Description: An Anomaly Correlation Coefficient (ACC) score of 1.0 on a particular valid date indicates that the forecast made 5 days prior to that valid date was highly accurate (i.e., essentially perfect) across the Northern Hemisphere. The closer the GFS (black line) is to an ACC score of 1.0 across the time series, the more accurate the GFS's 5-day forecasts valid during the last 31 days were. Colored lines represent other weather models, according to the key. The first number to the right of a weather model's name is that weather model's average ACC score during the last 31 days. The number of "days" written to the right of the average ACC score represents the number of 00Z forecasts included in the 31-day average. This number may be less than 31 if any 5-day forecasts valid during the last 31 days were missing.