Why Your Weather App Says 68 Degrees While Your Thermometer Says 74: Station Siting and the Nearest-Observation Problem

Have you ever walked outside, felt the heat hit you, and then looked at your phone in something close to disbelief? Your app says 68. The thermometer clipped to your patio umbrella says 74, and your body is voting with the thermometer.

Neither instrument is lying. They are measuring two different things in two different places, and the gap between them is one of the most under-explained facts in consumer weather.

Where The Number On Your Screen Actually Comes From

Almost every consumer weather app in the United States traces its current-conditions temperature back to the same backbone: the Automated Surface Observing System, or ASOS, jointly operated by the National Weather Service, the FAA, and the Department of Defense. There are roughly 900 ASOS sites nationally, plus around 500 AWOS stations run primarily for aviation.

Those sites exist because pilots need density altitude, visibility, and wind. They were not sited to describe your neighborhood, and they do not pretend to.

The consequence is structural. When your app says "Chicago, 68°," it usually means "O'Hare International Airport, 68°" — a specific patch of grass roughly 17 miles northwest of the Loop, surrounded by runway, not by the six-flat you live in.

What Station Siting Standards Actually Require

The World Meteorological Organization and the NWS publish siting criteria precisely because a thermometer's surroundings change what it reads. The standard is not arbitrary — each requirement removes a specific contaminating heat source.

The core requirements for a properly sited temperature sensor include but are not limited to:

• Height above ground. The sensor sits between 1.25 and 2 meters — about 4 to 6.5 feet — above the surface. Air one foot above asphalt on a sunny afternoon can run 15 or more degrees warmer than air at chest height.
• Natural low vegetation. The ground beneath should be short grass or the natural surface of the area, not gravel, concrete, or bare soil. Grass transpires and holds far less heat than a paved apron.
• Distance from obstructions. The sensor should be at least four times the height of any nearby obstruction away from it, and well clear of buildings, parking lots, and heat exhaust.
• Radiation shielding. The thermometer lives inside a louvered or aspirated shield that blocks direct sun and reflected radiation while letting air move freely across the sensing element.
• Aspiration. Many modern stations force air across the sensor with a fan, so the reading reflects the moving air mass rather than a pocket of stagnant, sun-warmed air.

All of these exist to isolate one measurement: the temperature of the free air, uncontaminated by the surfaces near it. That is a scientifically useful number and, on a hot afternoon, a number that describes almost nobody's actual experience.

Why Your Patio Thermometer Reads High

A thermometer in the sun is not measuring air temperature at all. It is measuring its own temperature, which is the sum of air temperature plus absorbed shortwave radiation plus longwave radiation re-emitted by every warm surface in view.

An unshielded sensor in direct September sun routinely reads 10 to 20 degrees above the shielded value, and on a still day with no wind to carry heat away, the high end is common. Move it into shade and much of that error vanishes, though not all of it.

Shade is not the end of the problem. A thermometer hung on a south-facing brick wall or mounted three feet above a composite deck is bathed in longwave radiation from surfaces that have been storing solar energy since ten in the morning.

Thermal mass is the mechanism. Brick, concrete, and asphalt have high heat capacity and low albedo, so they absorb energy all day and release it slowly into the evening — which is why a courtyard can still read 6 degrees above the airport at nine at night.

The Urban Heat Island Offset

Everything above scales up. What a brick wall does to one thermometer, a city does to itself, and the effect has a name and a measured magnitude.

Urban heat island intensity — the temperature difference between a city core and its rural surroundings — typically runs 1 to 7 degrees Fahrenheit in the daytime and can reach 22 degrees at night in large cities under clear, calm conditions. The nighttime peak is the counterintuitive part, and it is the part most consumer apps handle worst.

The reason is release timing, not absorption. During the day, cities and countryside both heat; after sunset, rural grass and soil radiate their heat to the sky quickly, while urban canyons trap longwave radiation between building faces and keep releasing stored energy for hours.

This is why the direction of your error flips with the clock. If your nearest station is a suburban airport and you live downtown, your app likely runs cool for you on a summer night — and the gap widens the calmer and clearer the night is.

How Far Away Is "Your" Observation, Really?

This is the question almost no app answers, and it is the one that explains the most variance. The observation labeled with your city name may be 3 miles away or it may be 25.

Some rough anchors worth knowing for your own location:

• Major metros. The default station is usually the primary airport. For New York that is often LaGuardia or JFK — both coastal, both cooler than Midtown on a summer afternoon by several degrees because of the sea breeze.
• Mid-size cities. A single regional airport frequently serves a 20-mile radius, meaning one sensor represents terrain, elevation, and land cover that vary substantially across that circle.
• Mountain and coastal towns. The nearest station may sit 1,500 feet below you or three miles inland. At a typical environmental lapse rate near 3.5°F per 1,000 feet, elevation alone can account for a 5-degree difference before any other factor.
• Rural areas. The gap to the nearest official observation can exceed 30 miles, at which point the app is interpolating from a model grid rather than reporting a measurement at all.

All of these add up to the same conclusion: the label on the number is a place name, and the number is a point measurement. Those are not the same claim, and the difference between them is where your disbelief on the patio comes from.

What's more, many apps do not show you a raw observation at all. They show a model analysis — a gridded estimate blending observations, satellite data, and model output — which can differ from the nearest actual station by a degree or two in either direction. We wrote about how those grids are built in our piece on how forecast models actually work.

Comparing The Measurements Side By Side

It helps to see the four common readings laid next to each other, because each one is answering a different question.

