Mars Weather and Predictability: Modeling and Ensemble Data Assimilation of Spacecraft Observations

Simple item page
dc.contributor.advisorKalnay, Eugeniaen_US
dc.contributor.authorGreybush, Steven J.en_US
dc.contributor.departmentAtmospheric and Oceanic Sciencesen_US
dc.contributor.publisherDigital Repository at the University of Marylanden_US
dc.contributor.publisherUniversity of Maryland (College Park, Md.)en_US
dc.date.accessioned2011-10-08T06:46:55Z
dc.date.available2011-10-08T06:46:55Z
dc.date.issued2011en_US
dc.description.abstractCombining the perspectives of spacecraft observations and the GFDL Mars General Circulation Model (MGCM) in the framework of ensemble data assimilation leads to an improved understanding of the weather and climate of Mars and its atmospheric predictability. The bred vector (BV) technique elucidates regions and seasons of instability in the MGCM, and a kinetic energy budget reveals their physical origins. Instabilities prominent in the late autumn through early spring seasons of each hemisphere along the polar temperature front result from baroclinic conversions from BV potential to BV kinetic energy, whereas barotropic conversions dominate along the westerly jets aloft. Low level tropics and the northern hemisphere summer are relatively stable. The bred vectors are linked to forecast ensemble spread in data assimilation and help explain the growth of forecast errors. Thermal Emission Spectrometer (TES) temperature profiles are assimilated into the MGCM using the Local Ensemble Transform Kalman Filter (LETKF) for a 30-sol evaluation period during the northern hemisphere autumn. Short term (0.25 sol) forecasts compared to independent observations show reduced error (3-4 K global RMSE) and bias compared to a free running model. Several enhanced techniques result in further performance gains. Spatially-varying adaptive inflation and varying the dust distribution among ensemble members improve estimates of analysis uncertainty through the ensemble spread, and empirical bias correction using time mean analysis increments help account for model biases. With bias correction, we estimate a predictability horizon of about 5 sols during which temperature, wind, and surface pressure forecasts initialized from an assimilation analysis are superior to a free running model forecast. LETKF analyses, when compared with the UK reanalysis, show a superior correspondence to independent radio science temperature profiles. Traveling waves in both hemispheres share a correspondence in phase, and temperature differences between the analyses are generally less than 5 K. Assimilation of Mars Climate Sounder (MCS) temperature profiles reveals the importance of vertical distributions of dust and water ice aerosol in reducing model bias. A strategy for assimilation of TES and MCS aerosol products is outlined for future work.en_US
dc.identifier.urihttp://hdl.handle.net/1903/12110
dc.subject.pqcontrolledAtmospheric sciencesen_US
dc.subject.pqcontrolledMeteorologyen_US
dc.subject.pquncontrolledAtmospheric Scienceen_US
dc.subject.pquncontrolledData Assimilationen_US
dc.subject.pquncontrolledEnsemble Data Assimilationen_US
dc.subject.pquncontrolledMarsen_US
dc.subject.pquncontrolledModelingen_US
dc.subject.pquncontrolledWeatheren_US
dc.titleMars Weather and Predictability: Modeling and Ensemble Data Assimilation of Spacecraft Observationsen_US
dc.typeDissertationen_US

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Greybush_umd_0117E_12394.pdf
Size:
5.17 MB
Format:
Adobe Portable Document Format
Download

Collections

UMD Theses and Dissertations
Atmospheric & Oceanic Science Theses and Dissertations