HWT 2011 Spring Experiment Model Evaluation

HWT/DTC 2011 Evaluation - Model Description

During the 2011 Spring Experiment, the DTC will be providing objective evaluation as a supplement to the subjective evaluation portion. Below is a list of variables (and thresholds) to be evaluated during the 2011 Spring Experiment.  DTC will perform evaluation of the Storm Scale Ensemble of Opportunity, several sub-ensembles representing physics experiments, and the LAPS short-range ensemble retrospectively.  Additional variables, such as 1-hr Accumulated Precipitation and 1-km reflectivity and corresponding probability fields will also be retrospectively evaluated, as resources allow.

Key:Traditional statistics include Gilbert Skill Score (GSS), Critical Success Index (CSI), Frequency Bias (FBIAS), Brier Score, Receiver-Operator Characteristic Curve (ROC), Area under ROC Curve (AUC), and Reliability Diagrams.  The statistic calculated for the neighborhood method is Fraction Skill Score (FSS).  MODE attributes include centroid distance, intersection area, symmetric difference and more.  Rank Histograms (or Talagrand Diagrams) and Spread will be provided for SSEF and SSEO ensembles.