Assimilation of Both Column- and Layer-Integrated Dust Opacity Observations in the Martian Atmosphere

A new dust data assimilation scheme has been developed for the UK version of the Laboratoire de Meteorologie Dynamique Martian General Circulation Model. The Analysis Correction scheme (adapted from the UK Met Office) is applied with active dust lifting and transport to analyze measurements of temperature, and both column-integrated dust optical depth (CIDO), tau(ref) (rescaled to a reference level), and layer-integrated dust opacity (LIDO). The results are shown to converge to the assimilated observations, but assimilating either of the dust observation types separately does not produce the best analysis. The most effective dust assimilation is found to require both CIDO (from Mars Odyssey/THEMIS) and LIDO observations, especially for Mars Climate Sounder data that does not access levels close to the surface. The resulting full reanalysis improves the agreement with both in-sample assimilated CIDO and LIDO data and independent observations from outside the assimilated data set. It is thus able to capture previously elusive details of the dust vertical distribution, including elevated detached dust layers that have not been captured in previous reanalyzes. Verification of this reanalysis has been carried out under both clear and dusty atmospheric conditions during Mars Years 28 and 29, using both in-sample and out of sample observations from orbital remote sensing and contemporaneous surface measurements of dust opacity from the Spirit and Opportunity landers. The reanalysis was also compared with a recent version of the Mars Climate Database (MCD v5), demonstrating generally good agreement though with some systematic differences in both time mean fields and day-to-day variability. Plain Language Summary Data assimilation is a method of combining atmospheric observations, which are inevitably uncertain and incomplete in their coverage, with a global numerical model. It is commonly used for the Earth to initialize weather forecasts, with associated benefits for climate analysis and prediction. This technique has also been used for the Martian atmosphere, using measurements of temperature, dust and ice from satellites in orbit around Mars. But most previous efforts have only used measurements of the total amount of dust in a vertical column from instruments that look vertically downwards to the Martian surface. In new work presented here, however, we also use detailed measurements of the vertical structure of the dust distribution from an instrument that looks toward the edge of the planet. This is much more effective when atmospheric dust is not mainly concentrated near the ground. Such events are reasonably common on Mars, when elevated layers of dust are formed, which can strongly affect how the atmosphere is heated by the Sun. We present examples of situations when previous methods failed to recover the correct dust distribution, as verified against independent measurements for example, from the Spirit and Opportunity Rovers, and compare with the ESA Mars Climate Database.

Automated summary

A new dust data assimilation scheme has been developed for the UK version of the Laboratoire de Meteorologie Dynamique Martian General Circulation Model, which combines different observations to optimize dust analysis and improve the accuracy of Martian atmospheric process simulations. It has been found that combining CIDO and LIDO observational data produces the most effective dust assimilation results.