CanSIPS Historical Forecasts

• Main Authors: Flato, Greg, Woo-Sung Lee, Boer, George, Scinocca, John, Polavarapu, Saroja, Kharin, Viatcheslav, Youmin Tang, Merryfield, William
• Format: Dataset
• Language: English
• Published: Canadian Cryospheric Information Network 2013
• Subjects: Arctic; Climatic models; El Niño; Forecast; Global; Sea ice; Beaufort Regional Environmental Assessment; Canada
• Online Access: https://dx.doi.org/10.5443/11812; https://www.polardata.ca/pdcsearch/?doi_id=11812

CanSIPS historical forecasts or hindcasts consist of output from two Environment Canada climate simulation and forecast models, CanCM3 and CanCM4, with ensembles of 10 forecast runs from each model in order to address the large uncertainties inherent in climate forecasts and to enable the probabilities of different forecast outcomes to be quantified. Hindcasts are initialized from the start of each month from January 1979 to December 2010, and have a duration of 12 months. Data consist of daily and monthly mean values for an array of atmospheric, surface, sea ice and ocean variables and are regridded onto a 2.5 degree grid (1 degree for ocean data) as described in ftp://dapp2p.cccma.ec.gc.ca/pub/goapp/CHFP/CHFP2_Data_Guide_v1.pdf, and is provided in NetCDF format. A subset of variables relating to sea ice, snow and ocean surface temperature is also provided on the original model grid. : Purpose: Climate forecasting systems such as the Canadian Seasonal to Interannual Prediction System (CanSIPS) require for their operation an extensive set of historical forecasts or hindcasts, typically covering a period of 30 years or more. While these hindcasts provide a basis for bias correction, skill assessment, and calibration of operational forecasts, in addition they serve as an essential resource for research efforts directed at understanding the nature of predictability in the climate system and improving climate prediction methods. This hindcast dataset is intended to serve that purpose with a particular emphasis on research into seasonal prediction of sea ice, enabled by its inclusion of an extensive range of sea ice-related data products which include daily and monthly averages of forecast sea ice concentration, thickness, and drift velocity components. : Summary: Forecasting agencies including Environment Canada produce long-range forecasts of monthly and seasonally averaged conditions up to a year or so in advance using climate models that represent the detailed interactions of the atmosphere, ocean, land and sea ice. The Canadian Seasonal to Interannual Prediction System (CanSIPS), based on two versions of the Canadian Centre for Climate Modelling and Analysis (CCCma) climate model, produces Environment Canada's official seasonal forecasts and at present is one of the few operational climate forecast systems in the world capable of predicting future evolution of sea ice. This set of CanSIPS hindcasts thus provides an important resource for research into sea ice and Arctic climate prediction.