I wanted to circle back on this really quick. Since GEFS runs on a 0.5 grid (roughly 55 km by 38 km), Sumner was falling in a cell centered in the Cascade foothills that is at 770 m, while Sumner sits around 20 m. I made some changes to help mitigate that. Meteograms now compare the terrain of the four cells around your point to your real elevation and use the best nearby match. Sumner now reads the same lowland cell as Tacoma, and its member average of snowfall has dropped to about zero after this change (sorry Kolk!). Where no nearby cell matches (valley towns like Wenatchee and Leavenworth), the chart now shows a note with the model's elevation vs the real elevation so you know that snowfall may run high. Since CartoSky allows meteograms to be created for any town, unlike WB that has preset airports/locations, this will hopefully allow for a better, more accurate cell to be chosen for the given town on coarse resolution models. High-resolution models are unaffected and remain accurate to the degree that they can see the differences in elevation.