stability (Figs. 6b–d).

Fig. 6.
Fig. 6.

(a) The empirical CDF of the 975-hPa stability [θz; the vertical derivative of potential temperature with height (K m−1)] for JAS. The box plots show the distribution of the stability index for cases of nonnorthwesterly winds (Other; 45°–225°), northwesterly winds (NW; 290°–340°), and extreme northwesterly winds (NWext; northwesterly with speed >9 m s−1) for all hourly observations during the months of (b) July, (c) August, and (d) September.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

The CDF plot of θz indicates that the stability of the atmosphere is quite different in July, August, and September (Fig. 6a). The CDFs are shifted such that in July the atmosphere is less likely to have negative (unstable) or extremely small (neutral) values than in August, which in turn is less likely to be unstable or neutral than September. The box plots of θz indicate differences in the stability during periods when northwesterly winds are observed at Tuktoyaktuk compared to periods when the winds are not northwesterly. For July, the box plots indicate small decreases in the stability between periods of northwesterly and nonnorthwesterly winds. In August, the distribution of θz is shifted toward smaller values for northwest surface winds compared with nonnorthwesterly winds. For the case of the extreme northwesterly winds, the median is smaller and the distribution is more skewed toward lower stability. In September, the distribution of the stability is also more highly skewed and shifted toward smaller values for the case of northwesterly winds, although the shift toward lower stabilities for the extreme northwesterly winds is not as pronounced as in August. To confirm that the lower stabilities are conducive for momentum to mix to the surface, the wind speeds at different pressure levels in the atmosphere are directly compared to the surface wind observations.

The NARR zonal and meridional winds on pressure levels were linearly interpolated to the location of the Tuktoyaktuk station. The observed surface wind speed at Tuktoyaktuk is directly compared to the 850-hPa wind speed with two-dimensional histograms of the 850 hPa and observed surface wind speeds at all 3-hourly observation times of the NARR data. Linear regression is also used to estimate the strength of the linear association between the winds at the two levels. The histogram and linear regression are estimated using those observations with westerly to northwesterly winds (270°–360°; Figs. 7a–c) and nonnorthwesterly winds (45°–225°; Figs. 7d–f) at the surface in July, August, and September. The surface wind speed is then correlated to the wind speed at each pressure level from 850 to 1000 hPa for cases of northwesterly and nonnorthwesterly winds to demonstrate the importance of mixing through the atmospheric column.

Fig. 7.
Fig. 7.

The 2D histogram of the 850-hPa winds from the NARR dataset and observed surface winds for (top) nonnorthwesterly (Other) and (middle) northwesterly (NW) winds for (a),(d) July, (b),(e) August, and (c),(f) September. The solid line in each plot is the best-fit line from the linear regression between the 850-hPa and observed surface winds; the dashed line indicates where the surface and 850-hPa wind speeds are equal. The correlation (R) is also shown for each case. (bottom) Correlation between the surface winds and the winds on several pressure levels for cases of northwesterly (dashed) and nonwesterly surface (solid) winds for (g) July, (h) August, and (i) September.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

The two-dimensional histograms of the surface and 850-hPa winds in the case of nonnorthwesterly surface wind indicate that the relationship between the winds at the two levels is extremely weak in July, August, and September (Figs. 7a–c). The magnitude of the correlation between the 850 hPa and surface wind speeds is very small, suggesting that momentum mixing from 850 hPa to the surface does not explain a significant fraction of the variance of nonnorthwesterly surface winds. For the case of northwesterly surface winds (Figs. 7d–f), the histograms and regression analysis suggest that the strength of the linear association between the wind speeds at the two levels is much stronger. The histograms indicate that the distribution is also shifted toward higher values, suggesting that northwesterly surface winds are associated with higher wind speeds aloft. Positive correlations of 0.32, 0.47, and 0.39 are observed in July, August, and September, respectively, consistent with the suggestion that momentum mixing is an important mechanism for the enhancement of the northwesterly winds at Tuktoyaktuk. The high correlation between the surface wind speed and the wind speed at higher levels also supports this suggestion. When the surface winds are northwesterly, the magnitude of the correlation with the 850-hPa winds is high and remains relatively constant throughout the boundary layer down to 1000 hPa (Figs. 7g–i). When the surface winds are not northwesterly, the correlation with the 850-hPa winds is nearly 0 and only becomes large near the surface. The magnitude of the correlation is larger in August than during July or September and reaches approximately 0.60 at the 950-hPa level. The results suggest that westerly surface winds tend to increase when stronger winds are observed aloft. This is not the case for nonwesterly winds, consistent with the box plots in Fig. 3, which showed that only the distribution of northwesterly winds are significantly enhanced.

