The channels in UVIS can operate in diﬀerent configurations and this results in diﬀerent types of data products. The result of the measurements performed during a time range and with a particular instrument configuration is called observation, and has a unique identifier. The data in a UVIS obser-vation are a copy of unprocessed data in the UVIS memory buﬀer, in binary format. An observation corresponds to one of four diﬀerent types of data product: a spectrum, a time series of spatial-spectral images, a time series of detector counts, or an image at one wavelength. One or more of these data products and associated ancillary data products (1 ancillary product for many data products is possible) are stored in a data volume. Several volumes form a data set. Data products conform to Planetary Data System (PDS) standards and therefore have data objects and an associated PDS object label, which completely defines the product. This includes instrument configuration information, window and binning (see below) and integration time specifications. For a more detailed description of the diﬀerent product types the reader is referred to the data documentation in the PDS. The observations dealt with in this work correspond to the time series of spatial-spectral images product type. This type of data is called spatial spectral cube. For a description of the spatial spectral cubes and their processing see Esposito et al. (2004) and Capalbo (2010).
A UVIS spatial spectral cube is a time ordered sequence of 1024 x 64 matrices. Each element of a matrix is the number of detector counts for an individual detector pixel during a fixed time interval or integration time. In more complex cases, to reduce data size, windowing and/or a binning of the spatial and/or spectral dimensions are defined in the 1024 x 64 matrices. See Figure 2.3 for a representation of data from an observation in the form of a PDS cube. In these cases, the detector is divided into a set of active rectangular sub-regions (windows). Each window can also be binned. The result is that some integers in the cube correspond to a range or ‘window’ of detector cells and their values are derived by summing over the spatial and/or spectral dimensions accordingly to the windowing and binning. The resulting data are located in the upper-left corner of the defined windows, and the rest of the locations have non-valid values. For a detailed description of the format see the PDS data documentation.
The knowledge of the geometry configuration during observations is important for interpreting the data. It permits the derivation of useful quantities like the tangent altitude corresponding to a particular measurement time, and the corresponding planetary coordinates. Ancillary information about the instrument, spacecraft, planets, etc., are essential to derive geometry configurations. This information includes pointing, position and attitude of the instrument and spacecraft, position and motion of the sources of radiation, the target objects, angles and distances to calculate parameters that determine the characteristics of an observation, etc. The Navigation and Ancillary Information Facility (NAIF) provides an information system for this and other purposes. The SPICE system assists scientists in planning and interpreting scientific observations from space-based instruments. A description of SPICE and its use in this work can be found in Acton (1996), Capalbo (2010), and references therein. SPICE provides the ancillary data needed for the calculations of necessary information (like geometry) and also a toolkit to process the ancillary data. The primary SPICE data sets are called ”kernels”. SPICE kernels are composed of navigation and other ancillary information that has been prepared to be easily used by the planetary science and engineering communities. They should include or be accompanied by metadata that provide pedigree and other descriptive information needed. The SPICE system includes the SPICE Toolkit, a large collection of allied software. The principal component of this Toolkit is a library of portable subroutines needed to read the kernel files and to calculate observation geometry parameters of interest. Users can integrate these SPICE Toolkit subroutines into their own application programs. The SPICE Toolkit was originally implemented in ANSI FORTRAN 77, but is now available in C, IDL and MATLAB as well. Some of the routines in the Toolkit IDL version, called ICY, were used in the present work to compute geometry related quantities.
Absorptive UV occultations
Several physical and chemical processes can take place in an atmosphere. An important factor driving those processes is the energy input/output. In the case of the upper atmosphere, this input can be due to radiation from stars (mainly from the Sun for Solar System bodies), the interstellar medium (γ and X-rays), or energetic particles from the interstellar or interplanetary medium (e- and ions from surrounding bodies, protons from stellar wind).
