When does grass start growing across the US? What 11 cities of backtesting found

"When does grass start growing" doesn't have one answer in a country that spans Seattle's mild winters and Minneapolis's short, intense summers. So instead of guessing, we checked.

The closest independent check available

No citizen-science project tracks "first lawn mow" specifically across the US. The closest independent, quantitative dataset instead is the USA National Phenology Network's Extended Spring Index — First Leaf: a 30-year average (1991-2020) tracking when lilac and honeysuckle shrubs leaf out each spring, queried directly at each covered city's exact coordinates.

First Leaf is a different, earlier event than first mow — a shrub leafing out isn't a lawn becoming mowable — so this isn't a same-event match. What it can check is whether the gap between the two events stays consistent across cities (expected, if the model behaves the same way everywhere) or not (a sign something's off in a specific climate).

Two groups, and one real pattern

Seven East and Midwest cities cluster in a fairly tight 12-30 day gap between first leaf and the model's predicted first mow. Four Western cities — Portland, Salt Lake City, Denver, Seattle — run 41-58 days, roughly double.

Elevation was ruled out first: Open-Meteo's grid-point elevation was checked against the USGS's independent Elevation Point Query Service at Denver and Salt Lake City's exact coordinates, landing within 3m of ground truth at both. The weather data isn't misplaced in the terrain there.

Two real, separate explanations accounted for the rest:

Denver and Salt Lake City are the two most water-stressed cities in the whole set — the highest watering-alert frequency, the lowest mow frequency. The model already stretches the mow interval under water-balance drought stress; a semi-arid climate genuinely producing slower real growth than a temperate one is exactly what that mechanism is for.

Seattle and Portland don't fit that explanation — their watering-alert rates are unremarkable, similar to several East/Midwest cities. Checking the raw soil-temperature data directly (not the model's output, just how long it takes soil to warm from 5°C to 15°C each spring) found the real cause: both cities take 50-90% longer than even the slowest of the other nine — 117 and 122 days, against 33-79 elsewhere. A mild, maritime winter lets soil cross the low threshold very early, which is what First Leaf picks up on. But a cool, cloudy, slow-warming spring then takes far longer to reach growing temperatures — a real climate signature, not a model artifact.

What this means

All four Western cities' divergence from the East/Midwest pattern is now explained by checkable climate mechanisms rather than left as an open question. Nothing in this check found a reason to distrust the model in any of the 11 cities it covered.

A 12th city, Pittsburgh, was added to the model in a later pass and has its own mow-interval and drought-year checks — but hasn't been run through this specific phenology comparison yet. Worth knowing if you're in western Pennsylvania: one rung less independently checked than the rest, for now.

See the full validation approach on how it works, or sign up and let it track your own city.