INTRODUCTIONOverthepasttenyears,national,pro vincialandlocalagencieshavecarriedoutstudie stoimproveandcomprehensivelyexploittheredso ilresourceofsouthernChina,andthemanagingand utilizingpracticesofredsoilresourcesweredev elopedbytheeducationalandresearchsectors.Wi ththeeffortsandfundsbeingcommittedtothesepr acticesandstudies,theestablishmentofpropera griculturaldevelopingzonesinredsoilregion,w hichisthefirstandalsothemosteffectivework,i sbecomingmoreandmoreimportant.Withoutthiswo rkeffortswouldbemisdirected,fundswouldbemis spent,andmostsignificantly,developmentinred soilregionwouldcontinuetobeirrational.Howev er,fullresearchonandestablishmentofproperag riculturaldevelopingzonesrequirelargegeogra phical-referenceddatasetsowingtoregionalvar iationsinsoilproperties,environmelltalcondi tionandhumanactivities.Therefore,mailyconve lltionalmathematicalmethodsweredifficulttob eusedinthiskindofstudy.Geographicinformatio nsystem(GIS)isaveryeffectivesystemforinput, manipulation,querylandoutputofalargeamounto f*lProjectsupportedbytheEuropeanCommunity(E 03-137)andZhejiangProvincialScienceandTechn ologyCopomission(913060).spatialdata.Onceth espatialdatabaseinaGISisprepared,theintegra tionbetweenGISandmathematicalmodelcanbeeffi cielltlyusedforstudyofagriculturaldevelopin gzones.Recently,severalpapersdiscussingthei ntegrationbetweenGISandappliedmodelhavebeen published(Hanatal.,1995;Chuvieco,1993;Timan dJolly,1994),however,fewofthemarefromChina. Interactiveself-organizingdata(ISODATA)isam athematicaltechniquewhichhasbeenusedsinceth elate1960s(BallandHall,1967).Itsapplication hascontributedtoresourceclassification,patt ernrecognitionandgeographicalzoning.General lyspeaking,ISODATAisnotaspatialanalysistech niquebecauseitdoesnottakeintoaccountthespat ialdistributionofthedecisionvariables.Howev er,itcanbetailoredtospatialproblemslinkingw ithgeographicaldatabase.Theobjectivesofthis studyweretodevelopanintegrationmethodbetwee nsoilinformationsystemcreatedusingARC/INFO( ESRI,1993)andISODATAcodedinQ-BAISCandtostud yagro-developingzonesbasedonthismethod.MATE RIALSANDMETHODSISODATAalgorithmISODATAalgor ithmisaniteratedprocedureandisfreeofthechoi ceofthenumberofclusters,thereforesimplifyin gcalculatingproceduresandsavingcomputingtim e.Itincludesthefollowingsteps:1)chooseanini tialmatrixoffuzzyclustering;2)computethemat rixoftheclustercenters'whereidenotesthenumb erofindexes,jrepresentsthenumberofsamples,s istakenasthenumberofclusters,Irepresentsthe timeofiteration,Vsiistheelementofclustercen tersmatrix,xijistheelementoforiginaldatamat rix,andrsjisthemembershipdegreeoffuzzyclust eringmatrix;3)computetheadjustedfuzzycluste ringmatrix:whereyisthenumberofclusters,andV viistheelementofclustercentersmatrix;and4)i fmaxstop,otherwisesetl=l+1andreturntoStep2, whereErepresentstherequirederror.StudyareaL ongyouCountylocatedinthemiddlepartoftheJinq uBasin,ZhejiangProvincewasselectedasthestud yarea.Itischaracterizedbymid--tropicalclima tewithanaverageannualtemperatureof17.3"C,an dannualprecipitationof1631.7mm.Accordingtot hesecondgeneraldetailedsoilsurveybyLongyouC ountyAgriculturalBureauin1984,thesoilareaco vers98274hectares,and63.3%ofthemwereredsoil andyellowsoil.LongyouCountyisoneofthemostim portalltagriculturalproductionbasesinZhejia ngProvinceandakeyareaforstudyingredsoilreso urce.SISdata6aseThesoilresourcedatabaseofLo ngyouCountyiscreatedusingARC/INFOsoftwareon SUNworkstation(ShiandWang,1997).ARC/INFOiss uitableforthelinkagebecauseofthecapabilitie sofitsinternalprogramminglanguage.Theoverla y,query3andmapgenerationcapabilitiesarealso majorreasonsforselectingthisGISsystem.Allma pssuchassoiltypemaps,soilsamplemap,terrainm apandtemperaturemapweredigitized,cleaned,ch ecked,andtransformed.Severalattributetables weredeveloped.ResearchscenariosTheprocedure stouseARC/INFOandISODATAfordevelopingzonesw eredesigned(Fig.1).Itincludesthefollowingst eps.Fig.1Theflowchartofsoilinformationsyste m(SIS)andISODATAinterface.StePone.SelectingthemainfactorsaffectingagriculturaldevelopmentusingDelphimethod(TableI).ThespatialfactorswereextractedfromSISdatabase,andthenonsi
More abstracts about the Integration of Soil Information System and Interactive Self-Organizing Data for Agricultural Develop