The results suggest that the high frequency of northwesterly surface winds observed at Tuktoyaktuk in the late summer months is associated with a high frequency of northwesterly geostrophic winds and stability conditions favorable for momentum to be mixed to the surface. In the next section, we describe the creation of SLP, wind, and temperature composites used to document the circulation anomalies favorable for persistent northwesterly wind events.

c. Climatology and composites of persistent northwesterly wind events

The monthly climatology of the SLP and 925-hPa temperature over the Beaufort coast and North Pacific is shown in Fig. 8. The climatology represents the SLP and 925-hPa temperature calculated from the 3-hourly NARR values over the period from 1979 to 2006. The climatological mean 925-hPa temperature field indicates a tight temperature gradient stretching from the Siberian coast across the Chukchi and Beaufort Seas with temperatures decreasing from July into August and September. The temperature gradient suggests the presence of a baroclinic zone along the coast, although one should be cautious when inferring baroclinicity from time mean fields (Simmonds and Lim 2009). Several interesting features are observed from the mean SLP field. In July, a closed surface anticyclone is observed over the Beaufort Sea with lower pressure over the continent producing a strong pressure gradient parallel to the coast favorable for easterly or northerly winds, consistent with the observed wind roses for July (Fig. 4). The closed contour in the high pressure is still evident in the August climatology (though 2 hPa weaker) and is no longer present in September. The mean SLP over the Beaufort does not decrease significantly from July to September, but the SLP gradient weakens dramatically in August and September. As the ice melts and the water surface warms, the climatological pressure gradient favorable for easterly or on-shore winds in July weakens. This is consistent with the observed wind roses at Tuktoyaktuk, which indicate that onshore flow is much more frequent in July than August or September. The monthly climatology explains the decrease in the frequency of onshore flow at Tuktoyaktuk in August and September but yields no insight into the atmospheric conditions favorable for persistent northwesterly wind events. Composites of SLP, 10-m and 925-hPa wind, and 925-hPa temperature during persistent high northwesterly wind events at Tuktoyaktuk were created for the relevant month to identify the atmospheric conditions favorable for such events.

Fig. 8.
Fig. 8.

Climatological SLP (solid contours) and 925-hPa temperature (dashed contours) for (a) July, (b) August, and (c) September. The gray circle marks the location of Tuktoyaktuk.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

Persistent high wind events were identified directly from the hourly station observations at Tuktoyaktuk. Any continuous period with wind speeds exceeding the 90th percentile value of 9 m s−1 and a mean wind direction exceeding 260° was considered to be an intense northwesterly wind event. The minimum separation between independent events was considered to be 6 h. Any period with six or fewer missing values adjacent to periods with 90th percentile wind speeds is therefore considered to be part of the same event. This allows otherwise continuous periods of intense northwesterly winds to span the 6-h period of missing observations between 0500 and 1100 UTC over most of the record. For this study, only the top 25% of events based on duration (those lasting 12 h or longer) are included in the composites. This criterion yielded 99 events, with 25 in July, 44 in August, and 30 in September. Composites of SLP, 10-m wind, and 925-hPa wind and temperature were created by averaging over the all the 3-hourly periods identified as being part of an extreme northwesterly wind event. Anomalies are defined as the difference between the composite mean and the climatological mean (1979–2006).

The surface (10 m) wind and SLP anomalies from the NARR are shown in Fig. 9. The anomalous wind is plotted as a wind barb with contoured isotachs of the wind speed anomaly. The wind anomalies are defined as the difference between the composite mean wind speed and the 1979–2006 mean of the wind speed. During the wind events, anomalous northwesterly winds are observed along the Beaufort coast north of Tuktoyaktuk in all three months, with speed anomalies as large as 15 kt (1 kt = 0.5144 m s−1) in August (Figs. 9a–c). The anomalies are larger in August and September than in July, consistent with the station observations at Tuktoyaktuk.

Fig. 9.
Fig. 9.

