Sub-kilometer numerical forecasting of atmospheric boundary layer in an Arctic fjord-valley system during polar night

Terrain-induced atmospheric features pose challenges for prediction of near-surface atmospheric conditions even in kilometer-scale numerical weather prediction (NWP) models. In this study, the ALADIN-HIRLAM NWP system was applied with 2.5 km and 0.5 km horizontal grid spacing in an Arctic fjord-valley system. The impact of the decreased grid spacing on the ability to represent boundary-layer processes was investigated for the period from 12 to 16 February 2018 during the Year of Polar Prediction Special Observing Period Northern Hemisphere 1. The model simulations were compared against a wide range of observations conducted during a measurement campaign in the Adventdalen area in the Svalbard archipelago. The model configuration with sub-kilometer grid spacing shows improved spatial structure of the near-surface temperature, wind and humidity forecasts. This includes wind channeling through the valley, down-slope wind intensification and cold-air pool formation. Although these phenomena are better captured, they show different magnitudes than observed. This leads to degradation of classical statistical verification scores for some variables compared to the 2.5 km forecasts. These results are in general agreement with previous studies on sub-kilometer NWP systems and offer valuable knowledge needed to progress towards operational sub-kilometer forecasting for Svalbard.