This past summer, the first in Eugene’s records to have no rain in June, July, and August, the driest 11 month period on record, was almost not the driest. In late July the weather models showed a significant storm system about 10 days out, then 9 then 8 then 7 then 6 days that was going to deliver a big soaker in early August, almost unheard of. Given how dry we had been, this was eagerly awaited, and after all, 6 day forecasts aren’t too bad. At 6 days out, the TV meteorologists were all over this storm. The next morning, first thing I did was check the models: to my chagrin, both the GFS and the Euro forecasted that storm wasn’t going to happen. It was gone. Kaputt. Sayonara. Weg. Hasta la bye-bye. Evaporated from fantasy. Never happening. It was a real bummer to me, and as I learned recently, to many others as well.
For those few of you who are model riders, like me, it’s OK. I understand. I feel your pain.
I follow the Portland Weather Blog (TV weather caster Mark Nelsen), and the California Weather Blog (Daniel Swain), Cliff Mass in Seattle, reading their comments, but it wasn’t until I read an article in Bay Nature that I realized there was a kindred group down in the Bay Area. They coined the term “model rider,” not me.
A model rider logs on a few times daily to check the long range weather models: the GFS, or Global Forecast System (American); the ECMWF (the European Centre for Medium-Range Weather Forecasts, or the Euro), and the GEM model (Global Environment Multiscale, Canadian.) I don’t check the latter too often, but I look at the other two. I even get the detailed European ones, which I have to pay for, but it’s worth it. I have access to world wide weather on the world wide web, and I can look at high pressure domes baking Scandinavia, or get 2 meter temps in Casablanca, should I want to. I don’t, but I do have the anomalies (differences from normal) for the upper level winds, the 3,9, and 14 km models for precipitation, the jet stream, and the low level winds. I can get local temperature forecasts using seven different models, four of which I had never heard of. I can scan what I need in a few minutes, and if I’m traveling, I can check out a lot of other stuff. I can get 500 mb heights for the upper level, anomalies, or departures from normal, all color coded, run them fast or slow out 3, 10, or 16 days, and even 46 days if I want to take a look at the precipitation anomaly.
Following the models gives me an idea of what I might expect in the long range. I have watched the models come into agreement about a storm, or a prolonged high pressure system (more on that shortly), or a big wind event, days before it makes the forecasts or even the National Weather Service forecast discussions, which are also about four times a day. I follow them in several places—Portland is my home area, but I also look at Medford, Seattle, Reno, and Sacramento. Some of these—Medford especially—give a great synoptic summary, how something we have here, like a persistent high pressure system, which many of us refer to as a “Death Ridge,” is related to and stays as a result of deep troughing, or low pressure, downstream on the East Coast. Medford is realistic; Portland is a little behind. This morning, Medford said the next storm was more than a week away and might split. Portland was still talking about the sprinkles Wednesday that won’t hit where I live. Medford NWS used the term “sacrificial front,” the first front that slams into a high pressure system, weakening both the front and the system, but allowing a second and additional fronts to break through.
That assumes the high pressure system doesn’t build back in behind the front, which it often does. That has happened a lot this year.
The Bay Area group refers to forecasts out 200-300 hours as “fantasyland.” They are right, and they confirm what I discovered my first winter here: numerous models kept showing big storms that were going to hit us in the long range time frame. About 7 days out, those storms disappeared from the models. For those of us who think rain is good, like real weather, these models sometimes pull a “pull the football up,” referring to Lucy and Charlie Brown, in the classic Peanuts cartoon, where Charlie Brown hoped it wouldn’t happen and always got fooled by Lucy at the end. Fantasyland is really wet. Reality is really dry.
It’s not often I read something where the comments and feeling expressed match mine so perfectly: if the models say persisting high pressure is going to set up for two weeks, persisting high pressure will set up for at least two weeks. If the models say it’s going to rain in 7 days, there will be a high pressure system instead. We model riders know that. Storms get delayed a day, then 2 days later delayed another 2 days, then disappear. Want wet weather? Fantasyland awaits.
One of the commentators on the Portland weather blog was bemoaning that “Seattle got all the good stuff,” meaning weather. I wrote back and said that those of us (like me) in the South Valley thought that Portland got all the good stuff. Nobody is happy, unless we get some good action.
The models over predict the rain we do get: I cancelled a hike in the Cascades last week because of what was going to be a good soaking with a lot of wind in the mountains. I was leery about cancelling it, because in my own mind I knew we were going to get minimal rain down here, because that has been the pattern for almost the last eleven months. That’s what a drought is, and we are in a severe one. We got minimal rain. I wouldn’t have needed rain gear.
I waited 20 years in Arizona for the drought to break, and it finally did—after I left. I stopped looking at the Climate Prediction Center for Arizona, because it invariably predicted higher than normal temperatures and less than normal rain. And for at least seven years that I looked, it was always right.
The latest GFS model said 10 days ago showed a good rain this weekend, then the models backed the rain off until the following Wednesday. What I am now hearing from the forecast discussions is “rain later this month, near Thanksgiving.” The problem is that Thanksgiving this year is early. Besides, for the last few runs, the models have shown dry weather out 10 days. Today, they show nothing for 16. Thanksgiving is the 22nd.
Maybe they meant “rain later this year, near Christmas.” Or, as I learned to say during the monsoonal busts in Arizona, “maybe it will rain next year.”
Bummer.
Tags: General writing
Boreal B[l]og 
Leave a Reply