(always under construction)

In general the AVN/GFS model tends to overestimates temperatures at the summit by about 1 C. However, different synoptic and mesoscale conditions warrant changes to this bias and the models prediction. For example, strong winds will promote mixing of cold air aloft and thus at night could lower the temperatures by 2 to 3 C. While during periods of weak winds one may expect radiational cooling to cool the temperature off even more. However, the cold air quickly slides right down the mountain. So little, if any, nocturnal inversion sets up; if it does, it is very shallow. Thus the bias to the models prediction is 0 to -0.5 C. During periods of strong subsidence, however, things are bit difference. Here the AVN/GFS underestimates temperatures as adiabatic warming can significantly increase temperatures, which the model doesn't get a good hold of. Thus add 1 to 2 C to AVN/GFS's forecast. Here is a general table that describes different scenarios that temperatures should be adjusted.

No Subsidence
With Subsidence
Daytime Bias
Nighttime Bias
Daytime Bias
Nighttime Bias
Light Winds (<= 15 mph)
+5 to +6 oC
-0.5 to -1 oC
+7 to +8 oC
+1 to +2 oC
Moderate Winds (15 < winds < 30 mph)
+3 to +4 oC
-1 to -1.5 oC
+5 to +6 oC
0 to +1 oC
Strong Winds (>= 30 mph)
+1 to +2 oC
-1.5 to -2 oC
+3 to +5 oC
-0.5 to 0 oC

In general the AVN/GFS model underestimates precipitable water by as much as 0.5 to 1 mm, although when the precipitable water is very low (< 2 mm), the bias is on the order of 0.25 to 0.5 mm. There is also a small concern regarding low-level moisture ascent. Normally the precipitable water value is the integration of the water vapor content above 600 mb. Obviously mixing of low-level moisture will taint that value, so those instances must be taken into account. On the other hand, during times of strong subsidence, which AVN/GFS doesn't have a good handle on, precipitable water will rapidly decrease.


The ANV/GFS model has a pretty good handle on the environmental winds at the approximate summit height, although there are a few exceptions that I have noticed. The first is during periods of light winds, the wind direction is not accurate, but that is hardly a factor. Another exception occurs during periods of fog, precipitation, and/or rain at the summit. The wind speeds under these conditions are definitely underestimated, as mesoscale outflow boundaries from precipitation are not evident in the model. The last, more serious, issue is when the model forecasts winds above 35 mph. During these periods, the AVN/GFS grossly underestimates wind speeds. For example, the model may forecast 40 mph winds, but winds exceeding 70 mph was recorded. On another occasion the model expected winds near 70 mph, but 125 mph was recorded.

Last Modified: Friday, January 17, 2003
Comments to: