简介：四维的变化数据吸收(4DVar)是最有希望的方法之一为数字天气预言(NWP)提供最佳的分析。在世界上的五个国家NWP中心成功地在他们的全球NWP使用了4DVar方法，由于增长方法和伴随技术。然而，4DVar的应用程序被在许多NWP中心和研究院可得到的计算机资源仍然限制。因此，进一步减少4DVar的计算费用是必要的。这里，实现4DVar的一条节俭的途径被建议，用dimensionreduced设计(DRP)的技术，它被称为建议途径基于尺寸减小使用的鈥淒RP-4DVar.鈥?历史的样品到的一个整体定义subspace。它直接由恰当的观察在减少的空间获得一个最佳的答案，历史的时间系列由模型产生了到形成一致预报状态，因此不要求正切的伴随的实现线性近似。在吸收mesoscale观察的不同类型上评估DRP-4DVar的性能，一些观察系统模拟实验用MM5被进行，比较用一个6小时的吸收窗口在基于伴随的4DVar和DRP-4DVar之间被做。关键词4DVar-伴随-尺寸减小-历史的样品-观察系统模拟实验

简介：Basedontheobserved2-yeartemperaturedataforfourkindsoftypicalurbanunderlyingsurfaces,includingasphalt,cement,barelandandgrassland,theannualvariationsandinfluencingfactorsoflandsurfacetemperatureareanalyzed.Thenfittingequationsforsurfacetemperatureareestablished.Itisshownthattheannualvariationofdailyaverage,maximumandminimumtemperatureanddailytemperaturerangeonthefoururbanunderlyingsurfacesisconsistentwiththechangeinairtemperature.Thedifferenceoftemperatureondifferentunderlyingsurfacesinthesummerhalfyear(MaytoOctober)ismuchmoreevidentthanthatinthewinterhalfyear(DecembertothefollowingApril).Thedailyaverageandmaximumtemperaturesofasphalt,cement,barelandandgrasslandarehigherthanairtemperatureduetotheatmosphericheatinginthedaytime,withthatofasphaltbeingthehighest,followedinturnbycement,barelandandgrassland.Moreover,thedailyaverage,maximumandminimumtemperatureonthefoururbanunderlyingsurfacesarestronglyimpactedbytotalcloudamount,dailyaveragerelativehumidityandsunshinehours.Thelandsurfacecanbecooled(warmed)byincreasedtotalcloudamount(relativehumidity).Thechangesintemperatureonbarelandandgrasslandareinfluencedbyboththetotalcloudamountandthedailyaveragerelativehumidity.Thetemperatureparametersofthefourlandsurfacesaresignificantlycorrelatedwithdailyaverage,maximumandminimumtemperature,sunshinehours,dailyaveragerelativehumidityandtotalcloudamount,respectively.Theanalysisalsoindicatesthattherangeoffittingparameterofalinearregressionequationbetweenthesurfacetemperatureofthefourkindsoftypicallandsurfaceandtheairtemperatureisfrom0.809to0.971,passingtheF-testwithaconfidencelevelof0.99.

简介：这研究检验与一个整体Kalman过滤器(EnKF)联合确定的四维的变化吸收系统(4DVAR)为数据吸收生产一条优异混合途径的性能。当在阻止过滤器分叉利用4DVAR时，从使用州依赖者的不确定性的联合吸收计划(E4DVAR)好处由EnKF提供了：4DVAR分析通过费用的最小化生产以后的最大的可能性答案整体不安关于被转变的功能，和产生整体分析能为下一个吸收周期并且作为整体预报的一个基础向前被宣传。这条联合途径的可行性和有效性与模仿的观察在一个理想化的模型被表明。E4DVAR能够在完美模型、有瑕疵模型的情形下面超过4DVAR和EnKF，这被发现。联合计划的性能比为标准EnKF或4DVAR实现的那些对整体尺寸或吸收窗口长度也不太敏感。

简介：Minimizationalgorithmsaresingularcomponentsinfour-dimensionalvariationaldataassimilation(4DVar).Inthispaper,theconvergenceandapplicationoftheconjugategradientalgorithm(CGA),whichisbasedontheLanczositerativealgorithmandtheHessianmatrixderivedfromtangentlinearandadjointmodelsusinganon-hydrostaticframework,areinvestigatedinthe4DVarminimization.First,theinfluenceoftheGram-SchmidtorthogonalizationoftheLanczosvectorontheconvergenceoftheLanczosalgorithmisstudied.TheresultsshowthattheLanczosalgorithmwithoutorthogonalizationfailstoconvergeaftertheninthiterationinthe4DVarminimization,whiletheorthogonalizedLanczosalgorithmconvergesstably.Second,theconvergenceandcomputationalefficiencyoftheCGAandquasi-Newtonmethodinbatchcyclingassimilationexperimentsarecomparedonthe4DVarplatformoftheGlobal/RegionalAssimilationandPredictionSystem(GRAPES).TheCGAis40%morecomputationallyefficientthanthequasi-Newtonmethod,althoughtheequivalentanalysisresultscanbeobtainedbyusingeithertheCGAorthequasi-Newtonmethod.Thus,theCGAbasedonLanczositerationsisbetterforsolvingtheoptimizationproblemsintheGRAPES4DVarsystem.

