Freezing Drizzle

Image 1 of 3

The classical formation process (shown on the left) is very similar to that of freezing rain. In this case, small ice crystals will fall through a warm layer in the lower atmosphere and melt, before continuing to fall into a layer of below freezing air closer to the surface. The drizzle droplets will not have enough time to refreeze in this layer, but will instead freeze on contact with the ground or other surfaces. This formation model is the less common process, and is most often associated with approaching warm fronts. It is short lived and is oftentimes mixed with snow, before transitioning to rain or drizzle.

Important notes:

Freezing drizzle (FZDZ) is similar to freezing rain, but differs in terms of droplet size and the formation processes.
Drizzle droplets are much smaller than rain with a typical diameter of 0.1mm versus 1-4mm for rain droplets.
Drizzle develops in very moist and generally stable stratus cloud, with cloud top temperatures warmer than -10°C. These conditions must be satisfied; otherwise ice crystal formation would occur and lead to snow or rain/freezing rain.
There are mechanisms by which FZDZ develops: classical and non-classical, and examples of the associated atmospheric profiles are available in this figure and the one shown next on the carousel. Although counter-intuitive, the non-classical FZDZ mechanism is the most commonly observed.
The requirement of stratus for the development of FZDZ means that it is typically accompanied by IFR conditions.
Snow grains (SG) develop in the same process as FZDZ, but in generally colder temperatures. SG starts as freezing drizzle in the cloud, but since the cloud is cold enough, it will freeze before reaching the surface. This can be thought of as “frozen freezing drizzle”. It typically is not associated with icing conditions at the surface, but does imply the risk of moderate icing conditions within the cloud.

Image Source: Environment and Climate Change Canada

The non-classical formation process is the dominant process by which FZDZ develops. It is characterized by stratus cloud whose temperature is entirely below freezing. Small supercooled liquid water droplets (water droplets under 0°C) will develop and bounce around and coalesce into larger drizzle droplets. Once the drizzle droplets get large enough, they will begin to fall through the cloud and freeze at the surface, or upon encountering an object, such as an airplane.

Image Source: Environment and Climate Change Canada

The climatology of FZDZ in the image on the left shows median annual precipitation days for freezing drizzle between 1976 and 1990. Factors such as onshore and upslope flow promote the formation of FZDZ, which add moisture and lift that helps drizzle droplet development. Consequently, Atlantic Canada (particularly Newfoundland), the Arctic islands, and other parts of the country near large bodies of water (particularly near Hudson and James Bay) are favourable for the development of FZDZ (shown in this last image). Freezing drizzle can last for a relatively long time if there is no change in air mass or wind direction.

Image Source: Cortinas Jr. et al. (2004)

FZDZ is difficult to forecast, as it requires very specific conditions to develop. The main challenge is determining the depth of the stratus. The stratus needs to be deep enough to have enough moisture for droplets to coalesce into drizzle droplet size, and begin to fall. However, if the stratus is too thick, and cloud top temperatures too cold, you will tend to favour the development of ice crystals and snow, with light snow observed at the surface. The Goldilocks zone of stratus depth can be difficult to attain. This is why the risk of FZDZ in a TAF is usually represented as a PROB until it is observed. Freezing drizzle is a phenomenon that is poorly forecast by numerical weather forecast systems.

Another forecasting challenge is precipitation typing, and whether the freezing drizzle will freeze or not. Snow grains (SG) develop in the same process as FZDZ, but in generally colder temperatures. SG starts as freezing drizzle in the cloud, but since the cloud is cold enough, it will freeze before reaching the surface. This can be thought of as “frozen freezing drizzle”. It typically is not associated with icing conditions at the surface, but does imply the risk of moderate icing conditions within the cloud.

Image 1 of 3

These GFA panels from December 13, 2022 valid at 1200Z indicate a large swath of freezing drizzle across southeastern Saskatchewan and west-central Manitoba. Within the orange dashed line on the clouds and weather panel (left), conditions are ideal for freezing drizzle formation: low ceilings around 200-600ft AGL and tops between 5000-7000ft. In the clouds and icing panel (right), patchy to moderate clear icing is indicated from the surface to 2000ft within the area expecting freezing drizzle. Local freezing drizzle is also mentioned east of the stationary trough, however cloud bases in those areas are forecast to be higher which makes it harder for freezing drizzle to reach the surface.

The GFA panels valid at 0600Z on December 14^thshow that the organized and extensive area of freezing drizzle has dissipated, though patchy freezing drizzle is still forecast across much of the Prairies where cloud tops are still low and IFR conditions remain forecast at the surface. Over southern sections of Manitoba and far southeastern Saskatchewan, deeper more organized precipitation is forecast to move in associated with the trowal tied to the low centered over Nebraska. As the cloud tops associated with this organized precipitation are far higher than the stratus required to maintain drizzle formation, the freezing drizzle event is forecast to end as the trowal moves into the area.

