We got about half an inch of rain last night. Hooray! That put about 80 gallons of water in our catchment tank, bringing the level close to the very top. It gave me a reprieve from having to water our newly-planted trees, a chore which was on my list for today.
It puts our annual rainfall at 18 inches in the past 12 months, so we’ll have to have a very wet fall if we’re going to get anywhere near our usual annual of 70-100 inches for the calendar year. Yes, it has been really dry.
Fun fact: in just the two days before I moved here last August, we got 19 inches at the property, or more than we’ve gotten in the entire year since. So getting 60 or so inches in the next few months is certainly possible, though I would greatly prefer it not all happen in the same week.
What was notable about this rain, though, is that it was actually predicted by the most reliable local weather forecast model I’ve found (GFS). Kinda. Sorta. GFS had predicted that we’d get 0.05-0.08 inches, so it was off by a factor of ten or so. Still, I felt surprised that the prediction had been “correct,” until I realized that it really wasn’t. I was engaging in confirmation bias.
Shortly after I moved here, I was complaining to a neighbor (we’ll call him “Phil” because I don’t know anyone around here with that name) that I couldn’t find an even remotely accurate forecast. He said that he knew some good ones. I asked if he would share. He got out his tablet and proceeded to take me through five different weather forecasting sites. One showed rain. One showed heavy clouds. One was mostly sunny. I asked him how he knew which one to trust, and he said something along the lines of “if you look at all of them every day, you get a sense of which one is accurate for that day.” Over time, I have come to believe that Phil just looks at several forecasts, then at the end of the day remembers that at least one of them was correct and so feels he has a handle on the weather.
To be fair to the weather forecasters, predicting weather here isn’t easy. We’re a very small (but not tiny) land mass in the middle of the largest ocean on the planet. Altitude and climate varies by the foot. I’m at over 2,000 feet elevation. The ocean — “sea level” — is about three miles away. The summit of one of the tallest mountains on the planet is about fifteen miles away. Heck, Phil’s house is a short walk from here, and his weather is markedly different from mine (and it’s possible one or more of the forecasts he looks at are actually useful at his location).
I routinely check four forecasts, though, and while there are occasional hits, there are far more misses. Weirdly, GFS is a regional forecast (22 km radius) but is generally more accurate than the hyperlocal forecasts, including the “AI enhanced” ones. Which tends to make me think it’s all just chance.
Yesterday, for example, GFS predicted rain in the afternoon, yes. It also predicted it would be mostly cloudy all day, yet it was one of the best solar energy days we’ve had: I overcharged the car, heated the hot tub to its max of 105°F, did two loads of laundry (dried with the electric dryer) and still ended up wasting a significant amount of solar power for the day. The rain was predicted for the afternoon but didn’t materialize until after 19:00, then was much heavier than predicted. So to say that the forecast was correct would be allowing for an awfully generous interpretation of “correct.”
Last week, also, I had a rare alignment of all four forecasts. They all predicted “mostly cloudy.” I thought surely they’d be close. It was one of the sunniest days we’ve ever had — and that’s saying something in this tropical paradise.
Does any of this really matter? I don’t think I have ever paid a lot of attention to weather forecasts, and my quality-of-life doesn’t seem to have suffered. Then again, I lived much of my life pre-internet, living in parts of the West with particularly boring weather, and I was thoroughly connected to the grid and my energy and water supplies weren’t weather dependent. Now, advancing or delaying laundry by a day or putting off a long drive can actually make something of a difference.
Perhaps this is something like what being a farmer feels like.
—2p