Read the Tale and a description of the figure to the left. I believe the apparent 0.8ºC increase in Global Temperature over the past 150 years is due to three major causes and one minor one, as indicated by the parts of the "tiger".
This posting is about DATA BIAS which I think is responsible for about 30% of the apparent warming. In other words, 0.2ºC to 0.3ºC of the apparent 0.8ºC temperature increase is not actual warming.
(However, the remaining 0.5ºC to 0.6ºC is actual warming, and needs to be addressed.)
NASA/NOAA GISS MODIFICATION OF THEIR OWN BASE DATA
The figure below is a "blink comparator" of the official NASA/NOAA GISS (Goddard Institute for Space Studies) USA temperature anomaly (from http://wattsupwiththat.com/2008/11/14/the-evolution-of-the-giss-temperature-product/. Credit for the graphs goes to http://zapruder.nl/images/uploads/screenhunter3qk7.gif.)
More about the above graph near the end of this posting!
CLASSES OF OFFICIAL MEASUREMENT SITES
NASA/NOAA specifies measurement sites in five classes, with the best at least 100 m (over 300 ft) from any source of artificial heating or land development and the worst located right on an occupied building. According to a 2009 survey, only about 3% of official sites in the US are at Class 1. About 8% are in Class 2, at least 30 m from a source of artificial heat. About 20% are in Class 3, between 10 and 30 m. The remaining stations are closer than 10 m to an artificial heat source (58%) or right on a heat source (11%).
Thus, only about 11% are in the best two classes, reasonably distant from artificial sources of heat, while 69% are in the worst two classes, easily affected by nearby heat sources. Over 2/3rds of the official reporting stations in the US are close enough to artificial heating sources to be affected. We have no idea if the situation is better for foreign stations, but it is likely even worse!
Of course, since Global Warming has to do with changes in temperature, if a station has been at the same location for decades, any change in reported temperature should be consistent with actual trends in that area, right?
WRONG! Stations in urban areas, even if they have been in the exact same place, have been affected by lifestyle changes, such as installation of air conditioning in buildings that had none fifty years ago, more auto and truck traffic, and construction of nearby buildings. But, many stations have been moved from time to time and thus have not been in the same place all this time, and most have been affected and encroached by civilization and changes in land use.
Why are the stations so close to artificial heat sources? Well, fifty or more years ago, all the readings were taken manually by volunteer observers twice a day. They were not about to walk the length of a football field to do so. Even as automatic reporting stations were introduced, the stations had to be close to buildings so the telephone and power wires could be run to them.
ANECDOTAL EVIDENCE (Source)
"One swallow doesn't make a spring." (Aristotle). Even two doesn't prove it is spring. Likewise, the following anecdotal evidence does not prove the case for measurement bias. However, human beings seem to be moved more by specific examples than by statistical evidence. So have a look at these two stations about 50 miles apart in north-central California with very different temperature plots.
Marysville, CA, around 50 miles away, has seen considerable encroachment by development. The shelter that houses the MMTS temperature sensor is close to air conditioning exhaust fans and you can see their effect in the 2ºC mean temperature rise from 1940 to the present.
Sadly, the second example is far more representative of reporting stations than the first.
How could two temperature measurement stations, only 50 miles apart, experience such large differences in annual mean temperature trends? The only explanation is measurement bias due to encroachment of development at the second station.
THAT IS WHY NASA/NOAA GISS CORRECTS THE RAW DATA
Indeed, NASA/NOAA has retroactively corrected the data to account for measurement error. Here is that "blink comparitor" again.
Remember, this is OFFICIAL published data from NASA/NOAA GISS in 1999 and then in 2008. They have adjusted the data by about 0.3ºC which is more than 30% of the apparent Global Warming over the past 150 years.
Since the data has been biased by proximity to artificial heating sources which tend to increase over time, you would expect them to adjust the data to show less actual warming in later years. However, they seem to have made the "correction" in the wrong direction!
Check the blinking graphs! For data before 1970, they have REDUCED the temperatures by up to 0.1ºC. For the years after 1970, they have INCREASED the temperatures by up to 0.2ºC. This retroactive "correction" has added 0.3ºC o the apparent warming.
I have only scratched the surface on reasons to suspect data bias. There are dozens of photos showing official temperature measurement sites that are clearly too close to artificial heat sources that have obviously encroached relatively recently. There is evidence that painting the sensor enclosures with modern, longer-lasting latex paint rather than the old method using whitewash has increased temperature readings in the warm part of the summer.
What if I suggested that all official sites in Class #4 and Class #5 should be abandoned and replaced by new sites that met the requirements of Class #1 (or at least Class #2)? Well, the Global Warming alarmists would be up in arms, and for good reason, because that would wipe out most of the apparent temperature increase.
I agree with them! Since much of the data from years ago came from stations that were in very close proximity to human development, the old data is biased towards higher readings.
On the other hand, I think everyone should agree that encroachment of measurement sites by human development, especially over the past 40 years as air conditioning and vehicular traffic became more common, has biased recent data towards higher readings.
I don't have the power to correct the measurement stations that have been encroached, nor do I have the base data to make exact corrections. However, I think it is clear that a substantial portion of the apparent Global Warming increase is definitiely due to data bias. I estimate this bias as about 30% (+ 0.2ºC to +0.3ºC). Does anybody agree or disagree with that estimate? On what basis?
(However, I hasten to repeat, the remaining 0.5ºC to 0.6ºC is actual warming, and needs to be addressed.)