There are some important changes here in the forecast so you need to pay attention.

Over the past few weeks we have had several frost scares   appear on the long-range models either at Day 9 or 10 or in the 11 to 15 day. Everyone of these scares has been brought about by the GFS model or the CMC -- the Canadian Model. Not a single time over the last three weeks has the European weather model show a pattern which is even remotely conducive to a frost and large-scale colder outbreak.

That has now changed. The 0z Monday operational European model is developing a major colder outbreak for late September which IF correct ...would bring a significant frost and much of the central and Upper Plains and the Midwest in the Sept 23 - 26 time frame .

In yesterday's report I mentioned how European was developing a new deep trough over Western Canada at Day 10 of which seemed to want to drop into the Plains at day 11. The New Monday morning European model run has done just that by developing a TWO major troughs... one in the 6-10 day and the 2nd one at day 10.

The first trough develops the major storm over Western Canada which tracks into northern Manitoba Northwest Ontario and an optimum Hudson's Bay Sept 19-22. This Low's associated cold front moves through the Plains and Midwest September 20 -- 21 that reaches the East Coast on the 22nd. This front will probably have significant showers and thunderstorms with it and a second low pressure area could develop on the front over Arkansas in southern Missouri on the 22nd  . If that happens additional rain will fall over the ECB September 23... but this aspect of the cold front passage is still uncertain and could change.

The second trough moves in to Western Canada September 22... while a strong ridge in the Jet stream builds along the West Coast and eastern Pacific. This forces the trough the drop into the northern Rockies and the upper Plains with a strong cold front running from southern COL to southern MN on the morning of the 24th. Temperatures are cold enough according to the model for significant frost over eastern Montana North Dakota all as a statue in Manitoba and Western Ontario.

The 0z GFS model handles things significantly different. The 1st deep trough over Western Canada moves due East according and becomes a major storm over eastern Canada with a trough over the Midwest and the Northeast. However according to this Model the cold air never plunges into the trough and as result there is no cold air outbreak and no frost. On the other hand the 0z Monday GFS model also develops this Fantasy hurricane off the East Coast in the 11-15 day which is a sign that the model is clueless and probably should be disregarded.

Indeed the new updated run of the GFS -- the 6z run which comes out between 6 and 7 am EDT --looks very much like the European as this model does show a massive trough coming into the Plains and the Rockies with a significant cold air outbreak coming south. Good forecasters know that when the 6z GFS looks like the ECMWF... it usually means that the European model is correct.

Lastly keep in mind that this is still a long way off. I do believe it is significant that the European model is now showing a significant colder outbreak for the first time. In addition the way in which the European model is developing this cold air outbreak makes sense: so far this autumn we have seen a lot of energy in the Pacific jet most of which has a moving into Western Canada. This tells me that THIS model solutions ...which brings the cold air in from Western Canada... into the Rockies than the Plains and then into the Midwest probably the correct solution... which is what the European is showing. And the scenarios s which have a cold air plunging out of Manitoba into the heart of the Midwest--- which is what the GFS has been showing a number of times since September 1-- is probably the incorrect solution.