Lavvas et al. (2011) comment on several of the sources mentioned above for the case of Titan. Solar photons dominate energy deposition and ionization/dissociation processes on the day side (Galand et al., 2010). Some of these photons will generate photoelectrons that will provide a smaller but important contribution to ionization and dissociation processes. Magnetospheric electrons have in general a smaller contribution than photons. Although the former can dominate ionization on the night side in some occasions (Cravens et al., 2009), their importance is always less than solar ionization on the day side. The contribution to neutral dissociation by energetic ions (O+, H+) in 500 – 1000 km is small under typical Titan conditions. The altitude of maximum energy deposition of pickup ions is in general above the solar EUV/UV energy deposition (as shown in Figure 5 of Westlake et al., 2011). There is a contribution to ionization in the 500 – 700 km by meteoroids ablation, but small. Galactic cosmic rays contribute to ionization in the lower atmosphere, with a peak at 65 km. Other sources of energy for the upper atmosphere can be particles from the solar wind, and breaking waves from the mesosphere into the thermosphere (M¨uller-Wodarg et al., 2008).
The characteristics of the atmosphere can be elucidated by studying its interaction with these sources. In this work the upper atmosphere of Titan is studied by means of its interaction with the radiation received from the Sun or other stars. These photons can interact with the atmosphere, more precisely exciting the molecules composing it. The interaction can be called:
absorption: the energy from the photons in the radiation field can be converted to internal energy leading to molecular excited states (electronic, vibrational, rotational). Some of the electronic levels can lead to dissociation or ionization. Part of the energy absorbed can also be converted to kinetic energy, increasing the temperature of the medium.
scattering: the molecule returns spontaneously to the original energy state, re-emitting the pho-ton. This emission can be of the same wavelength (scattering) or of diﬀerent wavelengths (Raman scattering). In the general case the intensity of the re-emitted radiation depends on the angle of emission; this distribution of energy depends in turn on the relative size of the molecule (or particle) with respect to the incident wavelength. Particular cases occur when the direction of emission is the same as that of incidence (forward scattering), opposite (backward scattering), or following the laws of reflection (diﬀuse reflection or just scattering).
Other processes in the atmosphere manifest as emission from it, for example: thermal emission: the energy gained from collisions (kinetic energy) are emitted as photons. airglow: emission of radiation caused by diﬀerent processes. These can be chemiluminescence, in which photons are emitted by the recombination of ionized particles, or by the reactions between certain molecules; or photoluminescence (fluorescence and phosphorescence), which is an energy release mechanism similar to molecular scattering but in which longer times are involved between absorption and emission. Apart from the phenomena just described, the radiation suﬀers from diﬀraction when traversing the atmosphere, which has diﬀerent refractive index at diﬀerent heights. This produces a deviation in the direction of the radiation. The combination of these phenomena determines the interaction of the atmosphere with the radiation received.
The eﬀects of radiation propagating through a medium can be modeled with the radiative transfer equation (see for example Rees, 2001). The spectral radiance L, in Watts per meter square per steradian, for a particular wavelength λ and propagating in a direction (θ, φ) is obtained solving the equation dL(λ, θ, φ) = −(αa(λ, T ) + αs(λ, T ))L(λ, θ, φ)R(λ) + αsJ(λ) + αaB(λ, T ). (2.1).
Here dD measures distance in the propagation direction, αa and αs are the absorption and scattering coeﬃcients, and R accounts for diﬀerential refractivity. The spectral radiance coming from emission from the medium is B. Equation (2.2) defines J, the radiation scattered into the direction of propa-gation from other directions. In this equation the primed angles reference the diﬀerent directions that contribute to the propagation direction. The phase function of the scattering, p(cos(Θ)), describes the angular distribution of the scattered radiation in terms of the angle Θ through which the radiation has been deflected (see Rees (2001) for details). J = 1 Z 4π L(θ′, φ′) p(cos(Θ)) dΩ′. (2.2).
Atmospheric composition from absorptive VUV stellar occultations
The overall picture of the procedure used in this work to derive composition and temperature is schematized in Figure 2.4. As Cassini performs a Titan flyby UVIS performs observations. These can include star occultations, particularly Sun occultations. The raw UVIS data are combined with ancillary data to calculate transmission as a function of altitude. At the same time absorption cross sections, model profiles, and the UVIS instrument function are combined according to analysis needs. Then the column densities and number densities are derived, and temperatures are calculated. The de-tailed procedure and particularities for stellar and solar occultations will be described in the remaining of this chapter.
