Evaluating radiance and retrieval assimilation of Mars Thermal Emission Spectrometer Spacecraft observations

Open Access
Kalogeras, Petros
Graduate Program:
Master of Science
Document Type:
Master Thesis
Date of Defense:
June 16, 2015
Committee Members:
  • Steven J Greybush, Thesis Advisor
  • Eugene Edmund Clothiaux, Thesis Advisor
  • Fuqing Zhang, Thesis Advisor
  • Data assimilation
  • Mars
  • NWP
  • Retrieval
  • Radiance
  • TES
In the current study two types of data assimilation approaches for updating atmospheric temperature profiles on Mars are explored. Namely, we investigate a retrieval and a radiance data assimilation, where in each case the assimilated observations are atmospheric temperature profiles and satellite spectra respectively. Temperature profiles alongside pressure distributions are simulated by utilizing the Mars Global Climate Model (MGCM). In particular, these temperature profiles serve as the prior to the applied filtering technique, an ensemble Kalman filter (EnKF). Direct satellite measurements in the form of Thermal Emission Spectrometer (TES) spectra of radiance available from the Planetary Data System (PDS), as well as Optimal Spectral Sampling (OSS) temperature retrievals deduced from PDS TES radiance spectra, are considered as the observations to be assimilated. Temperature retrievals, which are not direct observations, depend upon their respective acquisition procedures and in this sense are not unique. On the other hand, observed spectra cannot be compared directly with a weather model. Nevertheless, both geophysical quantities can be considered as observations and in this lies the focus of the current project. The findings suggest that both types of observations can yield physically sensible Kalman filtering analysis results. However, the quality of the results depends profoundly on how the filtering approach is carried out, with vertical and spectral localizations having the greatest impact. Comparing retrieval and radiance assimilations for the Martian atmosphere is also relevant to NWP on Earth; the Martian observations network is limited to satellite observations, whereas the terrestrial observations include satellite measurements, atmospheric soundings, and data procured from ground weather stations. Considering the paucity of data directly usable in a Martian weather model, the Martian case may then serve as suitable test bed being in the position to highlight viable assimilation procedures, in the context of operating only with such observed spectra.