简介:Background:Monitoringforesthealthandbiomassforchangesovertimeintheglobalenvironmentrequirestheprovisionofcontinuoussatelliteimages.However,opticalimagesoflandsurfacesaregenerallycontaminatedwhencloudsarepresentorrainoccurs.Methods:Toestimatetheactualreflectanceoflandsurfacesmaskedbycloudsandpotentialrain,3DsimulationsbytheRAPIDradiativetransfermodelwereproposedandconductedonaforestfarmdominatedbybirchandlarchinGenheCity,DaXing’AnLingMountaininInnerMongolia,China.Thecanopyheightmodel(CHM)fromlidardatawereusedtoextractindividualtreestructures(location,height,crownwidth).Fieldmeasurementsrelatedtreeheighttodiameterofbreastheight(DBH),lowestbranchheightandleafareaindex(LAI).SeriesofLandsatimageswereusedtoclassifytreespeciesandlandcover.MODISLAIproductswereusedtoestimatetheLAIofindividualtrees.CombiningalltheseinputvariablestodriveRAPID,high-resolutionopticalremotesensingimagesweresimulatedandvalidatedwithavailablesatelliteimages.Results:Evaluationsonspatialtexture,spectralvaluesanddirectionalreflectancewereconductedtoshowcomparableresults.Conclusions:Thestudyprovidesaproof-of-conceptapproachtolinklidarandMODISdataintheparameterizationofRAPIDmodelsforhightemporalandspatialresolutionsofimagereconstructioninforestdominatedareas.
简介:Themodernizationandinformatizationmanagementofancientandfamoustrees’managementisanimportantstandardwithwhichthemunicipalpublicresourcemanagementlevelofaregionorcitycanbemeasured.Designanddevelopmentofancientandfamoustrees’three-dimensionalmanagementinformationsystemwasrealizedbasedonvirtualplantbyusingintegratedtechniquesofvirtualplantsandGeographicInformationSystem.Thesystemarchitecturedesignwasdevelopedonthebasisoffunctionalrequirements,andthepracticalsystemwasachievedinVisualStudio2008developmenttoolsandOpenGLgraphicsstandards.Thesystemhasfunctionofancienttreesarchivesmanagement,three-dimensionalreconstructionofthetreessurroundingenvironment,individualtreeinformationinquiryinthree-dimensionalscene,etc.Applicationanddisseminationofthissystemwillgreatlypromotethemanagementandprotectionofancientandfamoustreesstandardization,informalizationandvisualization
简介:摘要 : 冠层光截获能力是反映作物品种间差异的重要功能性状,高通量表型冠层光截获对提高作物改良效率具有重要意义。本研究以小麦为研究目标,利用数字化植物表型平台( D3P)模拟生成了 100种冠层结构不同的小麦品种在 5个生育期的三维冠层场景,记录了从原始冠层结构中提取的绿色叶面积指数( GAI)、平均倾角( AIA)和散射光截获率( FIPARdif)信息作为真实值 ,进一步利用上述三维小麦场景开展了虚拟的激光雷达( LiDAR)模拟实验,生成了对应的三维点云数据。基于模拟的点云数据提取了其高度分位数特征( H)和绿色分数特征( GF)。最后,利用人工神经网络( ANN)算法分别构建了从不同 LiDAR点云特征( H、 GF和 H+GF)输入到 FIPARdif、 GAI和 AIA的反演模型。结果表明,对于 GAI、 AIA和 FIPARdif,预测精度从高到低对应的点云特征输入为 GF+H > H > GF。由此可见, H特征对提高目标表型特性的估算精度起到了重要作用。输入 GF + H特征,在中等测量噪音( 10%)情况下, FIPARdif和 GAI的估算均获得了满意精度, R2分别为 0.95和 0.98,而 AIA的估算精度( R2=0.20)还有待进一步提升。本研究基于 D3P模拟数据开展,算法的实际表现还有待通过田间数据进一步验证。尽管如此,本研究验证了 D3P协助表型算法开发的能力,展示了高通量 LiDAR数据在估算田间冠层光截获和冠层结构方面的较高潜力。
简介:用线粒体DNA的D-loop和Cytb基因序列分析方法研究了吉林延吉、敦化和辽宁法台3个区域的29尾拉氏PhoxinuslagowskiiDybowsky的遗传多样性。经PCR扩增和测序,获得了783~785bpD-loop和818bpCytb的同源序列。两者多态性遗传参数统计显示,29尾个体分别存在47(D-loop)和89(Cytb)个变异位点,分别检测出15(D-loop)和11(Cytb)个单倍型,总群体单倍型(H_d)分别为0.8966(D-loop)和0.8990(Cytb),核苷酸多样性指数(Pi)分别为0.0246(D-loop)和0.0498(Cytb),平均核苷酸差异数(K)分别为19.2857(D-loop)和40.7365(Cytb)。分子方差分析(AMOVA)结果表明,79.02%(D-loop)和81.69%(Cytb)变异来自群体间,20.98%(D-loop)和18.31%(Cytb)来自群体内。单倍型呈明显的地理差异,分成2个分支,一个以延吉群体为主,一个以法台群体为主。拉氏的遗传多样性水平较高,群体间遗传分化明显。该结果可为拉氏的种质资源保护提供参考。