Actually it should be arised by the 1D peak broadening. I checked some materials, and got three probabilities.
From Wuthrich's book: if the atom is kinetically stable,it should be related to intrisic T2 or pure T2; if the atom is active, there exists exchange processes,which cause line broadening; experimental artifacts, say the inhomogeneity of the static field Bo, which can be improved by shimming.
In protein NMR study, CPMG is usually employed to investigate the dynamics. I recommend two articles: Science,V.312,14 April,2006,Pp224-227; JBNMR, V.31,No.4,April,2005,Pp337-342. They are inspiring.
(Sorry, there is no chinese system in our lab computer!)