The table of maps below shows the predicted evolution of the 500 mb height (dashed lines) and the 850 mb temperature anomaly (departure from normal, shaded) from yesterday afternoon's GFS and ECMWF model runs. The maps are valid at 24-hour intervals from Tuesday afternoon in the top pair to Friday afternoon in the bottom pair. It's clear that the model agreement is excellent and therefore confidence is high. Remarkably, the models show 850 mb temperatures reaching more than 20 °C above normal over the eastern North Slope - I've created these maps before, but this time I had to adjust the color scale to accommodate the magnitude of the anomaly.
 |  |
 |  |
 |  |
 |  |
The good agreement between the models provides a nice opportunity to test my recent hypothesis that the National Weather Service temperature forecasts for Fairbanks are sometimes too conservative at the end of the short-range forecast period when the models agree in predicting a major anomaly. The 850 mb temperature anomalies predicted for the closest gridpoint to Fairbanks are shown below, and the current NWS forecast for the high temperature each day is also shown.
| Date (AKST) | GFS | ECMWF | Mean | NWS High Temperature |
| 3pm Nov 11 | +11.3C | +13.5C | +12.4C | 31F |
| 3pm Nov 12 | +17.2C | +15.3C | +16.2C | 34F |
| 3pm Nov 13 | +15.8C | +17.7C | +16.7C | 34F |
| 3pm Nov 14 | +13.6C | +14.3C | +13.9C | 26F |
We see that the NWS forecast is already calling for well above-normal temperatures, and in fact the forecast doesn't look conservative at all; the normal high temperature in Fairbanks on November 13 is 11 °F. However, the record high temperatures for this time of year are in the 40s, and the upcoming event looks like it intends to set some records. Could the thin (2-inch) snow pack be in danger in Fairbanks? We'll find out soon.