During an occultation the radiation from the source is measured before and during the time that the Line Of Sight (LOS) to the source traverses the atmosphere. From all the wavelengths that can be used, VUV is very popular because many gases present strong absorption features in this wavelength region. Almost all electronic transitions of molecules give rise to spectra in the visible and ultraviolet regions; only very few extend into the infrared region. Absorption is the dominant phenomena for the observations analyzed here, although extinction from aerosols is also considered. Information from UV absorption corresponds to the upper atmosphere, as the UV radiation is completely absorbed for the lower layers. Equation 2.1 can be simplified for the case of VUV occultations by Titan’s atmosphere observed by UVIS. The emissions in the UV from Titan’s atmosphere can be neglected when observing a star (see Koskinen et al., 2011 and Ajello et al., 2007, 2008), so the third term vanishes. Considering the Field of View (FOV) of the UVIS spectrographs (see Section 2.1) and the characteristics of occultations observed from an orbiting spacecrafts (for example, the distances involved), scattered radiation into the instrument FOV can be neglected (Smith and Hunten, 1990). Therefore the second term disappears, as well as the angular dependence. Finally, refractions eﬀects are negligible (Smith and Hunten, 1990). So, expressing the radiation flux as an intensity I in Rayleigh (instead of radiance L1), equation (2.1) can be written as dI(λ) = −αI(λ). (2.3).
Integrating equation (2.3) an expression for the attenuation of the radiation through the atmosphere can be reached, as shown in equation (2.4), I(λ) = exp(− Z αdD) = (λ), (2.4).
where I0 represents the intensity of radiation outside the atmosphere. The intensity measured above the atmosphere gives the source spectrum, the spectra observed while the LOS traverses the atmo-sphere, I, is the source spectrum modified by the extinction of the diﬀerent components of the atmo-sphere. is the transmission through the atmosphere, measured for the diﬀerent altitudes probed by the LOS. The integral extends over the whole path of the radiation through the atmosphere. This path varies with the altitude from the surface. The extinction coeﬃcient, α, is related to the extinction cross section σ via equation (2.5), α = σn, (2.5).
Stellar occultations measured by UVIS FUV
As shown in previous sections, stellar occultations are a powerful technique to identify molecules and to derive their vertical profiles in planetary upper atmospheres. Many molecules in Titan’s atmosphere present strong absorption features in the FUV region and can therefore be detected by UVIS occultations. From the many molecules measured or predicted in Titan’s upper atmosphere, here the focus is on CH4, C2H2, HCN, C2H4, C4H2, C6H6, HC3N, CH3, and aerosols (AER). These were chosen based on an analysis of theoretical detectability of diﬀerent species (see Chapter 5) and the visual inspection of spectra measured during occultations. This type of occultation can cover altitudes from 500 to 1300 km, which include several interesting regions in the middle and upper atmosphere. The altitudes below the homopause, which in Titan can be placed around 850 -1000 km (Cui et al., 2009; Vervack et al., 2004), include the region where a detached aerosol layer of varying height was observed (see for example Lavvas et al., 2010; West et al., 2011). The region above the homopause is a photochemically active region for several hydrocarbons and nitriles. Finally, the upper layers probed include the ionosphere, where the neutral abundances are determined in part by ion-chemistry. Therefore, diﬀerent layers with diﬀerent physics and chemistry are represented in a profile from a FUV occultation.
Stellar UVIS/FUV occultations by Titan were first presented by Shemansky et al. (2005). From the occultations measured during the flyby called Tb, they derived column density profiles for CH4, C2H2, HCN, C2H4, C2H6, and C4H2. The observations cover altitudes from 450 to 1600 km above the surface. They observed a change in scale height of the retrieved hydrocarbons at 700 and 1000 km. Below 600 km, the density of the species mentioned drops precipitously to undetectable levels, to mixing ratios four to six orders of magnitude below that of CH4. This result is in strong disagreement with inferred measured densities of higher order hydrocarbons and HCN from the Voyager experiment (Vervack et al., 2004) below 600 km. It should also be noted that the C4H2 absorption cross sections used by Shemansky et al. (2005) were saturated at 1445 ˚A and 1645 ˚A, which might have led to an overestimation of the abundance for this species. From the measured CH4 abundances they inferred a temperature profile above 400 km. The profile contains a convective region around 450 km where the temperature gradient is adiabatic and a mesopause at 615 km with a temperature of 114 K. The asymptotic kinetic temperature at the top of the atmosphere determined from this experiment is 151 K. By comparison, the HASI temperature was ∼ 175 K there. Liang et al. (2007) used the same UVIS data from the Tb flyby to detect and characterize aerosols at altitudes between 500 and 1000 km. The detection of aerosols in this region implies that they are formed at higher altitudes in the thermosphere.