(a)–(c) NARR mean 10-m wind anomaly and wind speed anomalies for extreme northwesterly wind events at Tuktoyaktuk for July, August, and September. The wind barbs are the anomalous 10-m wind and the contours are the anomalous 10-m wind speed. Both are in knots. (d)–(f) Composite mean SLP anomaly for extreme northwesterly wind events as in (a)–(c). For the SLP anomalies, negative contours are dashed and the 0 contour is highlighted. The contour interval is 1 hPa for both. The composites represent the mean of 25, 44, and 30 events with 323, 405, and 411 three-hourly observations in July, August, and September, respectively. The gray circle marks the location of Tuktoyaktuk.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

The composite SLP anomalies (Figs. 9d–f) indicate anomalous low pressure to the northeast of Tuktoyaktuk and higher pressure to the west along the Beaufort coast of Alaska but with significant seasonal differences in the magnitude and location of the anomalies. For events in July, anomalous low pressure to the northeast of Tuktoyaktuk and a weaker high pressure anomaly along the north coast of Alaska produce an enhanced pressure gradient favorable for strong northwesterly geostrophic winds at Tuktoyaktuk. During events in August and September (Figs. 9e,f), the SLP anomalies form a dipole with anomalous low pressure to the northeast of Tuktoyaktuk and positive anomalies along the western coast of Alaska that stretch from the Bering Sea over western Siberia. The positive SLP anomalies indicate that a persistent ridge located over Siberia in the composite mean (not shown) is significantly strengthened during the high northwesterly events at Tuktoyaktuk. In both August and September, the large positive SLP anomalies over the Bering Sea are displaced to the east along the Brooks Range, parallel to the north coast of Alaska. The elongation of the positive SLP anomalies along the mountains produces an enhanced northeast to southwest pressure gradient over the Beaufort Sea that is conducive for strong northwesterly geostrophic winds at Tuktoyaktuk. The U-shaped pressure anomalies along the mountains resemble those observed during cold-air damming events in the Rocky or Appalachian Mountains (Bell and Bosart 1988). To examine the role of trapped cold air along the northern edge of the Brooks Range in producing increased pressure, the composite mean 925-hPa temperature and wind and 925-hPa temperature anomaly are shown in Fig. 10.

Fig. 10.
Fig. 10.

The extreme northwesterly wind event composite mean of 925-hPa wind (barbs; kt) and temperature (solid contour; K). Also shown is the temperature anomaly (dashed contour; beginning at 3 K). The gray circle marks the location of Tuktoyaktuk.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

The composite mean 925-hPa temperature field shows that the baroclinic zone along the coast is significantly enhanced compared to climatology with the strongest temperature gradients very closely following the Brooks Range. The temperature anomalies demonstrate that the enhanced temperature gradients are associated with temperatures that are as much as 8° colder than climatology along the northern edge of the topography, with the coldest anomalies observed in August. The positive SLP and negative temperature anomalies north of the Brooks Range are consistent with cold-air damming as a possible mechanism contributing to the observed positive pressure anomalies.

The pattern of the composite mean 925-hPa wind demonstrates the importance of the anticyclone over the Bering Sea. The observed anticyclonic turning of the winds is conducive for cold air to become trapped along the northern edge of the Brooks Range. Such winds are conducive for both cold advection along the north coast of Alaska or adiabatic cooling from upslope flow forced by the turning of the westerly winds to the south by the Coriolis effect. The anomalously cold temperatures cause the pressure to rise north of the Brooks Range, producing an enhanced pressure gradient that is conducive for northwesterly geostrophic winds well to the east at Tuktoyaktuk. In all three months, the composite mean shows that the strongest 925-hPa winds near Tuktoyaktuk are perpendicular to the enhanced pressure gradient that forms along the coast parallel to the Brooks Range.

Persistent northwesterly wind events at Tuktoyaktuk tend to occur with cold temperatures and anomalous high pressure north of the Brooks Range. The map of 925-hPa temperature anomalies indicates that the coldest temperature anomalies are observed near Tuktoyaktuk. Composite temperature and wind profile were created at Tuktoyaktuk to confirm that momentum mixing occurs during the events. The temperature and wind were interpolated to the location of the Tuktoyaktuk station and a composite mean of the temperature and wind was calculated. The composite profiles for all the events without sorting by month are shown in Fig. 11. The temperature profile indicates a shallow mixed layer that extends throughout the lowest 100 hPa of the atmosphere. The wind profile also indicates that the winds throughout the lowest layers of the atmosphere are all northwesterly with similar wind speeds. Together, this confirms that momentum mixing contributes to persistent northwesterly surface winds at Tuktoyaktuk.

Fig. 11.
Fig. 11.

Composite temperature and wind profile for all high northwesterly wind events at Tuktoyaktuk.

