Althoughitiswellknownthatcoseismicgravitychangestakeplaceduringanearthquake,previousresearchhasnotyieldedconvincingevidencedemonstratingthatsignificantgravitychangesoccurbeforelargeearthquakes.Furthermore,evenifwesuspectthatgravitychangesoccurbeforelargeearthquakes,wehaveyettodemonstratehowtoconsistentlyobservethesechangesforusefulearthquakeforecastthatwouldbringbenefitstosociety.Weanalyzedgroundgravitysurveydataobtainedin1998,2000,2002,and2005atstationsoftheCrustalMovementObservationNetworkofChina(CMONOC)andexaminedgravitychangesbe-foretheoccurrenceofninelarge(Ms≥6.8)earthquakesthatrupturedwithinornearmainlandChinaandTaiwanfromNovember2001toAugust2008.Resultsfromthisanalysisshowthatsignificantgravitychangesoccurredacrossalargeregionbeforeeachoftheseninelargeearthquakes,andthesechangesweredetectedbyrepeatedgroundgravitysurveysthroughCMONOC.Althoughthesegravitychangesweresignificant,moreresearchisneededtoinvestigatewhetherthesegravitychangescouldbeviewedasprecursorsoflargeearthquakes.Limitationsanduncertaintiesinthedataincludesparsenessofthegravitymonitoringnetwork,longtimeintervalsbetweenconsecutivegravitysurveys,inevitablemeasurementerrors,hydrologicaleffectsongravity,andeffectsofverticalcrustalmovementsongravity.Basedontheseobservations,wemakeseveralrecommendationsaboutpossiblefuturedirec-tionsinearthquake-relatedresearchusinggravitymonitoringdata.
