The ground at the station reading 7 min ago
A sensor at the foot of the station mast measures the air right at the tips of the grass blades and the soil at 2, 4 and 8 inches deep, every 5 minutes. On clear, calm nights the grass cools faster than the air the station measures 5 feet up, so frost forms on the grass first; the soil gives back the heat it stored by day, more slowly the deeper you go. The grass probe was painted white and moved to its permanent mount on October 18, 2026; earlier grass readings are less reliable.
The grass is 13.0° colder than the air the station measures: the ground is losing heat to the sky faster than the air above it.
Snow on the ground
Bare. From the forecast models (no recent report).
The station's owner reports bare, patchy, covered or deep ground (and depth from snow stakes) when snow is around. Between reports the estimate is carried forward: new snow from the forecast models, melting from the station's own temperatures, settling. A report counts for less as it ages, and after ten days the models' snow depth takes over. A buried grass probe (one that stops following the sky, or sits at 32° while the air is well above freezing) also counts as snow on the ground.
The last 48 hours
Shaded bands are night (sunset to sunrise). The dashed line is the station's air, the dotted line its dew point, and the thin line at 32° marks freezing. Soil swings get smaller and later the deeper you go: the 8-inch soil barely notices one day's weather.
Night cooling last 24 hours
The grass got down to 27.9° at 3:50 AM, when the station's air was 37.0°. At most the grass ran 15.3° below the air (6:10 PM).
It was at or below freezing for 5 h 10 min, starting 1:30 AM. The grass reached the dew point, so frost formed.
Grass minus the station's air. Below zero (blue): the ground is losing heat to a clear sky faster than the air. Above zero (orange): sunshine is warming the ground. Big blue dips mean calm, clear nights, the ones our forecast finds hardest.
The column every level we measure, now
Soil profile now and 12 hours ago
The ground's season heat stored in the top foot, by day
The top foot of soil is losing heat: it has cooled 0.8° a day on average over the last 7 days, about 0.30 MJ per square metre a day. This is the ground's own autumn. It turned on November 13. The 8 inch probe is falling 0.7° a day.
Each bar is one day's change in the mean temperature of the top foot of soil, from the 2, 4 and 8 inch probes standing for the layers they sit in, with the daily wave averaged out. Bars below the line are days the ground gave up heat; bars above are days it stored heat. The sign changes once in autumn and once in spring: those two dates are this yard's own seasons, set by the sun and the air masses, with the soil following behind. The energy figure uses the heat capacity of this silt loam from the Studio's soil column fit, 2.27 MJ per cubic metre per degree Celsius.
Records since October 17, 2026 · 34 nights
Recent nights
| Night of | Grass low | Air low | Grass most below the air | At 32° or less |
|---|---|---|---|---|
| Thursday, November 19 (part) | 36.8° | 50.4° | 15.3° | - |
| Wednesday, November 18 | 27.9° | 37.0° | 14.9° | 5 h 10 min frost |
| Tuesday, November 17 | 24.1° | 33.3° | 17.8° | 13 h 50 min frost |
| Monday, November 16 | 22.7° | 29.3° | 14.1° | 14 h 10 min frost |
| Sunday, November 15 | 22.6° | 27.3° | 11.0° | 13 h frost |
| Saturday, November 14 | 29.4° | 35.3° | 6.8° | 1 h 50 min frost |
| Friday, November 13 | 26.4° | 29.3° | 7.2° | 10 h 50 min frost |
| Thursday, November 12 | 36.9° | 37.8° | 4.0° | - |
| Wednesday, November 11 | 39.8° | 43.9° | 7.5° | - |
| Tuesday, November 10 | 29.7° | 34.2° | 16.8° | 4 h 40 min frost |
Soil extremes
| Depth | Coldest | Warmest | ||
|---|---|---|---|---|
| 2 in | 39.3° | November 17 | 61.7° | November 4 |
| 4 in | 41.3° | November 18 | 59.6° | November 4 |
| 8 in | 43.4° | November 18 | 58.8° | November 4 |
A night runs from noon to noon and is named for its evening. "(part)" marks a night with gaps in the readings, such as the night the sensor went in. "Frost" means the grass reached both 32° and the dew point, so ice actually formed. Readings taken while testing the probes are left out. The station's air records (all-time extremes, monthly climate) are on Station records; normals and freeze odds: Climate at the mast.
Sensor health
Grass probe calibration: looks right. in 51.2 hours of well-mixed air (overcast day and windy night) the grass probe read -1.2 F against the station's air, within 2 F. When the wind stirs the air and there is little sun or clear sky to pull them apart, the probe and the station's air should read within about 2° of each other; on calm, clear nights a large gap is real.
The sensor wakes every 5 minutes, reads its four probes and radios them about 500 feet to a small receiver in the house. The battery chart is by hour: a single lithium cell reads about 4.2 V full and should be recharged below about 3.5 V. Radio signal is in dBm (closer to zero is stronger); the margin is how far the signal sits above the background noise. The station computer sends an alert if readings stop or the battery runs low.