Monte Carlo modeling of sodium in Mercury’s exosphere during the first two MESSENGER flybys

We present a Monte Carlo model of the distribution of neutral sodium in Mercury’s exosphere and tail using data from the Mercury Atmospheric and Surface Composition Spectrometer (MASCS) on the MErcury Surface, Space ENvironment, GEochemistry, and Ranging (MESSENGER) spacecraft during the first two flybys of the planet in January and September 2008. We show that the dominant source mechanism for ejecting sodium from the surface is photon-stimulated desorption (PSD) and that the desorption rate is limited by the diffusion rate of sodium from the interior of grains in the regolith to the topmost few monolayers where PSD is effective. In the absence of ion precipitation, we find that the sodium source rate is limited to ∼10⁶-10⁷ cm⁻² s⁻¹, depending on the sticking efficiency of exospheric sodium that returns to the surface. The diffusion rate must be at least a factor of 5 higher in regions of ion precipitation to explain the MASCS observations during the second MESSENGER flyby. We estimate that impact vaporization of micrometeoroids may provide up to 15% of the total sodium source rate in the regions observed. Although sputtering by precipitating ions was found not to be a significant source of sodium during the MESSENGER flybys, ion precipitation is responsible for increasing the source rate at high latitudes through ion-enhanced diffusion.     

Orbital effects on Mercury’s escaping sodium exosphere

We present results from coronagraphic imaging of Mercury’s sodium tail over a 7° field of view. Several sets of observations made at the McDonald Observatory since May 2007 show a tail of neutral sodium atoms stretching more than 1000 Mercury radii (R_m) in length, or a full degree of sky. However, no tail was observed extending beyond 120 R_m during the January 2008 MESSENGER fly-by period, or during a similar orbital phase of Mercury in July 2008. Large changes in Mercury’s heliocentric radial velocity cause Doppler shifts about the Fraunhofer absorption features; the resultant change in solar flux and radiation pressure is the primary cause of the observed variation in tail brightness. Smaller fluctuations in brightness may exist due to changing source rates at the surface, but we have no explicit evidence for such changes in this data set. The effects of radiation pressure on Mercury’s escaping atmosphere are investigated using seven observations spanning different orbital phases. Total escape rates of atmospheric sodium are estimated to be between 5 and 13 × 10²³ atoms/s and show a correlation to radiation pressure. Candidate sources of Mercury’s sodium exosphere include desorption by UV sunlight, thermal desorption, solar wind channeled along Mercury’s magnetic field lines, and micro-meteor impacts. Wide-angle observations of the full extent of Mercury’s sodium tail offer opportunities to enhance our understanding of the time histories of these source rates.     

Spatial variations of the sodium/potassium ratio in Mercury’s exosphere uncovered by high-resolution spectroscopy

High-resolution spectroscopy of Mercury has been obtained with two different instruments in 2006: the EMMI instrument at the 3.6-m NTT telescope of ESO La Silla Chile and the ESPADON spectrograph at the 3.6-m CFHT telescope on top of Mauna Kea (Hawaii). The disk of the planet has been scanned for spatial variation of the exospheric species. The large spectral range and high resolution allow simultaneous measurements of the integrated column density of Na and K.We measure Na/K ratio between 80 and 400 with values between 60 and 90 when the telescope was pointed towards the subsolar region of Mercury’s disk and much larger value when we looked to other part of the exosphere. Moreover, we observed that the Na and K exospheres display very different spatial distributions. Even if these two species are probably ejected with very similar mechanisms from the surface, their differences in mass and sensitivity to solar pressure acceleration imply very different behavior in Mercury’s exosphere.     

Self-consistent modelling of Mercury’s exosphere by sputtering, micro-meteorite impact and photon-stimulated desorption

A Monte-Carlo model of exospheres (Wurz and Lammer, 2003) was extended by treating the ion-induced sputtering process, photon-stimulated desorption, and micro-meteorite impact vaporisation quantitatively in a self-consistent way starting with the actual release of particles from the mineral surface of Mercury. Based on available literature data we established a global model for the surface mineralogy of Mercury and from that derived the average elemental composition of the surface. This model serves as a tool to estimate densities of species in the exosphere depending on the release mechanism and the associated physical parameters quantitatively describing the particle release from the surface.Our calculation shows that the total contribution to the exospheric density at the Hermean surface by solar wind sputtering is about 4×10⁷ m^-3, which is much less than the experimental upper limit of the exospheric density of 10¹² m^-3. The total calculated exospheric density from micro-meteorite impact vaporisation is about 1.6×10⁸ m^-3, also much less than the observed value. We conclude that solar wind sputtering and micro-meteorite impact vaporisation contribute only a small fraction of Mercury’s exosphere, at least close to the surface. Because of the considerably larger scale height of atoms released via sputtering into the exosphere, sputtered atoms start to dominate the exosphere at altitudes exceeding around 1000 km, with the exception of some light and abundant species released thermally, e.g. H₂ and He. Because of Mercury’s strong gravitational field not all particles released by sputtering and micro-meteorite impact escape. Over extended time scales this will lead to an alteration of the surface composition.     

