Here's something I keep thinking about: Thoreau spent seven springs walking the woods around Concord, Massachusetts, writing down when more than 300 species first bloomed. He was curious. That's it. He wanted to know what was happening in the place where he lived.
Over a century later, biologist Richard Primack retraced those routes and found that 43 common wildflowers now bloom an average of 10 days earlier. Some shifted three weeks. Of 21 orchid species Thoreau found, roughly six remain — the ones that couldn't track the warming vanished.
A Wisconsin ice business tracked Lake Mendota's freeze and thaw starting in 1855. The Park Service logged D.C. cherry blossom peaks from 1921. The USA National Phenology Network now holds over 40 million observations, many from volunteers noting what blooms in their yards.
What gets me is the accidental nature of it. People recording the ordinary rhythms of where they lived, year after year, with no idea they were assembling climate evidence. Their attention outlasted them. And now we can read it.
Concord, MA — 1851 to present Thoreau's wildflower records vs. modern observation: first bloom averages 10 days earlier. Flowers advance 3.2 days per 1°C of spring warming. Concord springs now ~3.3°C warmer than the 1850s.
Lake Mendota, WI — 1855 to present Longest continuous lake ice record in North America. Before 2000, the lake typically froze in December. Since then, a December freeze has happened less than half the time. Ice cover projected to shrink ~11 days per 1°C of warming.
D.C. Cherry Blossoms — 1921 to present Peak bloom has shifted roughly 8 days earlier. Arrived before the April 3 average in 16 of the last 20 years. Earlier than average for the 8th straight year in 2026.
USA-NPN — 2009 to present 40+ million records from volunteers and scientists. Spring arriving early in 3 out of 4 national wildlife refuges. Eastern U.S. growing season extended 4% over three decades.