The Earth Gravity Model 2008 (EGM2008) is now some years old, and yet information on how to use it to calculate Earth's gravity remains obscure outside the field of geodesy. We describe the mathematics necessary to implement EGM2008, and use this to discuss several points of the model: its sensitivity to the number of spherical harmonics being summed, nuances and a trap for physicists and mathematicians in the normalisation it uses, and a comparison with other work. We conclude that one must not overestimate the precision shown by a global-fit model such as EGM2008.
