Reading the two rows

Each hour carries two figures, and the whole value of the strip is that they disagree.

The relative humidity on top is the number people quote. It will swing enormously — 95% before dawn, 45% by mid-afternoon — and almost none of that swing is water arriving or leaving.

The dew point underneath is the water. Watch it across the same 24 hours and it usually moves a few degrees at most, unless a genuinely different air mass arrives. When it does jump, something real has happened: a front, a wind shift, air off the sea instead of off the land.

What each pattern means

  • Flat dew point, swinging humidity — an ordinary day. The air is not changing; the thermometer is. Nothing to plan around beyond the usual dawn dampness.
  • Dew point climbing through the day — humid air is moving in. Evenings will be heavier than the forecast high suggests, and overnight lows will not fall far.
  • Dew point and temperature converging overnight — the air is heading for saturation. That is dew on every surface, and if they meet, fog.
  • Dew point falling sharply — drier air arriving, usually behind a front. Good drying, big temperature swings, and static indoors a day or two later.

The overnight floor

The most practical thing in the strip is that the dew point sets a floor under the overnight temperature. Air cannot cool much past the point where it saturates, because condensing water releases heat almost as fast as the ground loses it.

So if tomorrow's dew point is 68°F, the night will not be a cool one, whatever the sky does. And if it is 45°F on a clear calm night, the low can go a long way down — which is the basis for working out whether there will be a frost when the forecast low is borderline.

Where the forecast comes from

Open-Meteo, read directly from your browser. The derived figures — the comfort bands, the condensation margin — are computed here with the same implementation the Hygrometer app uses, though the app's underlying observations come from Apple Weather rather than this source. Small differences between the two are expected and are about the weather model, not about the arithmetic.