Forecast data for the paper: "Assimilating summer sea-ice thickness observations improves Arctic sea-ice forecast"

1. “Short_term_forecast.mat” contains the short-term ensemble mean sea-ice thickness (SIT) and sea-ice concentration (SIC) forecast results. The dimension of each forecast result variable is [420, 384, 7, 30, 5]. 420 and 384 is the number of the horizontal grids. Seven denotes the seven-day lead time of a single forecast. 30 shows the number of days in which each forecast starts after the model restart. Five indicates that the model restarts monthly from the restart files from the Combined Model and Satellite Thickness (CMST) product in five summer months each year.

2. “Long_term_forecast.mat” contains the long-term ensemble mean sea-ice thickness and sea-ice concentration forecast results. The dimension of the forecast result variables in each year is [420, 384, 270, 5]. Similar to “Short_term_forecast.mat”, 420 and 384 is the number of the horizontal grids. 270 denotes the 270-day lead time of a single forecast. Five indicates that the model restarts monthly from the restart files from CMST in five summer months.

3. “Initial_condition_Short_term_forecast.mat” contains the ensemble mean initial conditions of SIT as plotted in Figure 1 and Figure S2.

For more details please refer to the paper “Assimilating summer sea-ice thickness observations improves Arctic sea-ice forecast”