Zur Seitenansicht


Assessing snow depth distribution on the basis of atmospheric reanalysis / by Christian Maier
VerfasserMaier, Christian
Betreuer / BetreuerinMarzeion, Benjamin ; Zeidler, Antonia
ErschienenInnsbruck : 2015, 2015
Umfangv, 79 Seiten : Illustrationen, Diagramme, Karten
HochschulschriftUniversity of Innsbruck, Masterarbeit, 2015
Datum der AbgabeDezember 2015
URNurn:nbn:at:at-ubi:1-3092 Persistent Identifier (URN)
 Das Werk ist frei verfügbar
Assessing snow depth distribution on the basis of atmospheric reanalysis [31.61 mb]
Zusammenfassung (Englisch)

Information on snow and on the physical processes affecting it is of great interest for a large number of environmental issues. The hydrological properties of a basin, with regard to run-out and avalanche characteristics, are in uenced by the amount of snow within a region especially during the melting season. Within this thesis, Central Asia acts as a test site for snow distribution modelling in a data-sparse region. On the basis of meteorological reanalysis and observational data, a semi-empirical model is developed. Daily temperature and precipitation fields are the input parameters of the algorithm. The semi-empirical model distinguishes between solid and liquid precipitation. Each snow fall initiates a new layer in the model, with an initial density that is calibrated using observations. Compaction processes are simulated, based on the relationship between snow-viscosity and settlement rate. In order to mimic the seasonal snow metamorphism a day degree approach describes melting procedures. ERA-Interim reanalysis were evaluated by available daily meteorological data. Within historical databases observational data of 12 stations served as a basis for calibration and evaluation procedures. By simulating seasonal snow packs within the area of interest downscaling procedures give the ability to simulate the snow pack where no observational data is available. Possible high resolution snow distributions depict another application example. Air temperature and precipitation evaluations show root mean square errors (RMSE) of 1.8C and 100 mm/y. Climate Research Unit (CRU) data in combination with in-situ data is used to modify the underlying raw ERA-Interim fields. With this simple optimization method, based on CRU-climatology and ERA-anomalies, improved temperature and precipitation fields are used to drive the semi-empirical model. Results show determination coeffcients between modelled and observed data from 0.4 to 0.8. RMSEs of mean daily snow-depth are within 1 cm for chosen stations. Depending on the time of the year, the mean snow pack density is found between 130 to 550 kg/m3. Model skill is higher in regions where daily snow depth data is available. Results are suitable for climatological research in a mean daily or monthly time-scale. Despite the relative coarse input resolutions of ERA-Interim data the semi-empirical model performs well for snow-climatological issues.