Mesoscale Aspects of Winter Weather Forecasting:
Polar Lows - Ungava Bay, Northern Canada 01-03 December 2000

Overview

A polar low formed over the open waters of Ungava Bay early on 02 December 2000 with winds gusting up to 40 kts at some coastal observing stations. It took only a few hours for the low to form and it lasted around 20 hours. As the low slowly drifted eastward, it decayed into an area of convection over the Torngat Mountains and the Labrador Sea.

The polar low formed as a consequence of the superimposition of upper-level conditions with favourable surface conditions. Very cold air at upper and low levels, together with an area of positive vorticity advection, interacted with a low-level baroclinic zone. The baroclinic zone set up as a result of the ice/water boundary near the western shore of Ungava Bay. The open water provided large amounts of latent and sensible heating to the boundary layer, resulting in an unstable thermal profile through a deep layer of the atmosphere. The resulting polar low moved across the Bay and eventually dissipated as surface and upper-level conditions became less favourable.

Note: This Summary provides an overview of the formation, evolution, and decay of the polar low. The Case Challenge provides a more in-depth analysis. Additional information on polar lows is referenced in the Supporting Topics.

Upper-level Precursors

One precursor in many polar low events is an upper-level trough containing very cold air. A rule of thumb is that the temperature in the trough at 500 hPa must be close to or below -40°C (see 1.3.1 Pre-development Conditions and 5.1.2 Identifying Pre-development Conditions). A triggering mechanism that provides some upward vertical motion, such as an area of PVA, is required to set things in motion (see 5.1.6 Identifying Trigger Mechanisms).

At 0000 UTC on 01 December the analysis indicates a 500-hPa trough lying over Foxe Basin and northern Hudson Bay. A well-defined vorticity center lies just south of Southampton Island, with an associated region of PVA to its east and south. Temperatures in the trough are very cold, near -40°C.

Upper-level Precursors (cont.)

The area of vorticity advection is well-correlated to a band of clouds composed of cirrus or thin mid-level clouds, as shown on the NOAA-14 IR imagery from 2050 UTC on 30 November.

no place names    view place names

Lower-level Precursors

Polar lows often form due to the interaction of a suitable upper-level pattern with favourable low-level and surface conditions. Typically, these low-level conditions consist of outbreaks of cold air across ice-water boundaries and their associated low-level baroclinic zones (see 1.1.1 Sensible and Latent Heat and 1.1.2 Baroclinic Instability).

At 0000 UTC on 01 December, the surface chart shows a trough of low pressure over Foxe Basin and northern Hudson Bay. Observed temperatures are at or below -25°C west of the trough. To the east, over the Quebec coast of Hudson Bay, temperatures are higher. These conditions point to a cold-air outbreak behind the trough.

Lower-level Precursors (cont.)

The ice analysis chart for this time period shows Ungava Bay to be nearly ice free, though there is a narrow band of 9/10ths ice cover along its western and southern shores. The ice-water boundary creates a local, shallow baroclinic zone. This is enhanced at 925 hPa and 850 hPa by similarly-oriented baroclinic zones created by the cold low-level air flowing eastward over the Bay.

Triggers and Development: Upper-level Circulation

Provided that a trigger mechanism is available, polar lows typically form under a cyclonic upper-level circulation (see 1.1.4 The Cold Upper Trough and/or Cold Low).

In this case, the upper-level and surface troughs are moving eastward together and arrive near the western shore of Ungava Bay shortly before 0000 UTC on 02 December. At that time, the upper pattern is still favourable as a polar low precursor and the area of positive vorticity advection remains well-defined.

Triggers and Development: Verification

This satellite imagery shows the cloud pattern associated with the upper trough and vorticity advection as it moves eastward over Ungava Bay. Surface observations indicate the passage of the low-level trough between 1800 and 0000 UTC on 01-02 December with the associated westerly winds and cold advection to the west of the trough. From the NOAA-12 IR 2211 UTC satellite imagery, it is clear that cloud streamers are forming over the open water of western Ungava Bay in the cold westerly flow following the passage of the trough.

Triggers and Development: Upper- and Lower-level Interaction

The interaction of certain upper- and lower-level patterns is thought to be important for polar low development and can be examined in a number of ways. One method is the potential vorticity (PV) approach. PV can be used to schematically illustrate how upper-level baroclinicity can reach down in the atmosphere to link up with low-level baroclinicity. In the case of a polar low, the diabatic heating of the cold atmosphere by open water plays a key role.

Over southwestern Ungava Bay, the GEM regional 6-hour forecast from 0000 UTC on 02 December defines fluxes of up to 600 watts/m² leading to diabatic heating in the lower layers of the atmosphere.

Triggers and Development: Upper- and Lower-level Interaction (cont.)

