The main aim of this RGB type is to discriminate snow covered land from fog or low water clouds. Although several RGB types are usable for snow detection and for discriminating between snow and low clouds, it is the Snow RGB which provides the highest colour contrast between snow and water clouds (red-orange against white). This RGB type was tuned especially for this purpose.

Fog frequently forms over cloud-free snow surfaces in wintertime high pressures. Identifying foggy areas is important, for example because of traffic security. The Snow RGB is a daytime RGB.

The Snow RGB is composed of the reflectivity images calculated from the VIS0.8, NIR1.6 and the 3.9 micrometer channel (IR3.9).

Physical basis

Daytime snow detection

Snow RGB consists of three images: the snow is bright in one and dark in two components, while the water clouds are bright in all three components. This is the key of snow-water cloud discrimination and will ensure good colour contrast between the two.

Fig. 1 shows typical reflectivity spectra of surface types. The snow reflectivity (blue curve) is high around 0.8 micrometer and low around 1.6 and 3.9 micrometers. The reflectivity of water clouds (not shown in Fig. 1) is high in all three channels.

Figure 1: Reflectivity spectra of different surfaces. (Credit: Andreas Kaab, University of Oslo.)

Water and ice cloud discrimination

Discrimination of water and ice clouds is possible because the absorption is stronger for ice than for water both at 1.6 and the 3.9 µm wavelengths (Fig. 2). Due to the stronger absorption of ice the reflectivity of the ice clouds will be (usually) lower than that of the water clouds.

Figure 2: Absorption spectra of water (blue curve) and ice (dashed red curve) and the spectral bandwidth of the SEVIRI channels. (Courtesy of Daniel Rosenfeld, HUJ)