The SPPI paper says that land weather station data is broken, temperatures are being reported as much higher than they should be, it's all the urban heat island, etc.
Yet the satellite data strongly corresponds with the land based temperatures and also shows an increase. See http://www.skepticalscience.com/pics/surface_satellite_compa... (or read the damn blog post properly)
In addition to this, both urban and rural stations show the same levels of temperature rise: http://www.skepticalscience.com/images/jones_london.gif
Which brings us back to my original question: If we take what you are saying about the land based temperatures at face value, how do you explain a) the extremely strong correlation between the satellite data and the land based data, and b) the extremely strong correlation in the trends between different stations if the data is flawed?