Ensemble Data Assimilation and Breeding in the Ocean, Chesapeake Bay, and Mars

Simple item page
dc.contributor.advisorKalnay, Eugeniaen_US
dc.contributor.advisorCarton, James Aen_US
dc.contributor.authorHoffman, Matthew Josephen_US
dc.contributor.departmentApplied Mathematics and Scientific Computationen_US
dc.contributor.publisherDigital Repository at the University of Marylanden_US
dc.contributor.publisherUniversity of Maryland (College Park, Md.)en_US
dc.date.accessioned2009-10-06T05:53:13Z
dc.date.available2009-10-06T05:53:13Z
dc.date.issued2009en_US
dc.description.abstractMy dissertation focuses on studying instabilities of different time scales using breeding and data assimilation in the oceans, as well as the Martian atmosphere. The breeding method of Toth and Kalnay finds the perturbations that grow naturally in a dynamical system like the atmosphere or the ocean. Here breeding is applied to a global ocean model forced by reanalysis winds in order to identify instabilities on weekly and monthly timescales. The method is extended to show how the energy equations for the bred vectors can be derived with only very minimal approximations and used to assess the physical mechanisms that give rise to the instabilities. Tropical Instability Waves in the tropical Pacific are diagnosed, confirming the existence of bands of both baroclinic and barotropic energy conversions indicated by earlier studies. For regional prediction of smaller timescale phenomena, an advanced data assimilation system has been developed for the Chesapeake Bay Forecast System, a regional Earth System Prediction model. To accomplish this, the Regional Ocean Modeling System (ROMS) implementation on the Chesapeake Bay has been interfaced with the Local Ensemble Transform Kalman Filter (LETKF). The LETKF is among the most advanced data assimilation methods and is very effective for large, non-linear dynamical systems in both sparse and dense data coverage situations. In perfect model experiments using ChesROMS, the filter converges quickly and reduces the analysis and subsequent forecast errors in the temperature, salinity, and velocity fields. This error reduction has proved fairly robust to sensitivity studies such as reduced data coverage and realistic data coverage experiments. The LETKF also provides a method for error estimation and facilitates the investigation of the spatial distribution of the error. This information has been used to determine areas where more monitoring is needed. The LETKF framework is also applied here to a global model of the Martian atmosphere. Sensitivity experiments are performed to determine the dependence of the assimilation on observational data. Observations of temperature are simulated at realistic vertical and horizontal levels and LETKF performance is evaluated. Martian instabilities that impact the assimilation are also addressed.en_US
dc.format.extent5460079 bytes
dc.format.mimetypeapplication/pdf
dc.identifier.urihttp://hdl.handle.net/1903/9514
dc.language.isoen_US
dc.subject.pqcontrolledPhysical Oceanographyen_US
dc.subject.pqcontrolledMathematicsen_US
dc.subject.pqcontrolledAtmospheric Sciencesen_US
dc.subject.pquncontrolledBreeding Methoden_US
dc.subject.pquncontrolledChespeake Bayen_US
dc.subject.pquncontrolledData Assimilationen_US
dc.subject.pquncontrolledEnergy Equationsen_US
dc.subject.pquncontrolledMartian Atmosphereen_US
dc.subject.pquncontrolledTropical Instability Wavesen_US
dc.titleEnsemble Data Assimilation and Breeding in the Ocean, Chesapeake Bay, and Marsen_US
dc.typeDissertationen_US

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Hoffman_umd_0117E_10532.pdf
Size:
5.21 MB
Format:
Adobe Portable Document Format
Download

Collections

UMD Theses and Dissertations
Computer Science Theses and Dissertations
Mathematics Theses and Dissertations