nai

Project Progress

Search for biomarkers on Mars

We use state-of-the-art infrared spectrometers to search for signatures of biology and geology on Mars. Searching for volatile species at Near InfraRed (NIR) wavelengths is most favorable, since the 2-5 mm range encompasses strong fundamental bands for a plethora of molecules (e.g., H₂O, HDO, CO₂, CH₄, C₂H₆, H₂CO, CH₃OH, C₂H₂, C₂H₄, NH₃ and HCN), and these are accessible with ground-based telescopes. In particular, several important hydrocarbons, including methane (the most abundant hydrocarbon in the Solar System), are symmetric molecules and can only be detected through their ro-vibrational transitions in the NIR since they have no permanent dipole moment and therefore no allowed pure rotational transitions. Furthermore, many of these molecules play a key role in our definition of “habitability” and in the possible origin of life in our planet (e.g. Urey-Miller experiment), ultimately driving their search across the Universe.

In the last two years, we acquired the deepest and most comprehensive search for biomarkers on Mars using powerful infrared high-resolution spectrometers (CRIRES, NIRSPEC, CSHELL) at high-altitude observatories (VLT, Keck-II, NASA-IRTF respectively). Among ground-based and spacecraft spectrometer/telescope combinations at infrared wavelengths, these instruments provide the highest spectral resolving power for spatially-resolved spectra in the 2-5 µm region, making these measurements of unique value. The measurements span a broad range of seasons on Mars: in L_S = 324 to 24° for the blue Doppler shift data, and L_S = 44 to 102° for the red Doppler shift. For certain observations, we enabled the adaptive optics mode, obtaining the highest spatial resolution achievable from Earth.

In order to analyze this unprecedented wealth of data, we developed highly automated and advanced processing techniques that correct for bad-pixels/cosmic-rays and perform spatial and spectral straightening with milli-pixel precision. In order to reach maximum sensitivities (photon-noise limited), we also developed new methods to account for many subtle instrumental effects, such as removal of internal scattered-light, correction for variable resolving power and removal of spectral fringing (using Lomb periodogram analysis). Spectral calibration and compensation for telluric absorption is achieved by comparing with highly precise atmospheric radiance and transmittance models synthesized at sub-Doppler resolution (λ/ δλ ~ 10,000,000). The proper synthesis of terrestrial spectra requires line-by-line, layer-by-layer radiative transfer modeling of the atmosphere. The quality of the synthesized spectrum depends directly on the robustness of the used set of radiative transfer equations, the precision of the adopted spectroscopic constants, and the accuracy of the assumed atmospheric conditions. Since 1981, our group has worked with three radiative transfer models (SSP, GENLN2, and LBLRTM) for the synthesis of telluric spectra. The Spectrum Synthesis Program (SSP, Kunde & Maguire 1974) is a robust model, although with limitations when synthesizing spectra at extremely high spectral resolution. The latest version (v4) of GENLN2 (Edwards 1992) provides highly realistic and Doppler-limited spectral synthesis of the terrestrial atmosphere, but we encountered problems in the calculation of spectral line shapes since the model incorrectly accounts for pressure-induced spectral shifts (Villanueva et al. 2008a; 2008b). We have recently integrated the highly robust and extensively validated Line-By-Line Radiative Transfer Model (LBLRTM, Clough et al. 2005) into our data processing methods, and obtained excellent results (e.g., Villanueva et al. 2011a; 2011b), and further modernized it by integrating realistic isotopic fractionation vertical profiles for water (H₂O, HDO) and CH₄ ([^12^CH₄, ¹³CH₄).

Perhaps, the biggest challenge of remote sensing of planetary atmospheres is the high-degree of incompleteness and numerous inaccuracies contained in the current spectral databases. For instance we found that a significant fraction of the C₂H₂, C₂H₆ and NH₃ databases in HITRAN have incorrect frequencies, energies, line IDs, Einstein-A values and statistical weights (e.g., Villanueva et al. 2011a; Gordon et al. 2011). These errors occur mainly because empirical databases are a collection of values obtained from diverse experiments employing different calibration techniques ultimately giving rise to inconsistencies (see H₂O section 2.1 in Rothman et al. 2009). Correction and completion of these atlases is particularly complex, and in most cases requires the development of ad-hoc quantum mechanical models (as we did for ethane: Villanueva et al. 2011a) and the compilation/ validation/integration of several spectral databases (as we did for water: Villanueva et al. 2011b). Even though ethane is a trace gas (~ 1 ppb abundances) in our atmosphere, it has a complex spectrum with numerous features in the main spectral region used to search for methane on Mars. Consequently, our new model for ethane allows us to remove telluric features in this spectral region with high precision, reaching unprecedented accuracy in the residuals.

