Wave clouds along the south side of the Alps

Wave clouds along the south side of the Alps

7 April 2017 00:00 UTC–23:45 UTC

Wave clouds along the south side of the Alps
Wave clouds along the south side of the Alps

Orographic or wave clouds occurred in the lee of the Alps on 7 April 2017, under northerly winds.

Last Updated

11 November 2022

Published on

07 April 2017

By HansPeter Roesli (Switzerland) and Jochen Kerkmann and Jose Prieto (EUMETSAT)

Operational numerical models still have problems forecasting orographic or wave clouds in the lee of mountain chains, as was the case on 7 April.

In the middle-upper troposphere moderately humid air was advected towards the Alps during most of the day, as shown on the animation of the Meteosat-10 Water Vapour (WV6.2 channel) imagery, 7 April 00:00–23:45 UTC. The advection only weakened towards midnight.

The Meteosat-10 High Resolution Visible (HRV) imagery, 7 April 06:00–16:00 UTC shows that an important wave cloud covered the central and eastern parts of the Po Valley during daylight, marring the temperature and sunshine forecasts for the area.

The photograph taken in an easterly direction from Locarno, by MeteoSwiss, (Figure 1) shows the northern edge of the cloud that can be identified as a massive multi-storey cirrostratus lenticularis.

 Photo of the wave cloud over Locarno. Credit: MeteoSwiss
Figure 1: Photo of the wave cloud over Locarno. Credit: HansPeter Roesli.

The imaging radiometer VIIRS on the sun-synchronous satellite Suomi-NPP viewed the area in two adjacent orbits, at 10:55 UTC (Figure 2, left) from an eastern position, and at 12:37 UTC (Figure 2, right), from a central position.

VIIRS Natural Color RGB imagery

Natural Color RGB, 12:37 UTC compare1
compare2
 

Figure 2: Suomi-NPP VIIRS images showing the orographic clouds two hours apart.

Both images are Natural Color RGBs at 375m spatial resolution, where the snowed-in alpine crests and some ice clouds stand out in a cyan colour.

What was surprising was that the wave cloud was white at 12:37 UTC, at least in its western part, hinting to consist of water droplets at its top.

The Milan radiosonde at 12:00 UTC showed a close-to-saturated layer between 420hPa (7.4km) and 200hPa (11.8km) with a temperature of -65 °C, suggesting that the wave cloud had a vertical extension of up to 4.4km and that its top had to be frozen. The cloud-top temperature was verified by the IR11.45 (I5) VIIRS band that also indicated -65 °C (208K), where the cloud was the most dense.

A second guess of the height was obtained from the shadow cast by the northern cloud edge: a reasonable match of ~10km was found.

So why do we see a white cloud top, although it must definitely be frozen? An explanation may given by looking at the temperature difference between the VIIRS bands IR11.45 (I5) and IR3.4 (I4).

Image comparison

Infrared band IR1.61, 12:37 UTC compare1
compare2
 

Figure 3: Comparison Suomi-NPP VIIRS infrared images.

On the IR11.45–IR3.74 difference image (Figure 3, left) the light blue streaks indicate that this part had an extremely high IR3.7 reflectivity, the difference even going beyond -80K.

This is confirmed by Meteosat-10 SEVIRI measurements in the Convection RGB with tropical tuning (Figure 4), which give a very high IR3.9r (solar component) reflectivity of around 15% and a NIR1.6 reflectivity of around 60% for the high-level wave cloud south of Switzerland (see cursor position in the image).

 Meteosat-10 Convection RGB with tropical tuning, 7 April 12:30 UTC
Figure 4: Meteosat-10 Convection RGB with tropical tuning, 7 April 12:30 UTC

It could be that the ice crystals were so small that they also reflected well in the IR1.61 (I3) band (Figure 3, right), thus whitening the RGB. In any case, further east where the difference is smaller at 12:37 UTC some 'correct' cyan colouring is evident.

High reflectivity at 1.6µm is unusual for ice particles. For reference, inside a typical frontal cloud, the differences between 0.8µm and 1.6µm for the ice cloud can reach 80 percent points because of 1.6µm ice particle reflectivity around 20%. In this case, however, the differences are much smaller, as shown on Figure 5, right panel.

If the crystals are less than 0.5µm in diameter, there cannot be much absorption at 1.6µm. Due to strong winds and low humidity, coalescence is restricted, and that explains why the crystal sizes are very small.

 Left: Meteosat-10 Natural Colour Composite, 7 April 14:00 UTC. Right: Diagram with readings for the yellow-shaded pixels on the left image. For them, the cloud is not opaque (0.6 µm reflectivities in the range 55–80%). Moderate differences between 0.8 µm and 1.6 µm occur.
Figure 5: Left: Meteosat-10 Natural Colour Composite, 7 April 14:00 UTC. Right: Diagram with readings for the yellow-shaded pixels on the left image. For them, the cloud is not opaque (0.6 µm reflectivities in the range 55–80%). Moderate differences between 0.8µm and 1.6µm occur.