It doesn't. You are looking at the noise part of the signal and the noise. This is why it can take 15-20 years for the warming to be statistically significant. The 15-20 years for warming to be statistically significant does not mean that the warming has slowed or stopped. That is the amount of data that is required to resolve the signal from the noise.
Only a few degrees of temperature change results in this.
http://en.wikipedia.org/wiki/File:Vostok...
And if the changes due to weather were more important than the smaller temperature changes due to climate change, they would need to show time scales of a few days rather than hundreds of thousands of years to show any change.
There are a number of interdependent factors involved in deciding what intrumentation to use to measure surface air temperature.
1. What air (where and at/over what times) do you want to index the temperature of? What is it beyond merely the instrument readings that you are trying to actually measure?
2. What influences might 'contaminate' your instrument readings?
3. What other readings do you need your index and/or changes in your index to be comparable with?
4. How much time and money do you have?
Temperature readings have historically and geographically been made with various answers to these questions and differing technical understandings. A huge amount of time and money has been spent on obtaining information about changes in the temperature of air near the surface, currently and historically. Key methods involved in this are
a. Utilise a variety of instrumentations in different conditions, according to consideration 2. above
b. Analyse historical and contemporary instrument readings which are/were obtained primarily for other purposes.
c. Make indices from different intrumentations comparable by setting them up in the same place and seeing how they respond slightly differently to the conditions. Contune/repeat for a variety of conditions in which the instrumentations have been implemented.
That's part of why you will usually see references to anomalies, rather than absolute temperatures.
While the temperature 5 feet off the ground may be different from the temperature 10 feet off the ground, the temperature *at 5 feet off the ground* will tend to be subject to the same forces all the time. So, you can get a valid result, that will tend to be the same under equivalent conditions, even if you'd get a completely different (though equivalently valid) measurement 10 feet off the ground.
We're not so much trying to measure the *absolute* temperature as we are trying to measure the *change* in temperature.
NASA numbers are an average of thousands of sampling points. take some statistics courses before claiming problems.
and where is this quote from- please provide a source.
Global Warming is based on temperature readings and the established Global average temperatures. Extreme fluctuations in temperature seem to confuse NASA also. Here's an explanation from them :
"What exactly do we mean by SAT?"
"I doubt that there is a general agreement how to answer this question. Even at the same location, the temperature near the ground may be very different from the temperature 5 ft above the ground and different again from 10 ft or 50 ft above the ground. Particularly in the presence of vegetation (say in a rain forest), the temperature above the vegetation may be very different from the temperature below the top of the vegetation. A reasonable suggestion might be to use the average temperature of the first 50 ft of air either above ground or above the top of the vegetation. To measure SAT we have to agree on what it is and, as far as I know, no such standard has been suggested or generally adopted. Even if the 50 ft standard were adopted, I cannot imagine that a weather station would build a 50 ft stack of thermometers to be able to find the true SAT at its location."
What does this say about how the records vary by as much as 0.5C in one month when it comes to these past temperature records : http://data.giss.nasa.gov/gistemp/tabledata_v3/GLB.Ts+dSST.txt
