Abstract
Verification has been completed for the NMC 6–10 day temperature outlook. The sample size involves 100 forecasts at each of 14 locations distributed uniformly across the continental United States. A separate regional analysis was performed at six locations in the northeastern United States.
The overall results indicate a marginal forecast skill versus climatology. Greatest skill is exhibited for much below normal forecasts which comprise 8% of all forecasts. Also, a conservative bias is observed in the outlook, as extreme temperature anomalies (i.e., much above and much below normal) are predicted less frequently than they are observed. The results also exhibit a striking similarity to the results of persistence-based forecasts, thus indicating that persistence is a principal source of guidance used in the forecast.