Skip Navigation
Back to PolicyBot

Nearly 90% of Temperature Stations Show Extreme Heat Bias, Study Says

August 1, 2009

Eighty-nine percent of official U.S. temperature measurement stations are corrupted by poor site selection that gives false warming signals, according to a new study by meteorologist Anthony Watts.

Eighty-nine percent of official U.S. temperature measurement stations are corrupted by poor site selection that gives false warming signals, according to a new study by meteorologist Anthony Watts.

According to the federal government’s own siting criteria, the corrupting influences at those stations create a margin of error larger than the entire asserted warming of the twentieth century.


Startling Degree of Error

The United Nations Intergovernmental Panel on Climate Change (IPCC), whose findings are often referred to by the news media and global warming alarmists, claims temperatures rose 0.74 degrees Celsius over the past century. However, Watts’ survey of official U.S. temperature stations shows 89 percent are so poorly located and maintained—according to U.S. National Oceanic and Atmospheric Administration (NOAA) published standards—they have an error margin of greater than 1.00 degree Celsius.

Worse, 69 percent have an error margin of greater than 2.00 degrees Celsius, and 11 percent have an error margin of greater than 5.00 degrees Celsius.

A variety of factors are responsible for the flawed data and resultant warming bias, the study reports.

Some stations have been corrupted by land-use changes and urban development nearby, resulting in an artificial warming signal over time. Other stations were moved from uncorrupted sites to corrupted ones for the convenience of persons reading the temperatures. Still others are and always have been located in places that are just plain ridiculous, the study shows.


Examples of Flawed Stations

A vivid example is the Forest Grove, Oregon official temperature station (see photo). As if the heat signature of the building, just 10 feet away, weren’t corrupting enough, the temperature station is located beside the exhaust grill of a window air conditioning unit.

The artificial heat signature from the unit’s exhaust did not exist to corrupt the data 50 to 75 years ago but results in an upward bias in recent temperature reports.


Similarly biased is the Aberdeen, Washington official temperature station (see photo). Not only is the station located beside an asphalt parking lot, but employees are encouraged to fire up their lighters and smoke cigarettes at the cigarette butt collector right beside the temperature station.

Another bizarre location choice is the Redding, California official temperature station (see photo). Consider what happens to the temperature reading every time that large light bulb, right above the thermometer, is turned on.

At the Tahoe City, California official temperature station, there’s a trash-burning barrel right beside the temperature station (see photo), with an exhaust vent set up to spew furnace-like heat just five feet away from the measuring device.

The before-and-after photographs of the official temperature station at Hopkinsville, Kentucky demonstrate how bias creeps in over time. Notice in the first photograph the temperature station was set up 30 years ago on a grassy lawn approximately six feet from the building. In the second photograph we see where it is today: adjacent to the building, right above an asphalt driveway and a barbeque. Hamburgers and hot dogs aren’t the only things heating up when somebody decides to cook on the grill!

“Changing the technology and locations of temperature stations and a blatant disregard for NOAA’s own rules about keeping sensors at least 100 feet away from heat sources and radiative surfaces have undoubtedly contaminated the U.S. temperature record,” Watts observes.


Changed Paint Guidelines

The stations with excessive temperature bias due to poor siting are only part of the story, Watts reports. Another source of heat bias was added in 1979 when NOAA changed its guidelines on temperature shelters to specify white latex paint instead of whitewash.

Watts found white latex paint on the shelters adds 0.55 degrees Fahrenheit (roughly 0.30 degrees Celsius) to the temperature readings when compared with whitewash. This heat bias alone is equal to roughly 45 percent of the asserted warming of the twentieth century.


Arbitrary Upward Adjustments

Watts also shows NOAA has introduced additional bias to the temperature record by arbitrarily adjusting temperature data ... upward.

Common sense dictates the nearly threefold growth in U.S. population during the twentieth century was accompanied by urbanization that created artificial heat islands corrupting the temperature data at even the most carefully located temperature station. Yet NOAA not only declines to adjust the raw temperature data downward to account for such inevitable heat bias, it instead adjusts the data upward, as if more asphalt, automobiles, electronics, and buildings create artificial cooling signatures instead of artificial heat signatures.

Anybody who lives near a large city knows urbanization creates artificial heat signatures, not cooling zones.

“Adjustments applied to ‘homogenize’ the data ... impart an even larger false warming trend to the data,” the study reports. “NOAA’s ‘adjustments,’ in other words, account for nearly one-half the agreed-upon rise in temperature during the twentieth century. The same adjustments are applied to the ... global temperature dataset.”


James M. Taylor (jtaylor@heartland.org) is a senior fellow of The Heartland Institute and managing editor of Environment & Climate News.

For more information ...

“Is the U.S. Surface Temperature Record Reliable?” by Anthony Watts: http://www.heartland.org/books/SurfaceStations.html.

Article Tags
Author
James Taylor is a senior fellow for environment and energy policy at The Heartland Institute.
jtaylor@heartland.org