Meteorological summer ends tonight in the Northern Hemisphere. That’s June/July/August. For much of the hemisphere those are the three warmest months of the year. The longest days are in late June plus the strongest sunshine of the year. The sun in September definitely doesn’t have the same “bite” as June and July!

Each September is different in our area. Some years the 80s and 90s keep going through early-mid September. In others (last year), “summer” just suddenly comes to an end. September was wet last year as the fall rains arrived early. We only made it to the 80 degree mark ONE DAY after the 6th of September. This September (at least the first half) promises to be much different.


  • A long spell of warm to hot weather begins Tuesday across the Pacific Northwest. It will last for at least 8 days, maybe longer. Expect mainly or all sunny days during that time.
  • There’s no sign of rain through at least the 8-10th of the month
  • Labor Day Weekend will be the hottest 3 day holiday weekend of the summer. It’s possible we approach 100 degrees Sunday and/or Monday west of the Cascades (not on the coastline).
  • Tuesday through Saturday those high temperatures range from warm to hot, but not record-breaking “crazy hot”. Our current 7 day forecast numbers look like this.

This has happened in the past. Back in 2011 we saw 9 days at/above 87 degrees in the first 10 days of the month. In 2017 we had 5 days at/above 90 in those first 10 days of the month, part of it filled with thick smoke and record warm nights due to the Eagle Creek Fire in the Gorge. Then in 1944 Portland saw 7 days at/above 90 degrees in those first 10 days. So it’s rare, but sometimes we do get a late season hot period. Good for your tomatoes, peppers, & squash, not great for sleeping.

Why are temps going to soar? Take a look at the 500mb map for today. Upper-level ridging to the west while a cool trough is slipping by to the east.

The result was our coolest day in almost three weeks…75 in Portland…nice.

That ridge moves right in over the West Coast, looking about like this by Thursday.

A sharp-looking ridge like that typically gives us some offshore flow and that will be the case on Thursday. Add in very warm 850mb temps (temperature around 5,000′ elevation) and that’s why I pushed highs into the lower 90s on that day. Then by Friday night that ridge has weakened a bit and upper-level heights are a bit lower. We could conceivably stay below 90 either Friday or Saturday for that reason, but I decided to keep temps right around 90 based on those very warm/hot 850mb temps. Models are keeping us in the 20-23 degree (Celsius) range Wednesday through Saturday.

You may noticed something else going on, another ridge is popping up just to the west of us Friday night. This one is going to be a real monster, check out the heights by Labor Day! A 597dm ridge over us, or just offshore, is quite rare. The maximum 500mb height over Salem based on many decades of atmospheric soundings is 598dm; this is close! It’s going to be a hot Labor Day Weekend…bit of an understatement there.

BTW, the colors represent the anomaly; you can see how unusual the ridge is up there in the Gulf of Alaska.

At this point (Sunday/Monday) models are in very good agreement showing 850mb temps around +26 to +27. That’s very close to the all-time record of +28 over Salem. If it was late July or early August, we’d be up in the 105+ range for a high temp in Portland. But it’s early September, and everything has to be just right (or wrong) for us to get to 100. In fact we’ve never gone above 100 after September 5th (PDX) or 7th (Downtown). That’s because nights are getting so long, sun angle is lowering quickly = it’s tougher to fully mix down the hot air overhead. The exact orientation of that upper-level ridge will determine whether we get easterly wind Sunday/Monday. Right now models are implying that will happen. That’s why I’ve got a record-breaking or record-tying numbers in the forecast for both Sunday and Monday.

Beyond Labor Day, models all agree the ridge sticks around for at least a few days, possibly much longer. The ECMWF model shows a full 11 days of very warm to hot weather through the 11th. These numbers are actually the ensemble average; the average of all 51 ensemble members

GFS ensembles are similar…

We will see how this all plays out, but prepare for a hot start to September!

Chief Meteorologist Mark Nelsen

Recommended for you

Load comments