Yeah, it certainly is hard to wrap your head around the transitions between zones sometimes where the numbers are wildly different in similar/neighboring zones. My sunset zone (which is 5) is basically a slightly cooler version of sunset zone 17. I’d argue that using numbers isn’t really ideal for systems that try to capture so many different factors.
I do like the Köppen classification system because it uses letters instead of numbers, but the temperature range in each zone is pretty wide. Under that system, Seattle (and most of the lowlands in PNW) are Csb (warm-summer mediterranean).
Maybe no single zone system tells the whole story, but when taken together you can glean a lot. So I’m in:
Yes, and when it’s a slow gradual sustained dip to -25 vs going from 50 to -25 overnight…maybe even several times. I think they should call this the “winter yo-yo coefficient”.
I do like how the Sunset zones account for so many different factors. And the individual zone descriptions have some nuance depending on where you are actually located. The USDA system not so much.
For example, Napa where I live, Sacramento, and Crescent City are all 9b in the USDA system, but on a hot day that could mean a 10 degree difference between Napa and Sacramento and a 50 degree difference between Sac and Crescent City.
Napa and Sacramento are both in Sunset zone 14, while Crescent City is zone 17, the difference between the two zones being the amount of coastal influence.
Here’s part of the description for zone 14:
Marine air moderates parts of Zone 14 that otherwise would be colder in winter and hotter in summer. The opening in Northern California’s Coast Ranges created by San Francisco and San Pablo bays allows marine air to spill much farther inland. The same thing happens, but the penetration is not as deep, in the Salinas Valley. Zone 14 includes the cold-winter valley floors, canyons, and land troughs in the Coast Ranges from Santa Barbara County to Humboldt County.
The milder-winter, marine-influenced areas in Zone 14 and the cold-winter inland valley within Zone 14 differ in humidity. For example, lowland parts of Contra Coasta County are more humid than Sacramento.
I think I will base my plant selections on the coldest recorded temperature in the last 20-30 years. Averages do not matter to a tree. My own rule of thumb is to choose plants that are hardy to a full zone below the minimum that I am likely to encounter.
I’d have to rule out trying to grow so many fruits if I was that conservative.
I know that if I try growing something that it might be killed by a freeze the very first winter… Or it might be fine for several years, provide some fruit, then die.
Balancing fruit that requires chill accumulation that will be met more years than not usually gets a tree / vine that has higher minimum temps it can tolerate as an established tree with a few years if higher cold sensitivity in the beginning.
I’m trying to fit in as many varieties that I can in my very limited space to at least have a handful that each year might fruit even while losing some along the way.
This raises a good point regarding my discussion of how hardiness ratings are set. Depending on when and where plants are tested and which iteration of the map the tester chooses, you can get some real wild cards thrown in. For example, if I trialed a plant for 5 years at my location, now rated as 6a, but I reference the map from 20+ years ago, I might erroneously rate the plant as Zone 5a hardy. Oops!
went from 3a 40+ years ago. now this current one says 4b! i may get chestnuts, persimmons and peaches to survive here yet! as long as snow depths continue to be somewhat normal, it opens up alot of opportunity here.
I think making big plans on a half step higher zone is problematic.
Obviously we all know this is based on a 30 year floating average for temps, but I would think those same areas are likely still to get as cold at least once every few years, so if we are only speaking of cold hardiness in choosing plants and trees, I’m not sure I’d change my decision making based on the new hardiness maps.
It may be more applicable to chill accumulation… Some places might increase as it spends a bit less time below freezing while other areas might lose hours being above 45 a bit more.
Looks like I’ve gone from 5a to 5b, and it bears out with my memory of the last few winters here in SE Wisconsin. It’s kind of a tossup nowadays if we’ll get a white Christmas.
Didn’t get our first freeze until the night the goblins came pecking on doors.
Guess that’s the reason you’re a half zone colder.
I still use all 100+ years, and that puts us back in zone 6. Selecting a one and a half to three decade period is setting up UDSA to look stupid (again) as the pendelum goes the other direction next time.
I graduated to 7b from 7a, and before that was 6b. I’m only five miles from 8a now. Maybe my figs will finally stop dying back. I sort of doubt it really, things are warmer on average but those rare bad winters seem pretty close to the older bad ones.
at least stuff like figs will come back from the roots so the warmer years should still give you fruit. i just put in Siberian peaches. a few seedlings and a bunch of pits. they are on their own roots so if they die on a cold year they should come back as the snow here will protect the roots. should be the same with persimmons and chestnuts. fingers crossed.
I agree to a point, but if you get those colder winters less often, any trees that need a few years in the ground before they become hardy enough for a bad year might have a better chance of getting established now than they used to. That’s my hope with avocados here. An avocado that’s been in the ground three or more years is much more likely to survive that rare “zone 8a range” winter than a first-year seedling.
I made this chart for 1948 through 2022, and you can see 1990 was the last time we had a day below 10°F, or more than one day below 15°F, and we haven’t had a year with more than 10 days below 25°F since 1993:
By gosh, you’re right! That interpolation though … whew! There’s no time dimension to daily lows, true, so no homogenization. The interpolation is over current data. But pick a point — any point — between several weather stations, and they have to deliver a gridded value for the 30-year average annual minimum temperature nearest that point. They weight the stations by reliability, by distance, and by nearness to one another to conjure the gridded value in advance. Then they apply a terrain model that adjusts the reported temperatures to what is likely to prevail at the elevation of the gridded location while considering the other things you mention, too. Yeah, it’s linear interpolation but on steroids.