Identification and measurement of neutron-absorbing elements on Mercury’s surface

MESSENGER Neutron Spectrometer (NS) observations of cosmic-ray-generated thermal neutrons provide the first direct measurements of Mercury’s surface elemental composition. Specifically, we show that Mercury’s surface is enriched in neutron-absorbing elements and has a measured macroscopic neutron-absorption cross section of 45-81 × 10⁻⁴ cm²/g, a range similar to the neutron absorption of lunar basalts from Mare Crisium. The expected neutron-absorbing elements are Fe and Ti, with possible trace amounts of Gd and Sm. Fe and Ti, in particular, are important for understanding Mercury’s formation and how its surface may have changed over time through magmatic processes. With neutron Doppler filtering - a neutron energy separation technique based on spacecraft velocity - we demonstrate that Mercury’s surface composition cannot be matched by prior models, which have characteristically low abundances of Fe, Ti, Gd, and Sm. While neutron spectroscopy alone cannot separate the relative contributions of individual neutron-absorbing elements, these results provide strong new constraints on the nature of Mercury’s surface materials. For example, if all the measured neutron absorption were due to the presence of an Fe-Ti oxide and that oxide were ilmenite, then Mercury’s surface would have an ilmenite content of 7-18 wt.%. This result is in general agreement with the inference from color imaging and visible-near-infrared spectroscopy that Mercury’s overall low reflectance is consistent with a surface composition that is enriched in Fe-Ti oxides. The incorporation of substantial Fe and Ti in oxides would imply that the oxygen fugacity of basalts on Mercury is at the upper range of oxygen fugacities inferred for basalts on the Moon.     

Mercury and Moon He exospheres: Analysis and modeling

Helium is one of the first elements clearly identified in the lunar exosphere (Hoffman, J.H., Hodges, R.R., Johnson, F.S., Evans, D.E. [1973]. Proc. Lunar Sci. Conf. 3, 2865–2875). Apollo 17 measured the He density at the surface during four lunations. It confirmed the expected day to night asymmetry of the He exosphere with a maximum density near the dawn terminator on the nightside. Few years later, the first detection of Mercury’s He exosphere was successfully obtained by Mariner 10 (Broadfoot, A.L., Shemansky, D.E., Kumar, S. [1976]. Geophys. Res. Lett. 3, 577–580). These observations highlighted similar global distribution of the He exosphere at Mercury and at the Moon, but also significant differences that have never been convincingly explained.In this paper, we model the He exosphere at the Moon and Mercury with the same approach. The energy accommodation of the exospheric He particles interacting with the surface can be roughly constrained using Apollo 17 and Mariner 10 measurements. Neither a low energy accommodation, as suggested by Shemansky and Broadfoot (Shemansky, D.E., Broadfoot, A.L. [1977]. Rev. Geophys. 15, 491–499), nor a full energy accommodation, as suggested by Hodges (Hodges Jr., R.R. [1975]. The Moon, 14, 139–157), can fit all the observations. These observations and their modeling suggest a diurnal variation of the energy distribution of the He ejected from the surface that cannot be explained satisfactorily by any of the present theories on the gas–surface interaction in surface-bounded exospheres.     

Whole-disk spectrophotometric properties of Mercury: Synthesis of MESSENGER and ground-based observations

Disk-integrated and disk-resolved measurements of Mercury’s surface obtained by both the Mercury Dual Imaging System (MDIS) and the Mercury Atmospheric and Surface Composition Spectrometer (MASCS) onboard the MErcury Surface, Space ENvironment, GEochemistry, and Ranging (MESSENGER) spacecraft were analyzed and compared with previous ground-based observations of Mercury at 11 wavelengths. The spectra show no definitive absorption features and display a red spectral slope (increasing reflectance with increasing wavelength) typical of space-weathered rocky surfaces. The MDIS spectra show evidence of phase reddening, which is not observed in the MASCS spectra. The MDIS spectra are commensurate with ground-based observations to within 10%, whereas the MASCS spectra display greater discrepancies with ground-based observations at near-infrared wavelengths. The derived photometric calibrations provide corrections within 10% for observations taken at phase angles less than ∼100°. The derived photometric properties are indicative of a more compact regolith than that of the lunar surface or of average S-type asteroids. The photometric roughness of the surface is also much smoother than the Moon’s. The calculated geometric albedo (reflectance at zero phase) is higher than lunar values. The lower reflectance of immature units on Mercury compared with immature units on the Moon, in conjunction with the higher geometric albedo, is indicative of more complicated grain structures within Mercury’s regolith.     

