Prediction of Moderate and Heavy Rainfall in New Zealand Using Data Assimilation and Ensemble

Table 2

Categorical measures used in this study along with their definitions based on a binary truth table, where is the number of events which were forecast correctly, the number of observed events not forecast, the number of nonevents incorrectly forecast, and the number of nonevents correctly forecast.

Metric	Formula	Range
Frequency bias		0 to ∞ (perfect: 1)
Hit rate		0 to 1 (perfect: 1)
False alarm rate		0 to 1 (perfect: 0)
False alarm ratio		0 to 1 (perfect: 0)
Critical success index		0 to 1 (perfect: 1)
Peirce skill score		−1 to 1 (perfect: 1)
Symmetric extremal dependence index		−1 to 1 (perfect: 1)