This cross section over Ungava Bay shows PV for the same 6-hour forecast. The 1.2 PVU contour extends downward to around 800 hPa, while another small 1.2 PVU contour is found from about 1000 to 900 hPa over Ungava Bay. This low-level PV max is likely related to the diabatic heating in that area. The 1.2 PVU contour symbolically connects the upper and lower PV maxima. The interaction between upper and lower PV maxima is one way to consider the relationship between upper and lower levels in the polar low development process.

Appearance in Satellite Imagery

Polar lows form in high-latitude areas over oceans or other bodies of water. These are generally data sparse areas, making satellite imagery the primary method of detecting polar lows. Images from polar orbiters are commonly used, while those from geostationary satellites can be used if the low is not too far north. (see 3.1.4 Satellite Imagery Characteristics and 3.2 Use of Satellite Imagery)

By 0515 UTC, the GOES IR imagery shows the old comma cloud rapidly spinning up and beginning to form a more distinct spiral-shape over Ungava Bay. At the height of its development, around 0900 UTC, the polar low has a classic shape with spiral arms whose individual elements have well-defined edges typical of convective cloud.

Surface and Upper-air Observations

Due to their remote location, the forecaster will be lucky to get a ship, buoy, or coastal land station observation close to a polar low. This calls for careful monitoring of all available surface observations for subtle signs involving pressure or wind speed and direction. (See 3.1 Nowcasting )

The GEM Regional surface analysis for 1200 UTC on 02 December does not resolve the polar low. Its MSL pressure is about 1008 hPa at WKW, and 1007 at WRH. The automated analysis shows only an open trough between those two stations, with pressures no lower than 1005 hPa in the trough. In reality, pressure at WKW is 1005 hPa, 3 hPa deeper than the initialization. The observed pressure at WRH is 1008 hPa. The gradient between the stations is tighter than shown by the initialization. In addition, observed winds at WRH are westerly, indicating a closed low rather than the open trough of the initialization.

Click the following link to view a more complete set of station observations:
• Station Map and Observations

Surface and Upper-air Observations (cont.)

It is very unusual to have a radiosonde observation in the vicinity of a polar low. However, in this case, the station YVP, on the south shore of Ungava Bay, did provide soundings representative of the conditions over Ungava Bay during the development of the polar low. The sounding from this site at 1200 UTC on 02 December illustrates the vertical structure of the very cold air mass in the area. The temperatures between 600 and 500 hPa are near -40ºC, and the lowest level air is at around -18ºC. This air mass, when placed over the open water of Ungava Bay, will exhibit a great deal of instability through a deep atmospheric layer.

Click the following link to view a more complete set of station observations:
• Station Map and Observations

Model Initialization

It is important to realize the limitations of the NWP guidance. In December 2000, the horizontal resolution of the GEM regional model was 24 km. Since at least 8 grid points are required to define a feature in an NWP model if it is to be well-forecast, wavelike structures with a scale of around 200 km, the approximate size of Ungava Bay, are at the limit of the model's ability. (See A 20 km Grid Accurately Depicts 40 km Features in the Ten Common NWP Misconceptions series for more discussion of this point).

NWP models depend on accurate initial analyses, but the high-latitude areas in which polar lows form are generally data-sparse areas in terms of traditional observational data. This means that model trial fields (first-guess fields) can be inadequate in these areas. Satellite data have barely started to fill the data void. For these reasons, the initial analysis of a NWP model can be incorrect on the details of position and depth of a small-scale high-latitude feature such as a polar low.

Another potential analysis problem is with the ice coverage used by the model. Ice analyses may be done on relatively coarse grids, so features such as ice-water boundaries can be "smeared out," even if there are adequate ice observations. During the polar low event, there is a discrepancy between the amount of ice coverage used by the GEM Regional model versus the coverage as interpreted by the Canadian Ice Service. Since the latter is based more closely on actual observations, it is considered a more accurate representation. Compare the GEM Regional analysis for the following 00-hour forecasts to the Canadian Ice Service product.

Dissipation

Polar lows will quickly lose their strength and decay as they cross over land or ice surfaces. (See1.3.5 Dissipation of Polar Lows and 5.1.9 Forecasting the Dissipation of Polar Lows)

As noted earlier, the Ungava Bay polar low dissipated as it crossed over the Torngat Mountains. By 1800 UTC, as the polar low moves eastward with the 500 hPa trough, it loses its shape and definition as it makes landfall on the east side of Ungava Bay. It then dissipates over the land of the Torngat mountains. The upper support also moved across the Torngats to the Labrador Sea, where a rather disorganized cyclonic surface pattern formed. The polar low did not reestablish itself over the Labrador Sea. Rather, the cyclonic pattern is best considered as a new and weaker system to the lee of the Torngats.

^top