Climate

154 appellations · normals 1991–2020 · vintages 1980–2025

Growing-season warmth by appellation

(4)(28)(59)(41)(22)

Hovering an appellation names it; a click opens it.

What the five heat classes mean

2,500 degree days or fewer4 appellationssparkling wine and the earliest-ripening varieties · e.g. Comptche 2,370
2,501 to 3,000 degree days28 appellationscool coastal valleys · e.g. York Mountain 2,878
3,001 to 3,500 degree days59 appellationsmuch of the North Coast · e.g. Covelo 3,257
3,501 to 4,000 degree days41 appellationswarm valley floors and foothills · e.g. California Shenandoah Valley 3,691
above 4,000 degree days22 appellationsthe hot interior — high yields, early picks · e.g. Diablo Grande 4,168

The ten coolest appellations

appellation	degree days	class
West Sonoma Coast	2,200
Comptche	2,370
San Luis Obispo Coast	2,443
Petaluma Gap	2,491
Sonoma Coast	2,583
Monterey	2,670
Edna Valley	2,684
Arroyo Grande Valley	2,771
Carmel Valley	2,786
Green Valley of Russian River Valley	2,819

The ten warmest appellations

appellation	degree days	class
Antelope Valley of the California High Desert	4,656
Cucamonga Valley	4,543
Madera	4,434
Sierra Pelona Valley	4,340
Temecula Valley	4,326
Winters Highlands	4,312
Squaw Valley-Miramonte	4,287
Dunnigan Hills	4,257
Paulsell Valley	4,244
Leona Valley	4,230

Harvest-window rain by appellation

Driest harvest windows

appellation	harvest	season
Antelope Valley of the California High Desert	14 mm	32 mm
San Lucas	16 mm	47 mm
Tracy Hills	16 mm	51 mm
Arroyo Seco	16 mm	49 mm
Hames Valley	17 mm	48 mm
San Bernabe	17 mm	50 mm
Madera	17 mm	57 mm
San Miguel District	17 mm	45 mm

Wettest harvest windows

appellation	harvest	season
Seiad Valley	93 mm	252 mm
Willow Creek	87 mm	267 mm
Fort Ross-Seaview	85 mm	249 mm
Rockpile	85 mm	220 mm
Dos Rios	81 mm	238 mm
Mendocino Ridge	81 mm	224 mm
Trinity Lakes	76 mm	246 mm
North Yuba	75 mm	253 mm

Harvest rain and growing-season heat, 1980–2025

Marked years sat furthest from the long-run average — labeled in red on the panel for the measurement that was unusual.

Years more than 1.5 from the 1980–2025 mean

2021high harvest rain · 180 mm at harvest
2020high growing season heat · 3,905 degree days
2016high harvest rain · 136 mm at harvest
2010low growing season heat · 2,943 degree days
1982high harvest rain · 118 mm at harvest
2004high harvest rain · 117 mm at harvest
2014high growing season heat · 3,777 degree days
1989high harvest rain · 114 mm at harvest
1982low growing season heat · 3,033 degree days
1998low growing season heat · 3,053 degree days
1980low growing season heat · 3,068 degree days
2017high growing season heat · 3,724 degree days

These are weather measurements, not vintage ratings. No published source of California vintage ratings is available here, so each callout is what the climate series measured that season.

Appellations are moving into warmer heat classes

Decade means — 1980–1989 against 2016–2025. Each numeral opens its definition.

6 → 4 -32,500 degree days or fewer
39 → 18 -222,501 to 3,000 degree days
60 → 55 -63,001 to 3,500 degree days
36 → 46 +113,501 to 4,000 degree days
13 → 32 +20above 4,000 degree days

Raw GridMET against the atlas's corrected record

California statewide

+281

GDD, 1980–1989 to 2016–2025

published · 3,281 → 3,562

+282

the same, undownscaled

raw 4 km · 3,374 → 3,656

The two deltas agree. Rebasing GridMET on PRISM moves the whole series down by about 93 GDD and leaves the slope where it was — so the warming here is in the underlying record, not in the correction applied to it. That is the only claim this plate makes, and it is why the raw series is published beside every downscaled column rather than discarded.

Modeled ripening is a week earlier than in the 1980s

1980–19892016–2025

variety	earlier by	appellations
Cinsaut	11.7 days	45
Mourvedre	11.4 days	26
Syrah	11.2 days	63
Grenache	10.9 days	69
Petit Verdot	10.6 days	75
Gamay Noir Au Jus Blanc	8.2 days	148
Chardonnay	7.7 days	150
Pinot Noir	7.5 days	150

Drought, year by year

California statewide

Categories, mildest to most severe — each opens its definition

How much the heat class depends on the dataset

Question one — how often do the three datasets agree?

109of 154

appellations get the same heat class from all three datasets

45of 154

land in a different class depending on the dataset used

42 of them differ under the raw 4 km series, 6 under the 800 m baseline alone

The atlas publishes all three on every region page rather than picking one. The three surfaces are raw at 4 km, ’s 800 m normals, and the correction between them; they disagree most where the ground is steepest.

Question two — how close is any of it to a thermometer?

5,876 complete station-seasons from 262 weather stations sitting inside 60 appellations. Lower is closer.

series	average error statewide	typical error, any one place
Published (corrected to the 800 m baseline)	-17	164
Raw GridMET 4 km	+78	149

average error statewide: -17
typical error, any one place: 164
average error statewide: +78
typical error, any one place: 149

(°F), against a station mean near 3,410. Statewide error is the average of signed errors, so opposite-signed regional errors cancel within it; the second column, like , does not.

The correction helps statewide and not region by region. Averaged over California it is effectively unbiased (-17 on a 3,410 mean). Taken one appellation at a time it moves 26 of 60 closer to their own stations and 34 further away. The statewide figure is not a statement about how accurate any single appellation is.

The widest gaps between a published figure and its own stations

Arroyo Grande Valley -773 · 1 station
Capay Valley -608 · 1 station
Gabilan Mountains -578 · 2 stations
West Sonoma Coast +526 · 2 stations