Reading	What it measures	Typical offset from ASOS	Useful for
ASOS / airport sensor	Shielded free-air temperature, 5 ft over grass	Baseline (0°)	Record-keeping, aviation, forecast verification
Backyard thermometer in shade	Air temperature plus local thermal mass	+1 to +5°F	Your actual microclimate
Backyard thermometer in sun	The thermometer's own temperature	+10 to +20°F	Almost nothing — this is instrument error
Car dashboard readout	Air near hot asphalt, sensor behind grille	+5 to +15°F while parked or idling	Rough trend once moving at speed

Note that only one row in that table is wrong in the strict sense. The others are all correct measurements of different things, which is exactly the problem.

Why This Matters More Than It Sounds

A 6-degree discrepancy is not a trivia item. It is the difference between a jacket and no jacket, between a safe run and heat stress, and between a frost that kills your tomatoes and one that does not.

Frost is the sharpest example. Cold air drains downhill and pools in low spots, so a valley-floor garden can hit 30°F while the station on the ridge reports 36 — a phenomenon we cover in detail in our guide to reading a frost forecast.

The dressing consequence is just as real. What your body registers depends on humidity, wind, and radiation as much as on the thermometer, which is the whole argument behind apparent temperature and why "feels-like" exists.

And because moisture drives so much of that felt experience, the number to check alongside temperature is usually dew point rather than relative humidity. Dew point is a station-independent measure of absolute moisture, which makes it travel across the siting gap far better than temperature does.

What Steps Should I Take To Get A Number That Describes My Yard?

You cannot move the airport, but you can build a local correction that is good enough to act on. Here's a list of the steps worth taking:

• Find your actual station. Look up which observation site your app is reading — most will name it in a details view, and the NWS station list will confirm the distance and elevation. Knowing it is O'Hare and not your block changes how you read the number.
• Fix your own sensor first. Move the thermometer into full shade, at least 4 feet off the ground, away from walls and pavement, with air moving freely around it. A cheap louvered shield costs less than dinner and removes most of the error.
• Log the offset for two weeks. Record your reading and the station reading at the same times — say 7 a.m., 2 p.m., and 9 p.m. Your personal offset is rarely a single number; it is usually small in the morning and large in the late afternoon.
• Apply the offset by time of day. If you consistently run 5 degrees warm at 3 p.m. and 1 degree warm at dawn, that is your correction. Add it to the forecast high, not to the current reading alone.
• Note the sky and wind conditions. Offsets balloon on clear, calm days and nearly vanish under overcast, windy conditions, because wind mixes the air and clouds suppress both incoming and outgoing radiation.

Two weeks of that produces something genuinely useful: a translation layer between the official record and the place you live. Keep in mind that the offset is seasonal, so the summer correction will not hold in January.

Personal Weather Stations And Their Own Problem

Networks like Weather Underground's PWS mesh and the Citizen Weather Observer Program have put tens of thousands of private sensors into neighborhoods, and some apps now blend them in. On paper this solves everything — a sensor two blocks away is far more relevant than one twenty miles away.

In practice, the siting standards that make ASOS boring are exactly what PWS data usually lacks. A large share of home stations are mounted on roofs, fences, or south-facing walls, which reintroduces every error the shield and the grass were designed to remove.

Our position is that a well-sited neighborhood station beats a distant airport, and a badly-sited neighborhood station is worse than either. The deciding factor is not proximity but installation, and proximity is the only one of those two an app can see.

How We Handle It

We think the honest move is to name the station. When a brief tells you the afternoon will reach the mid-70s, it should be clear whether that describes the airport, the model grid, or a corrected estimate for where you actually are.

We also think the raw number is the least interesting part of the forecast. The decision you are about to make — jacket or no jacket, shoot at 6:40 or 7:10, cover the tomatoes or leave them — depends on translation, not on a digit, which is the argument behind how we write a daily brief.

Remember that the station is not the weather. It is a carefully-controlled sample of the weather, taken somewhere specific, and the distance between that somewhere and your patio is the whole story of the missing six degrees.

Common Questions

Why do weather apps use airport sensors instead of city sensors?

Airport stations exist for aviation safety and are federally funded, maintained, and quality-controlled to a consistent standard. That gives them reliable uptime and a long, comparable record. City-center sensors are rarer, less standardized, and often privately owned, so apps default to the ASOS network for consistency rather than for relevance to your street.

Does the same siting gap affect the forecast high, or just the current temperature?

It affects both, because forecast highs are verified against the same station. A model predicting a 90-degree high is predicting it for the station's shielded sensor over grass. Your sun-exposed courtyard will likely exceed that, and the gap is typically largest on clear, calm afternoons when mixing is weakest.

Why is my car's temperature readout always wrong in a parking lot?

The sensor usually sits behind the front grille, inches above pavement that has been absorbing sun all day, with no airflow while stopped. It reads radiated heat from the asphalt and the engine bay. After ten minutes of highway driving the forced airflow brings it much closer to true air temperature.

Should I trust a nearby personal weather station over the airport reading?

Only if you can see how it is mounted. A shielded sensor over grass at chest height beats a distant airport easily. A sensor screwed to a sunlit wall or roof deck will run several degrees hot on clear afternoons and will mislead you more than the twenty-mile-away official number would.

How much does elevation change the temperature between me and the station?

The environmental lapse rate averages about 3.5°F per 1,000 feet of elevation gain, though it varies with moisture and stability. A station 1,500 feet below your house accounts for roughly 5 degrees of difference on its own — before land cover, wind exposure, or cold-air drainage enter the picture.

Why is the urban heat island effect strongest at night rather than in the afternoon?

Both city and countryside absorb solar energy during the day. After sunset, rural soil and vegetation radiate heat to the open sky quickly, while urban street canyons trap longwave radiation between building faces and keep releasing stored energy from concrete and asphalt for hours, widening the gap.