The weather of New York, every single day since 1869 — 158 years of the Central Park record. What it shows, read one day at a time: a warming city, its seasons coming apart, and the floods the next century is loading.
NOAA GHCN-Daily · station USW00094728 (NY City Central Park) · Jan 1, 1869 → Jun 19, 2026
One cell per day. Columns run January (left) to December (right); rows run 1869 (top) to today. By default each day is shaded by how far it ran above or below its seasonal normal — the slow reddening toward the bottom is New York warming. Switch the colouring to rainfall or snowfall and every wet day lights up. Hover any day.
Pick a year, then a yardstick to judge it by — each bar is that month against the chosen baseline. Switch the colouring above to rainfall or snowfall to compare those instead of temperature.
Each stripe is one year's average temperature, 1869 (left) to today (right), measured against the 1871–1900 average. Deep blue is a cool year; deep red is a hot one. Partial years are omitted.
Seasons drawn by temperature, not the calendar: winter is when the smoothed daily mean sits below the cold line, summer above the warm line, and spring and fall are the climbs between. Every year is one row, January (left) to December (right) — watch winter (blue) retreat from both ends and summer (red) swell toward the bottom. Drag the lines to set your own definition.
Days above the warm line.
Days below the cold line, wrapping the year-end.
Date the smoothed mean first clears the cold line. Lower = earlier.
Last spring freeze (low ≤32°F) to first fall freeze — the threshold-proof cross-check.
Pick any date and see its high and low in every year since 1869, with the all-time record for that date. Dashed lines mark the normal high and low.
The extremes of the Central Park record, and the trends underneath them.
How far each month ran above (red) or below (blue) its 1869–present normal. Pick a year, add a second to compare — El Niño and La Niña tagged.
Annual mean °F. Dashed line = 1871–1900 baseline.
Days the high reached 90°F or more.
Nights that dropped to freezing or below.
Muggy nights the low never fell below 75°F.
Central Park has logged precipitation every day since 1869. New York is measurably wetter than it was in the 1800s; snowfall, counted by winter, swings wildly year to year. Switch the matrix above to rainfall or snowfall to see every wet day, one at a time.
Monthly rain or snow against the 1869–present normal (grey line). Pick a year, add a second to compare, and see whether it fell in an El Niño or La Niña winter.
Total liquid precipitation per year. Dashed line = the 1871–1900 average.
Snowfall by winter season (July–June), labelled by the year it ends.
Days with at least 0.01″ of rain or melted snow.
Days that dropped an inch or more of rain — the downpours.
Heat is only half of comfort; the other half is moisture. Hourly observations — dew point, sky cover, rain rate — reach back to 1973 at LaGuardia and 2005 at Central Park. Dew point, not the humidity percentage, is the honest measure of mugginess: above 70°F the air feels oppressive; above 75°F, tropical.
Average June–August dew point. LaGuardia gives the long record; Central Park confirms it.
Days the air felt sticky. LaGuardia, 1973–2024.
Days the heat index reached 90°F. LaGuardia, 1973–2024.
Daytime sky cover, automated era only (2013+). Cloud reporting isn’t comparable across stations or to earlier years — a recent snapshot, not a trend.
Re-run New York's worst storms — or build your own — and watch two kinds of flooding: coastal surge and inland cloudburst. Warmer seas make storms wetter (Clausius-Clapeyron) and slower, so total rain = intensity × how long the storm lingers — a stalled storm is a Harvey. A simplified scenario model, calibrated to Sandy and Ida; sources below.
A weakening Gulf Stream piles ~0.5–1 ft of extra water on the Northeast coast; a blocking jet and warmer seas stall the storm — and a 10% slowdown doubles the rain.
Rough scenario estimates, calibrated to Sandy ($19B) and Ida ($0.36B). A harbor barrier stops surge but not rain — and as the sea rises, some areas flood at high tide with no storm at all.
Read the dispatch · The Barrier Stops the Wrong Disaster →Short data essays drawn from the Mercury record — every figure computed directly from the Central Park daily observations. Each opens as its own page.
Central Park's thermometer has climbed +4.5°F above its 1871–1900 baseline. The warmest year on record is 2023; the coldest, 1888 — and the gap between them is the story.
Read the dispatch →A Harvey stalled over New York drops ~82 inches in ~60 hours. The surge does ~$1B of damage. A $52.6B harbor barrier would stop the surge — and none of the rain.
Read the dispatch →A simplified Sandy redux floods South Street Seaport 8.3 feet deep and does ~$18B in damage — yet only 3 inches of rain fall. The surge does the work; the sky barely shows up.
Read the dispatch →Ida pushed still water to just 0.8 ft NAVD88 — a foot and a half below today's high tide. The flood arrived anyway, at 3.5 inches of rain per hour against sewers built for 1.75.
Read the dispatch →Central Park's rain gauge, read daily since 1869, now collects 50.5" a year — up 7.4" from the 1870s-90s — and it's arriving in fewer, harder bursts.
Read the dispatch →Replay Hurricane Sandy on the seas projected for 2100 — up to 5.4 ft higher — and the surge crests 16.7 ft, drowns Wall Street under 3.7 ft, and bills the city $43B.
Read the dispatch →Define seasons by warmth, not the calendar, and Central Park's spring now clears 40°F on Feb 20 — 27 days earlier than the 1869-1898 average of Mar 18.
Read the dispatch →