Back to overview

Inferring snow pack ripening and melt out from distributed ground surface temperature measurements

Type of publication Peer-reviewed
Publikationsform Original article (peer-reviewed)
Author Schmid Marc-Olivier, Gubler Stefanie, Fiddes Joel, Gruber Stephan,
Project Extension of Mountain Cryosphere Subgrid Parameterization and Computation (CRYOSUB-E)
Show all

Original article (peer-reviewed)

Journal The Cryosphere
Volume (Issue) 6
Page(s) 1127 - 1139
Title of proceedings The Cryosphere
DOI 10.5194/tc-6-1127-2012


Seasonal snow cover and its melt regime are heterogeneous both in time and space. Describing and modelling this variability is important because it affects diverse phenomena such as runoff, ground temperatures or slope movements. This study presents the derivation of melting characteristics based on spatial clusters of ground surface temperature (GST) measurements. Results are based on data from Switzerland where ground surface temperatures were measured with miniature loggers (iButtons) at 40 locations referred to as footprints. At each footprint, up to ten iButtons have been distributed randomly over an area of 10 m × 10 m, placed a few cm below the ground surface. Footprints span elevations of 2100–3300 m a.s.l. and slope angles of 0–55°, as well as diverse slope expositions and types of surface cover and ground material. Based on two years of temperature data, the basal ripening date and the melt-out date are determined for each iButton, aggregated to the footprint level and further analysed. The melt-out date could be derived for nearly all iButtons; the ripening date could be extracted for only approximately half of them because its detection based on GST requires ground freezing below the snowpack. The variability within a footprint is often considerable and one to three weeks difference between melting or ripening of the points in one footprint is not uncommon. The correlation of mean annual ground surface temperatures, ripening date and melt-out date is moderate, suggesting that these metrics are useful for model evaluation.