简介：Accurateestimatesofalbedosarerequiredinclimatemodeling.Accurateandsimpleschemesforradiativetransferwithincanopyarerequiredfortheseestimates,butseverelimitationsexist.Thispaperdevelopedafour-streamsolarradiativetransfermodelandcoupleditwithalandsurfaceprocessmodel.Theradiativemodelusesafour-streamapproximationmethodasintheatmospheretoobtainanalyticsolutionsofthebasicequationofcanopyradiativetransfer.Asananalyticalmodel,thefour-streamradiativetransfermodelcanbeeasilyappliedefficientlytoimprovetheparameterizationoflandsurfaceradiationinclimatemodels.Ourfour-streamsolarradiativetransfermodelisbasedonatwo-streamshortwaveradiativetransfermodel.Itcansimulateshortwavesolarradiativetransferwithincanopyaccordingtotherelevanttheoryintheatmosphere.Eachparameterofthebasicradiativetransferequationofcanopyhasspecialgeometryandopticalcharactersofleavesorcanopy.Theupwardordownwardradiativefluxesarerelatedtothediffusephasefunction,theG-function,leafreflectivityandtransmission,leafareaindex,andthesolarangleoftheincidentbeam.Thefour-streamsimulationiscomparedwiththatofthetwo-streammodel.Thefour-streammodelisprovedsuccessfulthroughitsconsistentmodelingofcanopyalbedoatanysolarincidentangle.Inordertocompareandfinddifferencesbetweentheresultspredictedbythefour-andtwo-streammodels,anumberofnumericalexperimentsareperformedthroughexaminingtheeffectsofdifferentleafareaindices,leafangledistributions,opticalpropertiesofleaves,andgroundsurfaceconditionsonthecanopyalbedo.Parallelexperimentsshowthatthecanopyalbedospredictedbythetwomodelsdiffersignificantlywhentheleafangledistributionissphericalandvertical.Theresultsalsoshowthatthedifferenceisparticularlygreatfordifferentincidentsolarbeams.OneadditionalexperimentiscarriedouttoevaluatethesimulationsoftheBATSla

简介：LASG/IAP气候海洋模型，的一个四维的变化数据吸收(4DVar)系统版本1.0(LICOM1.0)，命名LICOM-3DVM，用印射的观察(3DVM)的三维的变化数据吸收被开发了，一个4DVar方法最新在过去的二年里求婚了。有12年的模型集成的二个实验被设计验证它。一吸收被运用的、叫的ASSM，它合并了分析的每周的海表面温度(SST)由LICOM-3DVM每周一次从在1990和2001之间的雷纳兹和史密斯(OISST)回答。没有任何吸收，其它是控制跑，命名CTL。ASSM表演上面的海洋的模仿的温度(在50米上面)，特别赤道的太平洋的SST，与回归线空气海洋(道)绳索数据与一致，世界海洋地图集2001(WOA01)数据和遇见的办公室哈德利中心是海冰和海表面温度(HadISST)数据。它减少了在东方太平洋在CTL存在的冷偏爱并且生产了同意观察很好的一个Niño索引。确认结果建议LICOM-3DVM能有效地调整海洋温度的模型结果，尽管它是难的改正表面下的结果，它甚至由于仅仅表面数据的加入在一些区域使他们更坏。LICOM-3DVM的未来开发在situ观察是包括表面下的，进一步改善的卫星观察为模拟建模。

简介：NumericalexperimentsofadjointvariationalassimilationhavebeenperformedusingtheknownLorenzsystem.Withtheincreaseofsensitivityofmodel’sinitialvalues,itismoreandmoredifficulttousetheadjointmethodtogettheinitialvalueswhichareconsistentwiththedynamicsofthefore-castmodel.Undersomecircumstancesthealgorithmcompletelyfails.Thisshowsthatfour-dimen-sionalassimilationisrelatedtothelimitofpredictability.Ontheotherhand.withtheincreaseofmodelequation’serror,theresultofvariationalassimilationmaybecomeworseandworsesothatthepredictionhasnomeaning.Butifthemodelparametersarecorrectedwhenvariationalassimilationismade,theforecastresultscanbegreatlyimprovedbasedonLorenzmodel.

简介：一四维变化(4D-Var)数据吸收方法在热带太平洋的一个改进中间的联合模型(ICM)被实现。一个成双的实验被设计基于ICM在ENSO分析和预言上评估4D-Var数据吸收算法的影响。模型错误被假定仅仅从参数无常产生。SST异例的观察，从拿缺省参数价值并且把Gaussian噪音增加的真相模型模拟被取样，直接错误地与它的参数集合被吸收进吸收模型。结果证明4D-Var有效地减少ENSO分析的错误因此与非吸收盒子相比改进ENSO事件的预言技巧。这些结果提供一个有希望的方法让ICM完成更好即时的ENSO预言。