Some of the LR modeling are showing that the amplified pattern will last well into October(at least the first half) with some pretty chilly if not cold temperatures as far south as the upper southeast states. This makes sense with the affects of upper motion in the Caribbean and this persistent northwesterly flow pattern along the eastern periphery of the anomalous ridge situated over the western U.S. Starting this weekend a canadian air mass will move south into the area with the possibility of the first widespread frost or freeze across much of the great lakes, northeast, and parts of the mid atlantic. Temperatures across the plains will be mostly near to slightly below average and the west U.S. will be under the effects of predominant ridging that will generate mostly warmer and drier weather. However one particular model breaks down the ridging completely with a series of progressive low amplitude troughs swinging through the area. I'm not quite buying into this solution just yet.