Hello folks,

We started noticing a problem last Friday with our Eta-RUC 850 windspeed analysis intercomparisons performed in real-time using 211 grid data on our web site. An example of the impact on 850 windspeed analyses can be viewed at

http://taiga.geog.niu.edu/verification/ETA_RUC850S12Z21Feb99.GIF

The image name needs to be typed in exactly because we do not have a specific link to this image on our web site. Our web site just has the latest analysis intercomparisons with the Eta, but I store examples without links when I see particularly large differences.

The image above is a 4-panel image which is set up as:

Upper left: Eta 850 windspeed analysis
Upper right: RUC 850 windspeed analysis
Lower left: RUC-Eta analysis difference
Lower right: Absolute value of RUC-Eta analysis difference

The largest differences are over Saskatchewan (40 kts) but there are also large differences (~20 kts) over eastern Quebec and Newfoundland.

Of concern to me since I don't know how everything works is how surface data can affect 850 windspeeds where the surface is well below 850.