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

The composites suggest that the westerly wind regime at Tuktoyaktuk is associated with synoptic conditions favorable for a tight pressure gradient to produce strong northwesterly geostrophic winds along the coast. Cold-air damming, with the associated hydrostatic pressure rise along the northern edge of the Brooks Range appears to contribute to the enhanced pressure gradient. The composites likely smear the SLP, temperature, and wind fields because of the large number of observations included in the averages. The following example shows the SLP, wind, and temperature fields from one of the events included in the composite to demonstrate that conditions similar to the composites are observed during individual events.

d. Example from a persistent wind event at Tuktoyaktuk

Between 24 and 28 September 1999, Tuktoyaktuk experienced significant storm surge damage during a period of persistently high northwesterly winds (S. Solomon 2008, personal communication). The winds at Tuktoyaktuk turned northwesterly at approximately 1800 UTC 23 September after a cyclone that developed in the lee of the Rocky Mountains to the south and west of Tuktoyaktuk during the previous 12 h moved off to the east. Maps of the SLP, 10-m wind, and 950-hPa potential temperature and wind from the NARR are used to demonstrate that the composites describe realistic physical conditions (Fig. 12).

Fig. 12.
Fig. 12.

(a),(b) The SLP (contour) and 10-m wind (barbs; kt) for (a) 1200 UTC 24 Sep and (b) 0000 UTC 25 Sep 1999. (c),(d) As in (a),(b), but for the 950-hPa wind (barb; kt) and potential temperature (shaded; K).

Citation: Journal of Applied Meteorology and Climatology 50, 6; 10.1175/2010JAMC2606.1

At 0012 UTC 24 September, an anticyclone is observed along the Beaufort coast of western Alaska that along with the cyclone east of Tuktoyaktuk forms part of a deformation zone over the Beaufort Sea. At this time, strong northerly winds at 10 m and 950 hPa are observed in the confluence zone between the cyclone–anticyclone pair over the Chukchi Sea. The SLP contours along the Alaska coast north of the Brooks Range display the characteristic U-shaped ridge that is commonly observed during cold-air damming events in the Appalachians (Bell and Bosart 1988). The 950-hPa potential temperature field also indicates that a pool of potentially cold air has formed near the surface north of the mountains. The U-shaped ridge in the SLP field creates a tight pressure gradient favorable for strong northwesterly geostrophic winds along the Beaufort coast. Twelve hours later, at 0000 UTC 25 September, the anticyclone west of Tuktoyaktuk is still present while the surface cyclone is now located well to the east of Tuktoyaktuk. The pool of potentially cold air has moved to the east, following the coastal topography and the SLP has risen along the northern edge of the topography. The resulting increase in the pressure gradient north of the topography has caused the northwesterly winds at the surface to strengthen.

This example demonstrates the importance of a surface anticyclone over the Chukchi Sea that becomes displaced to the east along the northern edge of the Brooks Range in producing extreme northwesterly winds at Tuktoyaktuk. The anticyclonic flow over the Beaufort Sea is conducive for cold air advection and upslope flow to build a pool of cold air at low levels along the windward edge of the Brooks Range. The buildup of cold air causes the pressure to rise along the topography, contributing to a pressure gradient favorable for persistent northwesterly geostrophic winds at Tuktoyaktuk.

4. Discussion

In this study, the climatology of late summer (July–September) surface winds at Tuktoyaktuk was developed. We have documented that the distribution of westerly to northwesterly winds is shifted toward higher wind speeds compared to winds in other directions. The results demonstrate that northwesterly surface winds tend to occur when the overlying geostrophic winds are also northwesterly. Strong correlations between the surface winds and the winds throughout the atmospheric column from 1000 up to 850 hPa and a tendency for lower stability are also observed when the surface winds are northwesterly, strongly suggesting that the mixing of momentum throughout the atmospheric column is an important factor in the preferential enhancement of the northwesterly winds at the surface.

A composite analysis of persistent, high westerly to northwesterly wind events identified several interesting features of the synoptic-scale environment favorable for northwesterly geostrophic winds at Tuktoyaktuk. The composite mean indicates that persistently strong northwesterly winds are associated with anomalously low pressure northeast of Tuktoyaktuk and high pressure over the Bering Sea and eastern Siberia. This configuration of SLP anomalies alone, however, is not favorable for a tight pressure gradient and northwesterly geostrophic winds at Tuktoyaktuk. The high pressure anomalies over the Bering Sea also extend well to the east along the northern edge of the Brooks Range. The apparent topographic modification of the SLP field produces the pressure gradient favorable for strong westerly to northwesterly geostrophic winds at Tuktoyaktuk. Temperature composites indicate that the pressure rise along the Brooks Range is associated with anomalously cold temperatures, suggesting that cold-air damming might be contributing significantly to the wind regime at Tuktoyaktuk by altering the pressure gradient along the coast.

