Assessment of Sea Ice Predictability in CMIP5 Decadal Hindcasts

Thursday, 18 December 2014
Jiping LIU and Chao-Yuan Yang, SUNY at Albany, Albany, NY, United States
This study assesses the decadal hindcast/prediction simulations of CMIP5 coupled climate models. We focus on decadal hindcast/prediction that consists of simulations over a 10 year period initialized every five years from climate states from 1980 to present. The model prediction skill is evaluated by comparing seasonal mean (as well as September) sea ice extent from the observation and hindcasts of each model. To further examine the spatio-temporal prediction skill of the individual model and the multi-model ensemble, the average for the lead times of 1-year, 2-5 year and 6-9 year sea ice extent anomaly field is compared with the observation.