Measuring, validating and modeling of snow components in the upper eupharates basin

Precipitation, an important component of the global water balance, occurs in the form of snow in mountainous areas during the winter season. Permanent and seasonal snow covering much of the world show variations from year to year due to; atmospheric movements, geophysical changes, global warming and human impact. In the mountainous eastern part of Turkey, Europe's 4th highest country, snow remains on the ground more than half of the year. This snow dominated region feeds the Euphrates-Tigris Basins, that comprises many large dams of the country, and may also be called as "water tower of the Middle East". Therefore, the spatial and temporal change of the snow cover needs to be closely monitored and modeled. The focus of this study is the measurement of snow components on land, as well as monitoring the snow cover with remote sensing and modeling the snowpack to include its interaction on the soil-snow-atmosphere interface in the region that encompasses the Upper Euphrates Basin. The recent snow seasons of 2015-2018, when the snow observation network in Turkey has increasingly developed with the initiatives of water-related institutions, are selected as working periods. Snow component measurements are performed during the snow season using different measurement methods (automatic and discrete) at different elevation levels. Within this scope, snow pit analyses are carried out at suitable regions selected in the mountainous eastern Turkey. Within the same area, 33 snow observation station (KGI) data, ranging in altitude from 1250 m to 2400 m, are compiled and 4002 observation pairs (HS- ρ_snow- SWE) are analyzed for an average of 36 years. The long-term data from the region indicates on average a snow pack depth of 53 cm, snow density of 281 kg/m3 and a water potential of 16 cm. In addition, 11 Snow Pack Analyzer (SPA) snow observation station equipment established in Turkey before the 2015 snow season have been introduced, data analyzed and evaluated with satellite imagery. In general, SPA stations measured varying snow components according to region during the 2015-2018 winter seasons and their Snow Water Equivalent (SWE) performance with satellite imagery change from 30 to 480 mm in terms of RMSE. Finally, a fully distributed SNOW-17 snow model with 5 km x 5 km spatial resolution is applied as a modeling study. Snow model calibration and validation results show performances as 0.81 and 0.66 respectively, in terms of Nash-Sutcliffe Efficiency (NSE) compared with ground observations. In addition, the findings of the fully distributed model are evaluated with multiple satellite products (IMS, MODIS, SEVIRI and SSMI/S) that provide information about Snow Covered Area (SCA) and Snow Water Equivalent (SWE). Accordingly, successes are achieved between 0.81-0.95 NSE in terms of SCA and 22-65 mm RMSE in terms of SWE for the 2015-2018 periods. The region has different snow extent compared to meteorological conditions varying from year to year, and the model results are found to be very consistent both temporally and quantitatively with four satellite products. The thesis including many pioneer applications (snow pit analyses, multi-satellite snow monitoring and fully distributed snow modeling), is expected to pave new directions in operational snow hydrology and water resources management in Turkey.