The precise modeling of the solar spectrum is particularly important when observing planetary bodies in reflected sunlight (e.g., Mars, comets, outer planets, trans-Neptunian objects, etc.). The NIR infrared continuum observed from Mars is a combination of reflected sunlight (with Fraunhofer lines) and planetary thermal emission (featureless continuum). Sparse spectral lines of Mars’ atmospheric constituents are superposed on the continua according to the optical path experienced by the two components. Sunlight experiences a double optical path (Sun-to-surface + surface-to-Observer), while the Mars “thermal” continuum traverses only a single path (surface-to-observer). We determine the relative contributions of solar and thermal emission to the measured continuum by comparing the measured area (equivalent width) of Fraunhofer lines with their true value. This permits identification of the true reflected solar (and Mars thermal) continuum level needed to properly determine molecular abundances on Mars, and thus an accurate solar model has impact not only on the spectral residuals but also on the retrieval process.

The Solar spectrum in the infrared has been characterized recently with superior accuracy, although the latest ATMOS atlas (Abrams et al. 1996) has limited Signal-to-Noise-Ratio (SNR ~900 in the L-band) and the ACE spectrum (Hase et al. 2009) suffers from certain offset/gain inaccuracies, restricting the maximum SNR of our reduced spectra. Moreover, these atlases describe the solar disk-center and are not applicable for the sunlight from the integrated disk area. To provide an accurate representation of the Solar spectrum, we recently combined empirical and theoretical atlases to obtain an improved model for the solar spectrum (see Appendix B of Villanueva et al. 2011a). Our model includes the broadening effects introduced by differential Solar rotation and limb darkening, ultimately leading to a highly realistic characterization of the Solar features imprinted in the Mars continua.

Development of quantum mechanical models and application to cometary data

The vibrational C-H stretch region (~3.3 μm) is particularly bright in hydrocarbon-rich comets where efficient solar pumping (and inefficient collisional quenching) leads to strong fluorescent emission. Collision partners in cometary atmospheres usually lack sufficient energy to excite vibrational transitions, and the rate of quenching collisions is much smaller than radiative decay rates for (infrared active) excited states. For these reasons, the vibrational manifold is not significantly populated in LTE. Instead, solar radiation pumps the molecules into excited vibrational states, which subsequently de-excite by rapid radiative decay. Infrared photons are emitted through decay to the ground vibrational state, either directly (through resonant fluorescence) or through branching into intermediate vibrational levels (non-resonant fluorescence). Resonant fluorescence is expected to be the dominant factor in the excitation and is the sole pumping mechanism we consider here, although additional excitation cascading from levels with energies higher than n₃ may also be active. As states above, for a realistic description of the Solar pump, we developed a method that combines the accurate Hase et al. 2006 solar atlas, the continuum model by Kurucz 2009, and Doppler broadening introduced by the differential rotation of the Sun (see Appendix B of Villanueva et al. 2011).

To interpret high-resolution spectra of comets:

• We constructed a line-by-line model for the ν₇ band of ethane (C₂H₆), and applied it to compute telluric and Martian transmittances, including cometary fluorescence efficiencies.
• We constructed a line-by-line model for the ν₃ and ν₂ bands of methanol (CH₃OH), and applied it to compute telluric and Martian transmittances, including cometary fluorescence efficiencies.
• We developed a modern method to synthesize fluorescence efficiencies in comets. The new tool is readily available to compute cometary fluorescence emission rates for multiple molecules (e.g. C₂H₆, H₂O, HDO, NH₃, HCN, C₂H₂, CH₄, H₂CO, CO).
• We compiled spectral information for H₂O and HDO using 5 databases (BT2, VTT, HITEMP, HITRAN and GEISA), and created the most complete and accurate spectral database available for these isotopomers. Using these databases, we have computed highly precise terrestrial transmittance models using the advanced LBLRTM model, and obtained Martian residuals of high sensitivity. We compared the performance of the new database to spectral data of Mars taken with CSHELL/IRTF in 2008, NIRSPEC/Keck-2 in 2006 and using CRIRES/VLT in 2009. We extracted D/H ratios of water on Mars.
• We compiled spectral information NH₃ using 4 databases (BYT2, TROVE, HITRAN and GEISA), and created an extremely complete and accurate spectral database available for this molecule. We integrated these new developments into our radiative transfer models, computed cometary fluorescence efficiencies and extracted extremely sensitive searches for NH₃ on Mars.
• We characterized the chemical composition of comet C/2007 W1 (Boattini) using the long-slit echelle grating spectrograph at Keck-2 (NIRSPEC) on 2008 July 9 and 10. We sampled 11 volatile species, and retrieved three important cosmogonic indicators: the ortho-para ratios of H₂O and CH₄, and an upper-limit for the D/H ratio in water. The abundance ratios of almost all trace volatiles (relative to water) are among the highest ever observed in a comet. The comet also revealed a complex outgassing pattern, with some volatiles (the polar species H₂O and CH₃OH) presenting very asymmetric spatial profiles (extended in the anti-sunward hemisphere), while others (e.g., C₂H₆ and HCN) showed particularly symmetric profiles.