ALBERT

Description: We examine the simulated Arctic sea ice drift speed for the period 2003–2014 in the coupled Arctic regional climate model HIRHAM–NAOSIM 2.0. In particular, we evaluate the dependency of the drift speed on the near-surface wind speed and sea ice conditions. Considering the seasonal cycle of the Arctic basin averaged drift speed, the model reproduces the summer–autumn drift speed well but significantly overestimates the winter–spring drift speed, compared to satellite-derived observations. Also, the model does not capture the observed seasonal phase lag between drift and wind speed, but the simulated drift speed is more in phase with the near-surface wind. The model calculates a realistic negative correlation between drift speed and ice thickness and between drift speed and ice concentration during summer–autumn when the ice concentration is relatively low, but the correlation is weaker than observed. A daily grid-scale diagnostic indicates that the model reproduces the observed positive correlation between drift and wind speed. The strongest impact of wind changes on drift speed occurs for high and moderate wind speeds, with a low impact for rather calm conditions. The correlation under low-wind conditions is overestimated in the simulations compared to observation/reanalysis data. A sensitivity experiment demonstrates the significant effects of sea ice form drag from floe edges included by an improved parameterization of the transfer coefficients for momentum and heat over sea ice. However, this does not improve the agreement of the modeled drift speed / wind speed ratio with observations based on reanalysis data for wind and remote sensing data for sea ice drift. An improvement might be achieved by tuning parameters that are not well established by observations.

Description: We examine the simulated Arctic sea-ice drift speed for the period 2003–2014 in the coupled Arctic regional climate model HIRHAM-NAOSIM 2.0. In particular, we evaluate the dependency of the drift speed on the near-surface wind speed and sea-ice conditions. Considering the seasonal cycle of Arctic basin averaged drift speed, the model reproduces the summer-autumn drift speed well, but significantly overestimates the winter-spring drift speed, compared to satellite-derived observations. Also, the model does not capture the observed seasonal phase lag between drift and wind speed, but the simulated drift speed is more in phase with near-surface wind. The model calculates a realistic negative relationship between drift speed and ice thickness and between drift speed and ice concentration during summer-autumn when concentration is relatively low, but the correlation is weaker than observed. A daily grid-scale diagnostic indicates that the model reproduces the observed positive relationship between drift and wind speed. The strongest impact of wind changes on drift speed occurs for high and moderate wind speeds, with a low impact for calm conditions. The correlation under low-wind conditions is overestimated in the simulations, compared to observation/reanalysis. A sensitivity experiment demonstrates the significant effects of sea-ice form drag included by an improved parameterization of the transfer coefficients for momentum and heat over sea ice. However, this does not improve the agreement of the modelled drift speed/wind speed ratio with observations based on reanalysis for wind and remote sensing for sea ice drift. An improvement might be possible, among others, by tuning the open parameters of the parameterization in future.

Description: A new version of the coupled Arctic atmosphere-ocean-sea ice model HIRHAM-NAOSIM is described. This version utilizes upgraded model components for the coupled subsystems, which include physical and numerical improvements and higher horizontal and vertical resolution, and a revised coupling procedure with the aid of the coupling software YAC. The model performance is evaluated against observationally based data sets and compared with the previous version. Ensemble simulations for the period 1979–2016 reveal that Arctic sea ice is thicker in all seasons and closer to observations than in the previous version. Wintertime biases in sea-ice extent and near-surface air temperatures are reduced, while summertime biases are of similar magnitude as in the previous version. Problematic issues of the current model configuration and potential corrective measures and further developments are discussed.

Description: For large scale and long term Arctic climate simulations appropriate parameterization of the surface albedo are required. Therefore, the sea ice surface (SIS) albedo parameterization of the coupled regional climate model HIRHAM–NAOSIM was examined against measurements performed during the joint ACLOUD (Arctic CLoud Observations Using airborne mea-surements during polar Day) and PASCAL (Physical feedbacks of Arctic boundary layer, Sea ice, Cloud and AerosoL) cam-paigns which were performed in May/June 2017 north of Svalbard. The SIS albedo parameterization was tested using measured quantities of the prognostic variables surface temperature and snow depth to calculate the surface albedo and the individual fractions of the ice surface subtypes (snow covered ice, bare ice, and melt ponds) derived from digital camera images taken onboard of the Polar 5/6 aircraft. Based on data gained during 12 flights, it was found that the range of parameterized SIS albedo for individual days is smaller than that of the measurements. This was attributed to the biased functional dependence of the SIS albedo parameterization on temperature. Furthermore, a temporal bias was observed with higher values compared to the modeled SIS albedo (0.88 compared to 0.84 for 29 May 2017) in the beginning of the campaign, and an opposite trend towards the end of the campaign (0.67 versus 0.83 for 25 June 2017). Furthermore, the surface type fraction parameterization was tested against the camera image product which revealed an agreement within 1 %. An adjustment of the variables, defining the parameterized SIS albedo, and additionally accounting for the cloud cover could reduce the root mean squared error from 0.14 to 0.04 for cloud free/broken cloud situations and from 0.06 to 0.05 for overcast conditions.

Description: For large-scale and long-term Arctic climate simulations appropriate parameterization of the surface albedo is required. Therefore, the sea ice surface (SIS) albedo parameterization of the coupled regional climate model HIRHAM–NAOSIM was examined against broadband surface albedo measurements performed during the joint ACLOUD (Arctic CLoud Observations Using airborne measurements during polar Day) and PASCAL (Physical feedbacks of Arctic boundary layer, Sea ice, Cloud and AerosoL) campaigns, which were performed in May–June 2017 north of Svalbard. The SIS albedo parameterization was tested using measured quantities of the prognostic variables surface temperature and snow depth to calculate the surface albedo and the individual fractions of the ice surface subtypes (snow-covered ice, bare ice, and melt ponds) derived from digital camera images taken on board the Polar 5 and 6 aircraft. The selected low-altitude (less than 100 m) flight sections of overall 12 flights were performed over surfaces dominated by snow-covered ice. It was found that the range of parameterized SIS albedo for individual days is smaller than that of the measurements. This was attributed to the biased functional dependence of the SIS albedo parameterization on temperature. Furthermore, a time-variable bias was observed with higher values compared to the modeled SIS albedo (0.88 compared to 0.84 for 29 May 2017) in the beginning of the campaign, and an opposite trend towards the end of the campaign (0.67 versus 0.83 for 25 June 2017). Furthermore, the surface type fraction parameterization was tested against the camera image product, which revealed an agreement within 1 %. An adjustment of the variables, defining the parameterized SIS albedo, and additionally accounting for the cloud cover could reduce the root-mean-squared error from 0.14 to 0.04 for cloud free/broken cloud situations and from 0.06 to 0.05 for overcast conditions.