GISS is the only temperature index which shows significant warming since the year 2000. The animation below highlights the primary thing they are doing wrong.
The blink comparator flashes between their 250 km (measured) and 1200 km (extrapolated) 2000-2009 trend maps. The region at the center of the green circle is showing 1-2 C of warming, even though the closest actual measurement near Herschel Island shows no warming.
Their algorithm creates an imaginary hot spot at the North Pole (2-4 C based on no actual data within 800 km) and then weights it as heavily as a closer measured temperature reading, in the next step.
They appear to be using at least two steps of extrapolation/interpolation – which compounds error. In other words, their entire 21st century warming story is based on a defective interpretation of the Arctic.
“A likely explanation for discrepancy in identification of the warmest year is the fact that the HadCRUT analysis excludes much of the Arctic ….. (whereas GISS) estimates temperature anomalies throughout most of the Arctic.”
The GISS article with this statement has disappeared from their web site, but it used to be at http://data.giss.nasa.gov/gistemp/paper/gistemp2010_draft0319.pdf.
Note that the only red spots on the map are a few thermometers north of Russia, and they are exaggerated in size due to the map projection.
When GISS is done torturing the real data, it looks like this – with massive temperature and geographical distortion near the pole.
The video below shows a world tour of the thermometer data – as an equal area projection. Not much warming on this planet.
GISS graphs are broken, there is no significant 21st century warming trend.