Atmospheric response to changes in sea ice concentration within the marginal ice zone
In a newly published Nansen Legacy study, scientists Yurii Batrak and Malte Müller from the Norwegian Meterological Institute show how the resolution of the spatial sea ice characteristics in model simulations significantly influences the 12‐hr weather forecast in areas up to 500–1,000 km away from the sea ice edge.
The main idea of the study was to check how local kilometer-scale features in sea ice cover affect the atmosphere.
Modern weather and climate models, as well as reanalysis products, tend to have deficiencies in polar regions. One of the issues is the representation of the spatial structure of sea ice, and its interaction with the ocean and atmosphere. In most atmospheric modeling systems, sea ice cover is represented as a relatively smooth field, which does not reflect reality.
Standard deviation of differences induced by the modified sea ice cover over a 3.5-month model experiment in 6-hourly values of (a) 2-m temperature and (b) 10-m wind speed. Figure from Batrak & Müller 2018
In the current study, Batrak and Müller used a conventional low-resolution ice-concentration product and combined this with a lead product. To keep the total mass of sea ice constant, Batrak explains that "we did not just remove ice where leads had to be placed, but instead moved ice to the surrounding areas". The two scientists performed the experiments by using the ALADIN-HIRLAM NWP system with a 2.5 km horizontal resolution. They found that the presence of ice leads impacted air-sea turbulent fluxes. The resulting changes in near-surface temperatures and winds could be traced all the way to areas 500-1000 kilometers away from the ice edge. The results imply therefore that it will be beneficial for model applications to improve our capabilities in monitoring and simulation of sea ice lead characteristics.
The results show that utilizing high-quality fine-resolution datasets of sea ice cover in weather forecast models would have noticeable effect on the quality of forecasts.