The Climate Predictability Tool (CPT) provides an interactive Windows package for generating calibrated seasonal and sub-seasonal climate forecasts, performing model validation, and for verifying three-category probabilistic forecasts. A batch version of the software is also available, and is suitable more automation and/or for running large numbers of experiments. The CPT can be used for operational forecasting and/or for research purposes, and is designed to produce both empirical forecasting models and models for calibrating and potentially downsclaing the outputs of dynamical forecasting models. Although the software is specifically tailored for climate forecasting, it can be used in more general settings to perform canonical correlation analysis (CCA) , principal components regression (PCR) , multiple linear regression (MLR) , or probabilistic forecast verification (PFV) , on any data, and for any application. The General Circulation Model (GCM) option can also be used to validate models and produce single and multimodel forecasts, but is less likely to have broader application than the other options.

The following help pages provide guidance for using the software only in the applications for which it was specifically designed, but should be sufficiently general to provide guidance for other uses.

Last modified: