Difficult Forecasts

It is quiet tonight in the Fox house. Helaine and Stef are asleep. I am alone in my office, lit by a single lamp. It’s 64&#176 outside so the window’s open.

If you’re really still the sound of intermittent rain falling from leaves can be heard. It’s tough to say whether it’s actually raining or if this is residual moisture from earlier showers. It doesn’t matter. To be outside is to be enveloped by the moisture that saturates the atmosphere… and now too the ground.

This is an all weather area. Little that comes from the sky can’t be handled on the ground. In nearly 20 years I’ve never seen our nearby brook do much more than get angry and stray a few feet. It looks pretty when it’s wet.

I never know how to describe this area. We more rural that suburban, though we’re not rural. The homes are spread out here, more because of well water and septic systems than anything else. You need room to produce and process water!

This is probably the first house built on this land. A stone wall in my backyard marks some old property line. Walls were always built in the open, but it’s now hemmed in by mature trees. It’s been a long time since anyone tried to grow anything on this rock infused New England soil.

I am being driven crazy by the current weather scenario. Not because of what it is, but because of how difficult impossible it is to forecast beyond 60 hours! I know a lot of people think 8-days is too far ahead to forecast, but there is some utility–especially the long range temperatures. Now three days has me flummoxed.

The computer models are fine tuned on actual weather, so when something really unusual takes place they have trouble following what’s going on. Obviously, this pattern is exceptionally unusual.

The 12Z GFS (a computer model run at 8:00AM or 12Z) showed a cutoff low in the Northern Plains for next week which basically stood still for days at a time. Could it happen? Sure, but it would be the first I’ve ever seen in 25 years here! And, of course, everything else in the model is closely related to this large feature. If it’s wrong (and it most likely is) everything else is wrong.

“Discard it,” you say. “If it’s wrong just ditch it.”

The problem is going beyond a few days humans aren’t capable of producing a reliable forecast without this high level mathematical help. We may know it’s wrong, but we don’t know what is right and there are a variety of possible solutions.

If it was just a question of working harder to get the forecast right I would. In the meantime I grin and bear it trying not to mislead those who trust me.