Stellar occultations by Titan’s upper atmosphere observed by UVIS/FUV were also analyzed by Koskinen et al. (2011). These authors presented an analysis of other stellar occultations measured by UVIS: T53 and T41. The presence and profiles of some hydrocarbons in the upper atmosphere was then established, adding C6H6 and HC3N to the list of species detected by UVIS presented in Shemansky et al. (2005). Koskinen et al. identified absorption layers at 600 – 700 km dominated by hydrocarbons and nitriles and an absorption layer at 500 km dominated by aerosols. This last layer was less evident in the occultation occurring at higher northern latitudes among the two analyzed. From their CH4 profile they inferred temperature profiles for the upper atmosphere and presented an analysis of the temperature fluctuations in terms of gravity waves.
The results from the works commented are an example of the already highlighted potential of this observations to contribute to the understanding of Titan’s upper atmosphere variability and dynamics. The present work includes a thorough analysis and characterization of the retrieval techniques to derive composition from UVIS/FUV stellar occultations. The technique was applied to data from flybys T41 and T53 with the intention to compare results with those from Koskinen et al. (2011), obtained with an almost identical technique. This work resulted in profiles in good agreement with those in the reference mentioned.
UVIS/FUV stellar occultation analysis
The geometric calculations and data handling procedures common to both channels were described in Section 2.2; results and particularities in the data reduction for the FUV are presented in this subsection, the relevant calculated quantities can be consulted in Table 2.4. The overall picture of the procedure used in this work to derive composition from stellar occultations is schematized in Figure 2.7. The pointing direction to the source in these cases is calculated from the solar system.
Table of contents :
1.1 Titan’s atmosphere
1.1.1 A short tale about an intriguing atmosphere
1.1.2 General description – state of knowledge
1.2 Cassini investigations
1.3 Motivations and focus of the present work
2 UVIS stellar and solar occultation data analysis
2.1 UVIS instrument and data
2.1.1 UVIS instrument
2.1.2 UVIS data
2.1.3 Ancillary data
2.2 Absorptive UV occultations
2.2.1 Atmospheric composition from absorptive VUV stellar occultations
2.2.2 First steps in occultation analysis
2.2.3 Stellar vs. solar occultations
2.3 Stellar occultations measured by UVIS FUV
2.3.1 Introduction to stellar occultations measured in the FUV
2.3.2 UVIS/FUV stellar occultation analysis
2.4 Solar occultations measured by UVIS EUV
2.4.1 Introduction to solar occultations measured in the EUV
2.4.2 UVIS/EUV solar occultation analysis
2.5 Temporal/spatial coverage of data
2.6 Summary of chapter
3 Retrieval methods
3.1 Column density retrieval, stellar occultations
3.2 Column density retrieval, solar occultations
3.3 Altitude range for column densities
3.4 Number density and aerosol extinction retrieval
3.5 Temperature calculations
3.6.1 Choice of Species
3.6.2 Simulation of UVIS data and column density retrieval
3.6.3 Simulation of molecular number density and aerosol extinction retrieval
3.7 Summary of chapter
4 Benzene ACS in the VUV, relevance for Titan
4.1 An Introduction to Benzene on Titan
4.2 Introduction to benzene absorption cross sections
4.3 Experimental measurements of benzene absorption cross section
4.3.1 Absorption measurements with Synchrotron radiation, BESSY II facility
4.3.2 Absorption measurements with a deuterium lamp source, Meudon Observatory facility
4.3.3 Absorption cross section derivation and uncertainties
4.4 Absorption cross section results and interpretation
4.4.1 Temperature variations
4.5 Summary of chapter
5 Atmospheric composition and variability
5.1 Column density and number density profiles
5.1.1 Minor constituents abundance, from FUV stellar occultations
5.1.2 Major constituents abundance, from FUV stellar and EUV solar occultations .
5.2 Temperature and variability
5.3 Summary of chapter
6 Conclusions and further work
A Temperature calculations
B MPFIT, dependence of column density retrieval on initial guess
C MPFIT, uncertainties in column density retrieval