9:15 AM - 9:30 AM
[PCG19-02] Analysis of surface pressure distribution on Mars by CO2 2 μm absorption band observed by OMEGA/Mars Expres
Keywords:Mars, Radiation Transfer, Meteorological
Surface pressure measurements are crucial to characterize the dynamics of the atmosphere in metrology. On Mars, the only observations of pressure fluctuations associated with these events are at specific points by landers and by remote sensing from a limited orbiter.Remote sensing can retrieve the surface pressure from the CO2 column density. The only case to retrieve the surface pressure distribution from orbiter data is the work of Forget et al. (2007) and Spiga et al. (2007). They derived the surface pressure using CO2 absorption at 2 μm in the initial data (2004-2005) of near-infrared spectrometer OMEGA onboard Mars Express. Although they only used the data with the ideal low dust condition (optical depth < 0.4, 0.7% of the total data (29 of ~4000 nadir observations)), they could obtain the pressure distribution over 95 x 150 km (2.5 deg in longitude, 4 deg in latitude) and show some case studies such as the pressure gradients associated with the baroclinic storm track.
Near-infrared spectrometer MIRS is onboard MMX to be launched in 2024. MIRS can take wider spectral images and shorter time intervals than OMEGA with similar wavelength range and resolution, so the fast retrieval tool is critically requested. Our goal is to examine more of the horizontal distribution of surface pressure on Mars using MMX to understand meteorology. We developed a surface pressure fast retrieval tool using 25 OMEGA spectra between 1.8 and 2.2 μm which include the 2 μm CO2 absorption band observed OMEGA.
To process the hundreds of thousands of pixels present in each OMEGA image, retrieval of the horizontal distribution of surface pressure requires both high speed and high accuracy. In this study, we used a fast retrieval method of surface pressure by using a Look Up Table (LUT), following Forget et al. (2007). A total of 4,252,500 radiance spectra were derived and stored in the LUT by entering the physical parameters that affect the observed spectra. The LUT spectrum is then interpolated in multiple dimensions to retrieve the surface pressure. For this retrieval, we tried two methods: (1) a new method to use the equivalent width of the CO2 absorption band between observed and LUT spectra and (2) a method to fit the observed spectrum to the LUT spectrum. The former is less accurate but faster and the latter is slow but more accurate.
In the retrieval, fixed parameters are used for spacecraft locations and geometric ones. Temperature profiles, dust opacity, and water-ice cloud opacity are from the Mars Climate Database Version 5.3 (MCD v5.3). For the equivalent width method, the surface albedo is fixed based on the continuum. For the spectral fitting method, both surface pressure and surface albedo are retrieved as free parameters.
The validation was based on a comparison with Forget et al. (2007) at three locations (Hellas Planitia, Chryse Planitia, and Terra Meridiani). And the relative accuracy of surface pressure in the region was verified by the comparison between two surface pressure retrieved in the same area from two different orbits, 3 days apart. We confirmed that the derivation accuracy of the Equivalent width method provides the pressure with the accuracy from +3% to +42% at specific locations and 3σ ~ ±30 Pa (±4%) between 3-day differences at the same location. The spectral fitting method has better accuracy, which ranges from -0.1% to -2.3% at specific locations and 3σ ~ ±15 Pa (±2.0%) between 3-day difference, which is similar to Forget et al. (2007). In addition, a method for retrieval of the altitude corrections derived sea-level pressure in the fitting method was 3σ~±10 Pa (±1.7%). It is enough to study the meteorological analysis of Mars for atmospheric waves, etc. From them, we will identify metrological phenomena on a regional scale whose pressure distributions are not matched to the MCD and apply the Spectral fitting method for scientific analysis with the sea-level pressure maps with a relative accuracy of <5%. For practical studies, we will apply the equivalent width method for the event surveys and quick look analyses, and the spectral fitting method for the detailed analyses.
In this presentation, we will show the results of the initial analysis of the surface pressure distribution during Martian year 28 (January 2006-December 2007) using this retrieval tool. For the Equivalent width method, full analyses become realistic because the pressure map creation time can be ~50 days for 2348 data sets (the Spectral Fitting method: ~1200 days). We will also discuss the possible application to MIRS/MMX. It will first enable the derivation of a wide pressure distribution and its fluctuations on Mars, which are one of the main contributions to Martian meteorology by the MMX mission.
Near-infrared spectrometer MIRS is onboard MMX to be launched in 2024. MIRS can take wider spectral images and shorter time intervals than OMEGA with similar wavelength range and resolution, so the fast retrieval tool is critically requested. Our goal is to examine more of the horizontal distribution of surface pressure on Mars using MMX to understand meteorology. We developed a surface pressure fast retrieval tool using 25 OMEGA spectra between 1.8 and 2.2 μm which include the 2 μm CO2 absorption band observed OMEGA.
To process the hundreds of thousands of pixels present in each OMEGA image, retrieval of the horizontal distribution of surface pressure requires both high speed and high accuracy. In this study, we used a fast retrieval method of surface pressure by using a Look Up Table (LUT), following Forget et al. (2007). A total of 4,252,500 radiance spectra were derived and stored in the LUT by entering the physical parameters that affect the observed spectra. The LUT spectrum is then interpolated in multiple dimensions to retrieve the surface pressure. For this retrieval, we tried two methods: (1) a new method to use the equivalent width of the CO2 absorption band between observed and LUT spectra and (2) a method to fit the observed spectrum to the LUT spectrum. The former is less accurate but faster and the latter is slow but more accurate.
In the retrieval, fixed parameters are used for spacecraft locations and geometric ones. Temperature profiles, dust opacity, and water-ice cloud opacity are from the Mars Climate Database Version 5.3 (MCD v5.3). For the equivalent width method, the surface albedo is fixed based on the continuum. For the spectral fitting method, both surface pressure and surface albedo are retrieved as free parameters.
The validation was based on a comparison with Forget et al. (2007) at three locations (Hellas Planitia, Chryse Planitia, and Terra Meridiani). And the relative accuracy of surface pressure in the region was verified by the comparison between two surface pressure retrieved in the same area from two different orbits, 3 days apart. We confirmed that the derivation accuracy of the Equivalent width method provides the pressure with the accuracy from +3% to +42% at specific locations and 3σ ~ ±30 Pa (±4%) between 3-day differences at the same location. The spectral fitting method has better accuracy, which ranges from -0.1% to -2.3% at specific locations and 3σ ~ ±15 Pa (±2.0%) between 3-day difference, which is similar to Forget et al. (2007). In addition, a method for retrieval of the altitude corrections derived sea-level pressure in the fitting method was 3σ~±10 Pa (±1.7%). It is enough to study the meteorological analysis of Mars for atmospheric waves, etc. From them, we will identify metrological phenomena on a regional scale whose pressure distributions are not matched to the MCD and apply the Spectral fitting method for scientific analysis with the sea-level pressure maps with a relative accuracy of <5%. For practical studies, we will apply the equivalent width method for the event surveys and quick look analyses, and the spectral fitting method for the detailed analyses.
In this presentation, we will show the results of the initial analysis of the surface pressure distribution during Martian year 28 (January 2006-December 2007) using this retrieval tool. For the Equivalent width method, full analyses become realistic because the pressure map creation time can be ~50 days for 2348 data sets (the Spectral Fitting method: ~1200 days). We will also discuss the possible application to MIRS/MMX. It will first enable the derivation of a wide pressure distribution and its fluctuations on Mars, which are one of the main contributions to Martian meteorology by the MMX mission.