简介：Inthisstudy,thesampledatawasbasedon2190branchlengthandanglesamplesof30treesfromdahurianlarch（Larixgmeiinii）plantationslocatedinWuyingForestBureauinHeilongjiangProvince.Thestepwiseregressiontechniqueswereusedtodevelopbranchlengthandbranchanglemodels:BL=b1+b2DINC+b3DINC2+b4DBH·DINC2,BA=b1+b2DINC+b3DINC2+b4DBH·DINC.Then,thedevelopedmodelswerefittedusinglinearmixed-effectsmodelingapproachbasedonLMEprocedureofS-PLUSsoftware.Evaluationstatistics,suchasAIC,BIC,LogLikelihoodandLikelihoodratiotestwereusedformodelcomparisons.Theresultsshowedthatthebranchlengthandbranchanglemodelswithparametersb1,b2,b3asmixedeffectsshowedthebestperformance.Exponentialandpowerfunctionswereincorporatedintomixedbranchlengthandbranchanglemodel.Theadditionoftheexponentialandpowerfunctionssignificantlyimprovedthemixed-effectsmodel.Theplotsofstandardizedresidualsindicatedthatthemixed-effectmodelwithexponentialandpowerfunctionsshowedmorehomogeneousresidualvariancethanthemixed-effectsmodel.Validationconfirmedthatthemixedmodelwithcalibrationofrandomparameterscouldprovidemoreaccurateandpreciseprediction.Therefore,theapplicationofmixedmodelnotonlyshowedthemeantrendsofbranchlengthandbranchangle,butalsoshowedtheindividualdifferencebasedonvariance-covariancestructure.

简介：MODIS(中等决定成像分光辐射函数)是在地(曙光女神AM)和水(曙光女神下午)上的一台关键仪器卫星。线性光谱混合模型为陆地封面的亚象素分类被用于MOIDS数据。在中国的浙江省的Shaoxing县被选择是学习地点，早米饭作为学习庄稼被选择。从MODIS象素使用的陆地封面的导出的比例线性光谱混合模型与从在一样的天获得的TM数据导出的无指导的分类相比，它暗示MODIS数据能为米饭耕作被用作卫星数据来源区域评价，可能米饭生长在地区性的规模上监视并且产量预报。

简介：Basedonamultilevellinearmixedmodelapproach,anindividualdiameterincrementmodelwasdevelopedforfirplantationtreesgrowinginJiangxiProvince.Thedatasetusedinthisstudycamefromlong-termpermanentresearchplots.Thedatabaseconsistsoftotalof82counties,365plots,5416treesand16248observations.Thepaperchosemixedeffectsmodelsinsteadofregressionanalysisapproachbecauseitallowsforpropertreatmentoferrortermsandcorrelationinarepeatedmeasuresanalysisframework.Themodelwasdefinedasamixedlinearmodelwithparameterrandomeffectofplot,areaorplotandareasimultaneous.Inadditiontheheteroscedasticityandcorrelationwastakenintoaccount.Mixedmodelcalibrationofdiameterincrementwascarriedoutwiththeindependentdatausingadifferentsampleofcomplementaryobservations.Theresultshowedthatthetotalstandbasalarea,thediameteroftargettrees,theratioofbasalareaoflargertreestotargettreediameter,andaltitudewerefoundtobesignificantpredictors.Boththefittingmodelandthecalibratedmodelmeanasubstantialimprovementcomparedwiththeclassicalapproachwidelyusedinforestmanagement.Aftertakingintoaccountreasonablevariancefunctionofheteroscedasticityandcorrelation,themodelshowsbetterofgoodnessoffitthanonlytakingintoaccountparameterrandomeffects.Thistypeofmodelingmethodologyshowsflexible,preciseandaccurate.