The TAF for Portage la Prairie, MB, located approximately 70 km west of Winnipeg, MB, indicates a long-lived freezing drizzle event. The TAF issued at 1308Z on December 13 is the first one issued for the day at that airport. It shows southeasterly winds, freezing drizzle mixed with light snow, and overcast ceilings at 500ft. The forecast indicates that the freezing drizzle will taper off after 2000Z, with only light snow remaining.

Image 1 of 3

These images show the modeled atmospheric profiles for Portage La Prairie, MB (CYPG) between 1200Z December 13 and 0600Z on December 14 in 6-hour time steps. Click through the carousel of images to follow the timesteps discussed in this section. The 1200Z and 1800Z profiles on December 13 show that the lowest levels of the atmosphere are saturated with low cloud ceilings and cloud tops around 5000ft. These first two profiles capture the perfect environment needed for the development of freezing drizzle.

Image Source: Pivotal Weather

These images show the modeled atmospheric profiles for Portage La Prairie, MB (CYPG) between 1200Z December 13 and 0600Z on December 14 in 6-hour time steps. Click through the carousel of images to follow the timesteps discussed in this section. The 1200Z and 1800Z profiles on December 13 show that the lowest levels of the atmosphere are saturated with low cloud ceilings and cloud tops around 5000ft. These first two profiles capture the perfect environment needed for the development of freezing drizzle.

Image Source: Pivotal Weather

The 0000Z profile from December 14 shows the dewpoint and temperature coming together in the mid-levels, which would indicate deeper saturation, thicker clouds and higher cloud tops. These thicker clouds (which correspond to the approaching trowal seen on the GFA panels) and higher cloud tops bring about a shift from an environment that supports freezing drizzle to an environment that supports the growth of larger water droplets, ice crystals and eventually the development of snowflakes.

Image Source: Pivotal Weather

Numerical weather models have difficulty in detecting freezing drizzle, mainly because low-level moisture and clouds are poorly handled by weather models and this is usually where freezing drizzle develops. With “precipitation type” charts, freezing drizzle is quite often missed altogether or extremely over-forecast. With “precipitation rate” charts, quite often freezing drizzle is not detected at all as it does not yield much in terms of accumulated precipitation. In the animation to the left, hourly forecasted precipitation rates are shown from 1200Z on December 13 to 1200Z on December 14. The lobe of precipitation associated with the surface trough drawn on the 1200Z GFA panels from December 13 and the encroaching area of more organized precipitation associated with the low over the northern United States is shown. There is no freezing drizzle shown in this animation, likely because the accumulations were too small to be detected by the weather model.

This example shows why additional products and weather briefings are necessary for the proper diagnosis of the potential for freezing drizzle to occur.

Image Source: College of DuPage

This satellite imagery was taken on December 13^th between 1550Z-2020Z. The low-level clouds blanketing Saskatchewan and Manitoba producing freezing drizzle (to the north west of the image, smooth and unmoving) can be differentiated from the approaching higher-level clouds associated with the low pressure system and trowal moving in from the northern United States(to the southeast of the image, bumpy, moving). The difference in colour between these two areas indicates that most of this low-level cloud is composed of water, while the thicker, higher-level cloud is a mix of water and ice. It is these low-level clouds that support the formation and maintenance of freezing drizzle in Saskatchewan and Manitoba. Though exact cloud tops cannot be discerned from this particular satellite imagery, there is a clear delineation shown between the cloud supporting freezing drizzle and the cloud supporting snow. As shown in the GFA panels, the tops are forecasted to be 5000-7000 ft.

Image Source: CIRA

Radars can struggle with capturing the occurrence of freezing drizzle, especially when ceilings are often so low where the freezing drizzle is occurring that it is outside the range of the radar beam. The radar loop shown here, valid on December 13^th from 1900Z-2230Z shows little to no precipitation echoes in the Portage La Prairie (CYPG) area when in fact METARs indicate that freezing drizzle and light snow were occurring through that same timeframe. The more organized precipitation associated with the trowal and encroaching low is more readily captured by the echoes moving into southern Manitoba south of CYPG late in the event.

Image Source: RADARSCOPE

According to the surface observations, Portage La Prairie (CYPG) saw a freezing drizzle event that lasted at least 8 hours (as observations prior to 1200Z are not available for this site, the start time of this event is unknown). Ceilings throughout the event remain solidly IFR, ranging from 300-500ft. Notice that there are METARs issued intermittently that show higher ceilings (1724Z, 2000Z, after 2049Z) and are either associated with no precipitation or only light snow. This highlights the specific environment needed to support the development of freezing drizzle, where a change in the height of the cloud ceiling altered the precipitation height observed at the airport. After 2049Z ceilings lifted considerably, leading to the dissipation of freezing drizzle entirely.

Image Source: OGIMET

Legend

Legend

HAZARDS

About

Definition

Associated Terms

Freezing Drizzle

Visualization

MAIN CONCERNS

Service Providers

Users