Interannual and seasonal variability in landfast sea ice growth and properties at Barrow, Alaska: A comparison between observations and CICE model simulations

Abstract:

Studies of long-term, regional variations of growth and properties of sea ice are best served by a combination of field measurements and model simulations. We explore the utility of the Community Ice Code (CICE) model, designed for fully coupled global climate models. CICE was run in standalone mode, with an integrated module that includes ice salinity as a prognostic variable (Turner et al., doi:10.1002/jgrc.20171).

We examine performance of the CICE salinity module by hindcasting interannual variability in the seasonal cycle of first-year ice salinity and temperature. Landfast ice, monitored at Barrow, Alaska since 1999 (SIZONet.org), provides a suitable test case and reference. The model is forced with 6-hr weather data from the National Climate Data Center, except for precipitation and humidity provided by the NCEP reanalysis model. Simulations were run for the period 1948-2013, with a focus on 1999 to 2013 when observation data are available.

Based on validation with measured ice and snow thicknesses, a non-zero ocean-to-ice heat flux has to be specified to reproduce the full seasonal cycle. The model captures the broad seasonal trends of key ice properties and thickness, especially during the growth season. During melt, significant deviations between observations and model output for ice temperature and salinity are observed, in particular near the ice surface where meltwater flushing is only partially captured by the CICE mushy-layer salinity module. With the exception of two years, the model captures interannual variability of ice thickness and bulk ice salinity well.

Further work is required to improve the accuracy of the full seasonal range of modeled salinity variations, especially during the melt season when processes at the ice interface are not well reproduced by the model. However, in conjunction with local validation data, CICE may serve as a tool to assess regional variations in key ice properties on interannual to decadal time scales.