Not until you start thinking about the physics of the issue. Then you have an acutal cause (not just a correlation) for current warming-Carbon dioxide and other "greenhouse gasses" (released by humans) force the atmosphere to be warmer because the radiation emitted from earth is the the absorption wavelengths for these molecules. There are a few other things that matter for the energy budget of the earth too like sulfer from volcanoes and slight modulations of energy for the sun but using a causal model you can show that the its the additional gases that accuount for the current trend....
They use what are called "proxy" meausrements like cores from trees, corals and ice that show annual records of temperature. These data become more sparse (spatialy) as you go farther back in time but can often be overlaped with current "thermometor" temperature records for calibration.
Slashdot Mirror
Unless you live in Bangladesh
Not until you start thinking about the physics of the issue. Then you have an acutal cause (not just a correlation) for current warming-Carbon dioxide and other "greenhouse gasses" (released by humans) force the atmosphere to be warmer because the radiation emitted from earth is the the absorption wavelengths for these molecules. There are a few other things that matter for the energy budget of the earth too like sulfer from volcanoes and slight modulations of energy for the sun but using a causal model you can show that the its the additional gases that accuount for the current trend....
Its a variety of data sources and the error bars are estimated accordingly for each data type.
there are some good (slightly in-depth) discussions of the science behind global warming at realclimate.org if you are interested in learning more.
According to NASA the five warmest years on record are (in order) 2005 1998 2002 2003 2004
They use what are called "proxy" meausrements like cores from trees, corals and ice that show annual records of temperature. These data become more sparse (spatialy) as you go farther back in time but can often be overlaped with current "thermometor" temperature records for calibration.