Font Size: a A A

The Applyment Of Kalman Filter Method In Advanced Regional Eta-coordinate Numerical Prediction Model

Posted on:2006-09-11Degree:MasterType:Thesis
Country:ChinaCandidate:C MaFull Text:PDF
GTID:2120360152488756Subject:Traffic Information Engineering & Control
Abstract/Summary:PDF Full Text Request
This text mainly use kalman filter method to set up a data assimilate system, and apply to Arem number value predict mode , through dealing with initial assimilation field of Arem mode, we offer a more perfect initial field for Arem mode, in order to achieve the goal of optimizing Arem mode. Torrential rain is major casualty weather of our country,and big flood that torrential rain cause give people lives and properties and national economic construction of our country bring enormous losses often. Predict torrential rain weather for better prediction, our country academy of sciences atmospheric physics research institute number value predict set up one r| - number value of the coordinate of topography predict mode that can consider precipitous limited regional (REM), and develop into present Arem mode on this basis. The research purpose of this text is to look for a kind of rational assimilation method to offer a better assimilation module for Arem mode, optimize Arem mode. Some innovation of this text whether first time assemble kalman filter (ENKF) method is it get assimilation , Arem mode of the module to introduce, and analyse the test about these system through odd concrete an example of torrential rain , through is it divide comparative analysis of method to become with original Arem mode and four-dimension , improve set kalman filter assimilate the system progressively and then, and achieve the goal of optimizing Arem mode finally. The article has introduced the origin assimilated, basic method and developing newly first in the part of the introduction, introduce kalman strain basic thought and characteristic of wave , end whether set turn on graceful to strain wave business latest research that study develop and predict application of in set. Inside chapter two, this text set up four little kalman strain wave system and make simple analysis to four system these at first, combine set statistical method is it set up set kalman filter system to begin on the basis of these four systems then, before setting up and assembling kalman filter system, I strain to kalman filter and four-dimension equivalence divided to become is it prove and strain to set kalman filter and four-dimension is it is it compare to assignto become to make first, then begin to set up set kalman filter system. Set kalman that this text set up strain wave assimilate system one horizontal plane net, at the vertical direction the analyticl system P for longitude and latitude, Arakawa A of lattice point, the dimension in the horizontal and vertical direction is adjustable.The system is a software system that Fortran90 wrote, adopt the module structure. In is it assemble kalman filter to set up, I analyse to this system through several concrete an example of torrential rain in chapter three of this text, through compare original Arem mode with four-dimension become method of dividing , I make sure improvement and make the good result to the system. Therefore it can as a conclusion,ensemble kalman filter count with the advantages ofkalman filter as one expansion kalman filter that produced through combining ensemble method and kalman filter , prevent from calculate in a large amount that can't be yet bored at present at least, and the assimilation result to the non-linear mode is obviously improved, optimizing the number value predicted mode of torrential rain -Arem mode.
Keywords/Search Tags:assimilation, kalman filter, ensemble, Arem mode
PDF Full Text Request
Related items