Homogenisation of a gridded snow water equivalent climatology for Alpine terrain: methodology and applications

摘要

Gridded snow water equivalent (SWE) data sets are valuable for estimating thesnow water resources and verify different model systems, e.g. hydrological,land surface or atmospheric models. However, changing data availabilityrepresents a considerable challenge when trying to derive consistent timeseries for SWE products. In an attempt to improve the product consistency, wefirst evaluated the differences between two climatologies of SWE grids thatwere calculated on the basis of data from 110 and 203 stations, respectively.The "shorter" climatology (2001–2009) was produced using 203 stations(map203) and the "longer" one (1971–2009) 110 stations(map110). Relative to map203, map110underestimated SWE, especially at higher elevations and at the end of thewinter season. We tested the potential of quantile mapping to compensate formapping errors in map110 relative to map203. During a9 yr calibration period from 2001 to 2009, for which both map203and map110 were available, the method could successfully refine thespatial and temporal SWE representation in map110 by makingseasonal, regional and altitude-related distinctions. Expanding thecalibration to the full 39 yr showed that the general underestimation ofmap110 with respect to map203 could be removed for thewhole winter. The calibrated SWE maps fitted the reference (map203)well when averaged over regions and time periods, where the mean error isapproximately zero. However, deviations between the calibrated maps andmap203 were observed at single grid cells and years. When we lookedat three different regions in more detail, we found that the calibration hadthe largest effect in the region with the highest proportion of catchmentareas above 2000 m a.s.l. and that the general underestimation ofmap110 compared to map203 could be removed for the entiresnow season. The added value of the calibrated SWE climatology is illustratedwith practical examples: the verification of a hydrological model, theestimation of snow resource anomalies and the predictability of runoffthrough SWE.