The anomalously strong ridge over eastern Siberia and the Bering Sea is a well-known climatological feature in fall and winter (Serreze et al. 1993; Serreze and Barrett 2008) that has been previously linked to the wind regime at Barrow (Lynch et al. 2004). Our composite suggests that the anomalously strong anticyclone also alters the wind regime at Tuktoyaktuk by producing northerly and westerly winds over the Beaufort that are favorable for the trapping of cold air north of the Brooks Range. The increase in the pressure north of the mountains that alters the pressure gradient near Tuktoyaktuk is the hydrostatic response to the low-level cooling. Although cold anomalies are observed along the coast, the composite temperature profile at the Tuktoyaktuk station indicates that a well-mixed layer is present during the persistent high wind events, consistent with the suggestion that momentum mixing is the primary mechanism behind persistent northwesterly wind events at Tuktoyaktuk.

The suggestion that momentum mixing is an important factor in the enhancement of the northwesterly winds does not preclude a role for other processes, including reduced surface friction along the long northwesterly fetch over open water or the passage of cyclones. The relative importance of sea ice and sea surface temperatures in creating the conditions favorable for reduced stability or the strong seasonal differences in stability characteristic are important considerations that have not been addressed in the study. The importance of surface fluxes in reducing the stability and causing conditions favorable for momentum mixing along the Beaufort coast is especially important given recent findings that have linked changing sea ice conditions and sea surface temperatures to rapid surface warming and an increase in storm activity across the Arctic (Simmonds and Keay 2009; Screen and Simmonds 2010a,b). The question of how the late summer wind regime along the Beaufort coast will respond to such changes is crucial for making realistic assessments of future changes in rates of coastal erosion damage along the Beaufort coast. Our results suggest that realistic assessments of future changes to the environment of the region will require a better understanding how synoptic-scale circulations, surface fluxes, and the upstream topography all contribute to the coastal wind regime.

Acknowledgments

The authors acknowledge funding from the Natural Sciences and Engineering Research Council of Canada (NSERC) International Polar Year (IPY) program, the Canadian Foundation for Climate and Atmospheric Sciences (CFCAS), and the MITACS (Mathematics of Information Technology and Complex Systems) program.

REFERENCES

  • Atkinson, D. E., 2005: Observed storminess patterns and trends in the circum-Arctic coastal regime. Geo-Mar. Lett., 25, 98109.

  • Barber, D. G., and J. M. Hanesiak, 2004: Meteorological forcing of sea ice concentrations in the southern Beaufort Sea over the period 1979 to 2000. J. Geophys. Res., 109, C06014, doi:10.1029/2003JC002027.

    • Search Google Scholar
    • Export Citation
  • Bell, G. D., and L. F. Bosart, 1988: Appalachian cold-air damming. Mon. Wea. Rev., 116, 137161.

  • Cavalieri, D. J., C. L. Parkinson, and K. Y. Vinnikov, 2003: 30-year satellite record reveals contrasting Arctic and Antarctic decadal sea ice variability. Geophys. Res. Lett., 30, 1970, doi:10.1029/2003GL018031.

    • Search Google Scholar
    • Export Citation
  • Christensen, J., and Coauthors, 2007: Regional climate projections. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 847–940.

    • Search Google Scholar
    • Export Citation
  • Colucci, S. J., and J. C. Davenport, 1987: Rapid surface anticyclogenesis: Synoptic climatology and attendant large-scale circulation changes. Mon. Wea. Rev., 115, 822836.

    • Search Google Scholar
    • Export Citation
  • Comiso, J. C., 2003: Warming trends in the Arctic from clear sky satellite observations. J. Climate, 16, 34983510.

  • Danard, M., A. Munro, and T. Murty, 2003: Storm surge hazard in Canada. Nat. Hazards, 28, 407431.

  • Drobot, S. D., and J. A. Maslanik, 2003: Interannual variability in summer Beaufort Sea ice conditions: Relationship to winter and summer surface and atmospheric variability. J. Geophys. Res., 108, 3233, doi:10.1029/2002JC001537.

    • Search Google Scholar
    • Export Citation
  • Hudak, D., and J. Young, 2002: Storm climatology of the southern Beaufort Sea. Atmos.–Ocean, 40, 145158.

