Icy or snowy runway. Credit: Andrew Norris

Freezing precipitation at Helsinki-Vantaa airport

10-11 December 2021

Photo credit: Andrew Norris

Icy or snowy runway. Credit: Andrew Norris
Icy or snowy runway. Credit: Andrew Norris

In December 2021 the collision of weather regimes created freezing precipitation conditions over southwestern and southern Finland, that lasted for a few days.

Last Updated

04 May 2023

Published on

26 January 2023

By Robert Mäkitie

On 10 December a strong high pressure area occurred over western Russia, while warmer air and precipitation arrived from the west, along with several fronts. As the fronts started to approach southwestern Finland their movement slowed down due to the high pressure area over western Russia, which prevented the fronts from moving further inland over central Finland. The strong high pressure area also pushed colder surface air from the southeast towards Finland. The collision of these two weather regimes created freezing precipitation conditions over southwestern and southern Finland that would last for a few days.

NSWC chart over Fennoscandia, 12 UTC 12 December (FMI).
Figure 1: NSWC chart over Fennoscandia, 12 UTC 12 December 2021 (FMI).

Overall, three most-used weather models at FMI (including ECMWF HRES global model, FMI HIRLAM and MetCoOp) predicted the whole situation for those three days reasonably well. Weather models tend to perform well during the winter when a strong blocking high occurs and slows down the movement of weather systems.

Two out of three models predicted freezing precipitation over southern Finland, but one of the models predicted the warm air from the west to be a bit cooler, which decreased the likelihood of the freezing precipitation conditions. Instead of freezing precipitation, the specific model predicted more snowfall over southern Finland, which created some minor uncertainty on the coming conditions.

When the first signal of long-lasting freezing precipitation and snowfall was noticed, FMI informed the Helsinki-Vantaa airport about the coming weather. The airport started to closely monitor the weather and prepare for the problems the snowfall and the freezing precipitation might cause. The models predicted a few days in advance that, between 10th and 11 December, 4–10cm of snow and 1.5–3.5mm of freezing precipitation would accumulate at Helsinki-Vantaa airport. The actual amounts on those days were around 1cm of snow and 5–6mm of freezing precipitation.

The freezing precipitation started in the afternoon on 10 December and continued intermittently until the midday the following day. The longest continuous freezing precipitation conditions occurred in the evening  of 10 December, which led to issues at the airport. Freezing rain and drizzle continued non-stop for five hours, which started to delay flights and general operations on the ground. Closer to the midnight Helsinki-Vantaa airport was closed for two and a half hours, which has not happened since 2003. Several flights had to be redirected to other airports in Finland, Sweden and Estonia.

With the freezing precipitation continuing into the following day, there were more delays and cancellations, but the intensity of freezing precipitation was not as strong as the previous day.

The temperature started slowly to rise above zero degrees on Saturday afternoon, turning freezing precipitation into sleet and rain. The freezing precipitation slowly moved towards the east and the surface temperature slowly started to rise, due to the high pressure area in the east slowly weakening and moving further east on 11 and 12 December.

Using satellite imagery to track the event

This case clearly highlights the problems high latitude countries face in the middle of the winter when there is not much sunlight available, and what kind of problems day RGB satellite products face when the sun elevation angle is low (Figures 2-4). It also illustrates the big difference between geostationary satellites images and polar satellite images during the winter period (see also High-latitude simulator for MTG/FCI instrument, in the section Limitations of the method).

Day Microphysical (Metop-A, AVHRR) 11 December 2021 08:16 UTC
Figure 2: Metop-A AVHRR Day Microphysics, 11 December 08:16 UTC
Meteosat-11 SEVIRI 24h Microphysical, 11 December 2021 10:00 UTC
Figure 3: Meteosat-11 SEVIRI 24h Microphysics, 11 December 10:00 UTC
Meteosat-11 SEVIRI Day Microphysics, 11 December 2021 10:00 UTC
Figure 4: Meteosat-11 SEVIRI Day Microphysics, 11 December 10:00 UTC

During winter periods the most used RGB satellite product in the high latitudes is the SEVIRI 24h Microphysics (Figure 3) product because it is not dependent on the sunlight, though the 24h Microphysics RGB does have few small drawbacks. Its resolution is not the sharpest, compared to the HRV products, and the product also over-analyses high thin cirrus clouds, which can give a wrong picture of the real situation. In this case the amount of high clouds was over-emphasised in the 24h Microphysics product, making it impossible to see the clouds below the cirrus clouds. By comparing the SEVIRI images to the polar satellite images, it was possible to see that the high-level clouds that were interpreted in the 24h Microphysics RGB as thick clouds, were mostly just thin cirrus clouds, due to better resolution on polar satellite images.

