Snowfall Regime Evaluation for MetCoOp EPS at a Norwegian Mountain Site

Previous studies showed the importance to have information about the vertical distribution of precipitation to simulate snow and related cyclone development correctly in regional,mesoscale models. Understanding snowfall is important since snowfall has effects on human society, wildlife, and local and global climate. During Christmas 2016 an extreme weather event had a large impact on Eastern, Southern, and Western Norway. The WMO measurement site Haukeliseter (991 m a.s.l), in Southern Norway, is equipped with a double fence snow gauge. During winter 2016-2017 three additional instruments for the estimation of snow water content (SWC) were installed for the US National Science Foundation-funded field campaign, High-Latitude Measurement of Snowfall (HiLaMS). Snow measurements are compared with the ensemble model forecast of the Norwegian Meteorological Institute for Haukeliseter. We investigated if the Meteorological Cooperation on Operational Numerical Weather Prediction (MetCoOp) Ensemble Prediction System (MEPS) can forecast the surface snowfall and variation in the vertical for an extreme winter event as well as for winter 2016-2017. The local topography around Haukeliseter affects the surface snow accumulation and the vertical snowfall pattern. Continuous snow patterns are observed during weak south-easterlies. Strong westerlies show higher amounts of SWC with an alternating precipitation pattern within one hour. The 24 h surface precipitation amount is predicted too high by MEPS during the entire event (+33 %). The vertical snow water content shows lower values for MEPS than for the observations. During winter 2016-2017, simulated surface accumulation in MEPS is greater (+44 %) and related to high wind speeds during westerlies at Haukeliseter. An adjustment to the microphysical scheme in MEPS shows a discrepancy of +56 % for winter 2016-2017. The predicted SWC of the adjusted microphysics improved compared to the control run, with higher, but still too low values of MEPS.