Analysis of the time series

We aim to analyse the temporal change in snow characteristics between 11 and 19 January 2021, observing 1-2 images per day. Since no new snow fell, we assume that the snow grain size was increasing over time. Does the imagery confirm this assumption?

If the reflectivity of the continuous snow cover was primarily determined by the grain size, we would observe decreasing reflectivity over time in both near-infrared channels, i.e., the snow would appear progressively darker in the mountains in the Cloud Phase RGB. (This will be the case for geosynchronous satellite data due to the fixed viewing geometry.) However, the evaluation of the time series is complicated, since the measured reflectivity values are also highly dependent on the satellite zenith angle, which varies significantly for a given area in VIIRS images taken at different times.

In order to reduce the effect of viewing geometry, we studied time series in pre-defined satellite zenith angle classes: 0°-40°, 40°-50°, 50°-60° and "over 60°". Animation 1 shows the images in the 50°-60° satellite zenith angle interval.

Animation 1: Cloud Phase RGB and its components for 11 January 2021 12:05 UTC, 13 January 2021 12:17 UTC, 16 January 2021 12:11 UTC and 19 January 2021 12:04 UTC.

The reduction of the snow-covered area can be clearly observed in the VIS0.67 and Cloud Phase RGB images.

Does the blue colour really become darker with time in accordance with our assumption? In the image taken on 11 January 2021 at 12:05 UTC, the blue colour of the snow in the area of the Iberian Mountain Chain is bright, as the snow was fresh and continuous in most of the pixels. Later, the reflectivity in the NIR1.61 and NIR2.25 channels decreased, since the snow grain size was likely increasing as the snow was becoming older and/or dirty and/or discontinuous in more and more pixels. Therefore, the snowy area in the Cloud Phase RGB became darker.