Image comparison

24h Microphysics compare1
Figure 5: Day Microphysics (NOAA-20, VIIRS), 10 December 2021 11:24 UTC (left), 24h Microphysics (NOAA-20, VIIRS) 10 December 2021 11:24 UTC (right)

Because the 24h Microphysics does provide many details on cloud structure, the best way to get a view of the cloud structure is to use Day Microphysics RGB product. By looking at Day Microphysics images from SEVIRI (Figure 4) a big problem quickly arises. Because the product is largely based on visible channels, the difference between the angle of the sunlight and the imager 'burns' the colour scale, making the Day Microphysics RGB product unusable, even during the day in the middle of the winter. Luckily, the same problem does not occur on the polar satellite images because the imager always looks at the surface from a straight angle making the colour scale normal along the flight path (Figure 5). The same optical phenomena that SEVIRI is affected by can, to some extent, be seen at the edges of the polar satellite images where the colour scale starts to deteroriate due to the measurement angle of the imager.

On 10 December, NOAA-20 did a flyby over Scandinavia at 11:24 UTC, just before the freezing precipitation started at Helsinki-Vantaa (Figure 5). VIIRS data is a good proxy of what we can except from future Metop-SG METimage imagery. From NOAA-20’s Day Microphysics RGB image, it is possible to see that the cirrus clouds over Finland were very thin, and the super-cooled low-level clouds can be observed through small patches in the cirrus cloud cover (green colour), confirming that there was a risk of freezing precipitation. This was the only usable Day Microphysics image that day that FMI forecasters could use, because all the other polar satellites images of Finland were 'burnt' because the sun was just about to rise or set, or Finland is located at the edge of the images. This means that the next usable Day Microphysicsl image was available when NOAA-20 made the same orbital pass over Scandinavia the following day, creating a big time-gap between Day Microphysics images for the polar satellite.

The new generation of satellite images

The new RGB products that will arrive with the new Meteosat Third Generation Flexible Combined Imager (FCI) instrument are unlikely to bring much help to high latitudes in the middle of the winter, because most of the new RGB products are day RGB products, which need  sunlight. This is particularly true of the Cloud Phase RGB, because it's composition analyses the particle size, like the Day Microphysics RGB composition. If the sun shines from a low elevation angle the RGB composition loses information from the particles in the cloud. This can be seen in the proxy data (Figure 8), where it is hard to see the difference between the blue or purple colours.

The True Colour RGB composition might also have some problems from the low angle sunlight, because high and thick cloud will easily create long shadows behind the cloud that might cover some important low clouds.

But, the big question is, how well will the Cloud Type RGB composition work? By looking at the proxy data (Figure 4), it shows a large red area, or so-called thin cirrus cloud layer over southern Finland. The same layer cannot be seen in any other RGB composition. This raises the question, does the new thin cirrus channel detect some thin cirrus that cannot be seen in other compositions or is it the low sunlight angle that creates the wrong illusion of the shadows in this situation.

Cloud Type (FCI simulated from VIIRS) 10 December 2021 11:24 UTC
Figure 6: Cloud Type (FCI simulated from VIIRS) 10 December 11:24 UTC
24h Microphysical (FCI simulated from VIIRS) 10 December 2021 11:24 UTC
Figure 7: 24h Microphysics (FCI simulated from VIIRS) 10 December 11:24 UTC
Cloud Phase (FCI simulated from VIIRS) 10 December 2021 11:24 UTC
Figure 8: Cloud Phase (FCI simulated from VIIRS) 10 December 11:24 UTC

The new RGB compositions that the FCI will bring, will also be available on the new imager (METimage) onboard the new generation of the polar satellites. There might be a chance that the Cloud Phase and the Cloud Type RGB composition will work better onboard the polar satellites, as was seen with the Day Microphysics RGB product earlier on. This means that the new RGB products might work during the winter when there is sunlight.

Of course, the new FCI imager will not only bring four new RGB compositions, but it will also enhance its resolution, including better temporal resolution. This means the Night Microphysics and the 24h Microphysics (Figure 7) will be sharper and have a smaller time gap, making these products better for the winter period.