  • Jones, B. M., C. D. Arp, M. T. Jorgenson, K. M. Hinkel, J. A. Schmutz, and P. L. Flint, 2009: Increase in the rate and uniformity of coastline erosion in Arctic Alaska. Geophys. Res. Lett., 36, L03503, doi:10.1029/2008GL036205.

    • Search Google Scholar
    • Export Citation
  • Jorgenson, M. T., Y. L. Shur, and E. R. Pullman, 2006: Abrupt increase in permafrost degradation in Arctic Alaska. Geophys. Res. Lett., 33, L02503, doi:10.1029/2005GL024960.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471.

  • Keegan, T., 1958: Arctic synoptic activity in winter. J. Meteor., 15, 513521.

  • Lynch, A., J. A. Curry, R. D. Brunner, and J. A. Maslanik, 2004: Toward an integrated assessment of the impacts of extreme wind events on Barrow, Alaska. Bull. Amer. Meteor. Soc., 85, 209221.

    • Search Google Scholar
    • Export Citation
  • Manson, G., and S. Solomon, 2007: Past and future forcing of Beaufort Sea coastal change. Atmos.–Ocean, 45, 107122.

  • McCabe, G. J., M. P. Clark, and M. C. Serreze, 2001: Trends in Northern Hemisphere surface cyclone frequency and intensity. J. Climate, 14, 27632768.

    • Search Google Scholar
    • Export Citation
  • Mesinger, F., and Coauthors, 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343360.

  • Screen, J. A., and I. Simmonds, 2010a: The central role of diminishing sea ice in recent Arctic temperature amplification. Nature, 464, 13341337.

    • Search Google Scholar
    • Export Citation
  • Screen, J. A., and I. Simmonds, 2010b: Increasing fall–winter energy loss from the Arctic Ocean and its role in Arctic temperature amplification. Geophys. Res. Lett., 37, L16707, doi:10.1029/2010GL044136.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., 1995: Climatological aspects of cyclone development and decay in the Arctic. Atmos.–Ocean, 33, 123.

  • Serreze, M. C., and A. P. Barrett, 2008: The summer cyclone maximum over the central Arctic Ocean. J. Climate, 21, 10481065.

  • Serreze, M. C., J. E. Box, R. G. Barry, and J. E. Walsh, 1993: Characteristics of Arctic synoptic activity, 1952–1989. Meteor. Atmos. Phys., 51, 147164.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., and Coauthors, 2003: A record minimum Arctic sea ice extent and area in 2002. Geophys. Res. Lett., 30, 1110, doi:10.1029/2002GL016406.

    • Search Google Scholar
    • Export Citation
  • Serreze, M. C., M. M. Holland, and J. Stroeve, 2007: Perspectives on the Arctic’s shrinking sea-ice cover. Science, 315, 15331536.

  • Simmonds, I., and K. Keay, 2009: Extraordinary September Arctic sea ice reductions and their relationships with storm behavior over 1979–2008. Geophys. Res. Lett., 36, L19715, doi:10.1029/2009GL039810.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., and E.-P. Lim, 2009: Biases in the calculation of Southern Hemisphere mean baroclinic eddy growth rate. Geophys. Res. Lett., 36, L01707, doi:10.1029/2008GL036320.

    • Search Google Scholar
    • Export Citation
  • Simmonds, I., C. Burke, and K. Keay, 2008: Arctic climate change as manifest in cyclone behavior. J. Climate, 21, 57775796.

  • Solomon, S. M., 2005: Spatial and temporal variability of shoreline change in the Beaufort–Mackenzie region, Northwest Territories, Canada. Geo-Mar. Lett., 25, 127137.

    • Search Google Scholar
    • Export Citation
  • Stroeve, J. C., M. M. Holland, W. Meier, T. Scambos, and M. Serreze, 2007: Arctic sea ice decline: Faster than forecast. Geophys. Res. Lett., 34, L09501, doi:10.1029/2007GL029703.

    • Search Google Scholar
    • Export Citation
  • Whiteman, C. D., and J. C. Doran, 1993: The relationship between overlying synoptic-scale flows and winds within a valley. J. Appl. Meteor., 32, 16691682.

    • Search Google Scholar
    • Export Citation
  • Zhang, X., J. E. Walsh, J. Zhang, U. Bhatt, and M. Ikeda, 2004: Climatology and interannual variability of Arctic cyclone activity, 1948–2002. J. Climate, 17, 23002317.

    • Search Google Scholar
    • Export Citation
Save