GFS ENS Warm Forecast Bias in the West U.S.

Highlight: ECM ENS vs. GFS ENS temperature bias past 30 days points toward ECM ENS as forecast model of choice.

Fig. 1-2: ECM ENS and GFS ENS 15-day temperature bias across North America from the past 30 days clearly favors ECM ENS as the better model.

Discussion: An extraordinary difference in temperature bias shown in 15-day forecasts from the past 30 days appears when comparing ECM ENS vs. GFS ENS. The ECM ENS is reasonable with patchy relatively small temperature bias slightly favoring the warm side (Fig. 1). However, the GFS ENS is boldly too warm across the Southwest and Interior West U.S. and to a lesser degree the Carolinas (Fig. 2). The much too warm bias by GFS ENS may be related to the generally warm bias attached to regions of dry soil moisture (or drought).

Consequently, favoring the ECM ENS for the mid-summer forecast is preferred. The 15-day outlook is consistent maintaining a hot pattern across the Great Plains (Fig. 3). ECM ENS extends the anomalous warmth across the Northeast U.S. during the medium-range while the GFS ENS is cooler. The anomalous heat also affects central and southern California.

The precipitation forecast remains very dry across the central Great Plains to Texas and across California and the Northwest U.S. (Fig. 4). However, so far this week, the East/Southeast U.S. is trending wetter and the back edge of this wet regime is extending to the Tennessee Valley and eastern half of the Ohio Valley. The wet forecast into the Northeast suggests that the warm ECM ENS forecast for New England may be too warm.

“Flash drought” commentary in yesterday’s forecast remains valid in the central/south Great Plains with highest concern in Kansas to Nebraska where deep layer soil moisture deficits are driest.

Fig. 3-4: ECM ENS 15-day temperature/precipitation anomaly forecasts for the U.S. with annotation highlights.