Détermination objective des caractéristiques des grains de neige à partir d'images

Objective determination of snow grains characteristics from images

Centre d'Etudes de la Neige

Abstract

The determination of snow grains characteristics (such as type or size) are crucial to validate snow models and interpret data from remote sensing. Until now, the size and type of snow grains are estimated in the fields. Nowadays, the transport of snow sample in iso-octane allows laboratory analyses. An objective analysis of snow grains digital images has been developed at the Centre d'Etudes de la Neige since the mid 80's. The images are taken in a cold room and stored on a videodisk. The calculation of relevant parameters for snow characterisation was developed. Recently, the hardware and software systems were upgraded, allowing quicker and easier analyses. Finally, a procedure designed to determine the snow types from images analysis was developed and calibrated with 65 snow samples. For that purposes, independent experts determined the snow type (as defined by the International Classification) of each sample. The procedure was verified on an independent set of snow images.

This semi-automated determination gives 97% of accurate results on granular snow types. However, it is not reliable on precipitation particles.