After a dark Nordic winter, the Meteosat Natural Colour RGB provides a view of day time ice through the combination of 0.8 µm, 1.6 µm and 3.9 µm — optimal to separate land, ice or cloud.
By Vesa Nietosvaara and Jose Prieto (EUMETSAT)
At the end of March, the Gulf of Bothnia was frozen at its surface, with a thickness of around 40 to 50 cm (Figure 1, source: FMI).
Over the lake Meteosat's sensors in the solar domain measured ice surface reflectivities higher than for liquid water.
The highest ice-liquid difference occurred for 0.8 µm, which explains the reddish hues at the most northern part of the Baltic Sea on Figure 2.
The sea between continental Finland and the Aland island was also frozen and appeared red.
Between 27–31 March the ice started to melt (eventually turning to liquid) in large areas of the east side of the Gulf of Finland — the same date as the colour composite showing no red colouration apart from the easternmost part of the Gulf of Finland, near Saint Petersburg.
The analysis does not match the satellite picture. It could be that wet ice is similar to liquid water in its reflectivity, so it is also dark in the satellite composite.
The main advantage of the composite based on 0.8µm, 1.6µm and 3.9µm (Snow RGB) is the clear separation between snow on the ground and ice cloud, both looking similar in hue in the solar channels composite.
In the Snow RGB the second albedo at 1.6 µm is higher for the small crystals in the cloud as for the rough snow on the ground, creating the hue difference: red for snow (reflectivity at 1.6 µm ca. 15%), peach for ice crystals (ca. 40%). See the areas A (snow) and B (ice cloud) in the slider, Figure 3.
The MODIS reflectivity product from the Terra satellite on 30 March (Figure 4) is superior in resolution to Meteosat. At these latitudes, around 20 times better, or 400 times more pixels for a particular area. See also Aqua MODIS image from 10:50 UTC.
Baltic Sea ice thickness information (Baltice.org)
Previous case studies
Frozen Lake Baikal (19 March 2016)