Starting with the temperature pattern in September 1991 notice the amazing similarities to this year with the core of heat over the northern plains and cooling across in the northeast as well as southern texas, florida and arizona.
Below is the 500mb pattern so far this september. A few important features include the western canada ridge with a band of lower heights extending from alaska towards western greenland. Also take note of the ridging on the asian side of the pole as well as a ridge east of Newfoundland. The lower height centre south of Norway is very similar in both years as well.
Looking further into the fall and winter here are what the next few months looks like in 1990-91 compared to the european monthly forecasts. These two factors alone are a huge player in the fall forecast.
October 1990 surface temps
November and December should stay warm in the east this year with a fairly strong southeast ridge but I do see more arctic air entrainment in the west then the euro is showing. Storm track will remain inland towards the great lakes/ohio valley as well as the B.C coast. Heading into January the cold continues to bleed further and further east in 1991 which is where I see this winter going. By february, I have some disagreements with the analog years portrayal of the upcoming pattern. It floods the entire CONUS with mild pacific air and widespread warmth. A strong alaskan vortex is present with a lack of greenland blocking.
Given this years lower solar regime, warmer SST's in the northern atlantic and warming ENSO state I expect a pick up in blocking by february with a cold month on the way for the eastern u.s. Thanks for reading and have an amazing rest of your week!