Rising or Falling Barometric Pressure: What It Means

Every barometer gives you a number, and the number is the least useful thing about it. Whether your station reads 1008 or 1019 hPa says little on its own — what matters is which way the pressure is moving, and how fast. Pressure tendency has been the backbone of short-range forecasting since long before satellites existed, and it is still the most predictive single reading your weather station produces.

Why does the pressure trend matter more than the number?

The trend matters more because weather is driven by pressure changes, not pressure values. A falling barometer means air is being removed from the column above you — a low-pressure system or front is approaching. A rising barometer means denser, more stable air is moving in. The same 1010 hPa can precede a storm or follow one.

The absolute value also depends on season, latitude, and how well your elevation correction is set, so comparing your number to a friend's a hundred kilometers away tells you little. Professional observations handle this the same way: synoptic reports include a 3-hour "pressure tendency" — the amount and character of the change — precisely because that is the part with forecasting value. Treat the current reading as context and the 3-hour change as the actual signal.

How fast does pressure need to fall to signal a storm?

A fall of more than 3–4 hPa in three hours signals an approaching storm or front, usually arriving within 12–24 hours. Falls above 6 hPa in three hours are rare and point to severe weather — an intense low, a squall line, or a rapidly deepening system. Slower changes of 1–3 hPa mean gradual change.

These thresholds are rules of thumb, not laws. A modest fall in a region where storms are rare means more than the same fall somewhere lows pass weekly. The character of the change matters too: a steady, relentless fall is more significant than a wobble that recovers an hour later.

What is a normal barometric pressure reading?

Normal sea-level pressure ranges between about 980 and 1040 hPa, centered on the standard atmosphere of 1013.25 hPa. Deep winter storms can dip below 960 hPa, and strong winter highs can exceed 1050 hPa. The all-time recorded extremes are roughly 870 hPa (Typhoon Tip, 1979) and 1083.8 hPa (Agata, Siberia, 1968).

Within that range, high pressure generally brings settled, drier weather and low pressure brings unsettled, wetter weather — but the boundaries are soft, which is exactly why the trend beats the number. NOAA's JetStream explainer on air pressure covers the physics in more depth. One caveat before you compare your reading to any of these figures: they only make sense once your pressure has been corrected to sea level, which brings us to the most common point of confusion for station owners.

Why doesn't your station's raw pressure match the weather map?

Weather maps show sea-level pressure, while your barometer sensor measures station pressure — the actual air pressure at your elevation. Pressure drops roughly 1 hPa for every 8 meters of altitude, so a sensor at 200 m reads about 24 hPa lower than the map. Your console corrects for elevation to make readings comparable.

This is the chronic point of confusion with personal weather stations. The raw sensor value is not wrong; it is just answering a different question. Davis consoles and WeatherLink apply the sea-level reduction using the elevation you entered during setup, which is why an accurate elevation setting matters — get it wrong by 40 m and every reading is offset by about 5 hPa. Usefully, the trend is immune to this: a constant offset shifts the whole curve but leaves the rises and falls intact. If you want the fuller picture of what each instrument on your station actually measures, see our guide to how weather station sensors work.

What do classic pressure patterns look like?

Two patterns cover most mid-latitude weather. A slow, steady fall with backing winds and thickening high cloud signals a warm front, with rain typically 12–24 hours away. A sharp drop followed by an equally sharp rise, a veering wind, and a burst of gusts marks a cold front or squall line passing through.

The warm-front pattern unfolds gradually: pressure eases down over 6–12 hours, the wind backs (shifts counterclockwise, say from west around to south), and wispy cirrus thickens into a gray altostratus sheet. The cold-front pattern is compressed into an hour or two: the barometer bottoms out just as the front arrives, then jumps as colder, denser air floods in — often with a spike in gusts well above the sustained wind (the difference is explained in our post on wind gusts vs sustained wind). A third pattern is worth knowing in the cold season: a steady rise into a strong high means clear, calm nights, which is precisely the setup for radiation frost — see how to predict frost with your station.

How accurate are the forecast icons on a weather console?

Console forecast icons are mostly pressure-trend algorithms, and they are right roughly 70–75% of the time over the next 12–24 hours. Davis consoles combine the pressure tendency with wind direction, humidity, and time of year, in the tradition of the old Zambretti forecaster. Beyond about 24 hours their skill drops off quickly.

That 70–75% sounds unimpressive until you remember the icon is generated from a single location's sensors with no satellite, radar, or model data. The Zambretti forecaster — a 1915 slide-rule that mapped pressure value, trend, and wind direction to one of 26 forecasts — achieved similar results, and most console algorithms are its descendants. Where they fail is the same place any pressure-only method fails: fast-moving systems, convective pop-up showers, and local effects the barometer cannot see. Treat the icon as a sanity check on your own reading of the trend, not a forecast.

How do you read the pressure chart on your own site?

Read the pressure chart on two timescales. Use a 24-hour view to judge the tendency — compare now against three hours ago and apply the thresholds table above. Use a 7-day view to see whole systems parade through: each low is a valley, each high a broad dome, and fronts show as steep slopes between them.

On a Pro Weather site, the pressure chart does exactly this at the 24h and 7d ranges, drawn from your station's own barometer every 10 minutes. Two things to expect: a small semi-diurnal wiggle of 1–2 hPa (an atmospheric tide, most visible in calm weather), and the occasional steep ramp that is worth acting on. Pressure alerts are rare on any platform, but you can set alerts for what a falling barometer predicts — wind gusts and rain rate — so the storm the trend announced does not catch you off guard.

Common questions

What is considered a rapid pressure drop?

A fall of more than about 1 hPa per hour — 3–4 hPa over three hours — counts as rapid and usually signals an approaching front or storm within 12–24 hours. A fall exceeding 6 hPa in three hours is very rapid and associated with severe weather: deep lows, squall lines, and damaging winds.

What barometer reading means a storm is coming?

No single reading does — the fall is the signal, not the value. That said, a barometer already below about 1000 hPa and still falling fast deserves attention, and readings under 980 hPa indicate a deep low is overhead or nearby. A high reading falling rapidly means deterioration is coming even though the current number looks benign.

What's the difference between station and sea-level pressure?

Station pressure is what the sensor physically measures at your elevation. Sea-level pressure is that value corrected to what it would be at sea level, so stations at different altitudes can be compared on one map. The correction is roughly 1 hPa per 8 m of elevation. Consoles and weather sites display the sea-level value.

Why does my pressure differ from the airport's?

Small differences of 1–2 hPa are normal: you are some distance apart, readings are taken at slightly different times, and aviation reports use a slightly different reduction method (the altimeter setting). A large, constant offset almost always means your station's elevation setting is wrong — fix that and the readings should fall in line.