Observations of metallic species in Mercury’s exosphere

From observations of the metallic species sodium (Na), potassium (K), and magnesium (Mg) in Mercury’s exosphere, we derive implications for source and loss processes. All metallic species observed exhibit a distribution and/or line width characteristic of high to extreme temperature - tens of thousands of degrees K. The temperatures of refractory species, including magnesium and calcium, indicate that the source process for the atoms observed in the tail and near-planet exosphere are consistent with ion sputtering and/or impact vaporization of a molecule with subsequent dissociation into the atomic form. The extended Mg tail is consistent with a surface abundance of 5-8% Mg by number, if 30% of impact-vaporized Mg remains as MgO and half of the impact vapor condenses. Globally, ion sputtering is not a major source of Mg, but locally the sputtered source can be larger than the impact vapor source. We conclude that the Na and K in Mercury’s exosphere can be derived from a regolith composition similar to that of Luna 16 soil (or Apollo 17 orange glass), in which the abundance by number is 0.0027 (0.0028) for Na and 0.0006 (0.0045) for K.     

Variation in lunar sodium exosphere measured from lunar orbiter SELENE (Kaguya)

Resonant scattering of the lunar sodium exosphere was measured from the lunar orbiter SELENE (Kaguya) from December 2008 to June 2009. Variations in line-of-sight integrated intensity measured on the night-side hemisphere of the Moon could be described as a spherical symmetric distribution of the sodium exosphere with a temperature of 2400-6000 K. Average surface density of sodium atoms in February is well above that in the other months by about 30%. A clear variation in surface density related to the Moon’s passage across the Earth’s magnetotail could not be seen, although sodium density gradually decreased (by 20±8%) during periods from the first through the last quarter of two lunar cycles. These results suggest that the supra-thermal components of the sodium exosphere are not mainly produced by classical sputtering of solar wind. The variation in sodium density (which depends on lunar-phase angle) is possibly explained by the presence of an inhomogeneous source distribution of photon-stimulated desorption (PSD) on the surface.     

The calcium exosphere of Mercury

A tenuous calcium atmosphere at Mercury, principally seen in the polar regions, was first observed in July, 1998, using the High Resolution Echelle Spectrograph (HIRES) at the W.M. Keck I telescope (Bida et al., Nature 404, 159, 2000). We report four years of observations of the calcium exosphere of Mercury, confirming the initial findings of a very tenuous atmosphere. These observations show a persistent but spatially variable blue shift, indicating an excess velocity toward the observer of up to 3 km s⁻¹, with an average excess velocity of 2.2 km s⁻¹ above the south pole. In addition, the line profiles reveal a hot corona at the equivalent of 12,000-20,000 K in a thermalized atmosphere, indicating a large range of motion with respect to the observer. The calcium is not confined to the polar-regions: rare and low Ca abundance is seen in the equatorial regions. Strong emission was seen anti-sunward on 3 May 2002. Apparent weak emission on the sunward hemisphere may be due to scattered light from the surface, or may indicate a high latitude source. We show that the likely source of the calcium is either impact vaporization in the form of CaO and clusters, which are subsequently photo-dissociated, or ion-sputtering of atoms, molecules and ions. The column abundance is somewhat, but not strongly, correlated with solar activity. We predict a very hot (probably escaping) oxygen component to the hermean exosphere.     

Source dependency of exospheric sodium on Mercury

Due to a large solar radiation effect, the sodium exosphere exhibits many interesting effects, including the formation of an extended corona and a tail-like structure. The current suite of observations allows us to study some physical properties of the sodium exosphere, such as the source rates and the interaction with the surface, both experimentally and theoretically. In order to quantify the complex variations in the sodium exosphere in more detail, we use an exospheric model with the Monte-Carlo method to examine the surface interactions of a sodium atom, including the surface thermal accommodation rate and the sticking coefficient. The source rates from different components, such as the photon stimulated desorption (PSD), the meteoroid impact vaporization (MIV), and the solar wind ion sputtering (IS), can be constrained by comparing our exospheric model calculations with the published observational data. The detected terminator to limb (TL) ratio on the disk and the tail production rate can be explained with no sticking effect and small thermal accommodation rates. We also examine the best fit of the MIV source evolution, through comparison with the disk-averaged emission. The resultant discrepancy between the observations and the model fit may reflect the surface variation in the sodium abundance. A comprehensive mapping of the surface geochemical composition of the surface by the MESSENGER and Bepi-Colombo missions should give us more information about the nature of this surface-bound exosphere.     

Titan’s atomic hydrogen corona

Based on measurements performed by the Hydrogen Deuterium Absorption Cell (HDAC) aboard the Cassini orbiter, Titan’s atomic hydrogen exosphere is investigated. Data obtained during the T9 encounter are used to infer the distribution of atomic hydrogen throughout Titan’s exosphere, as well as the exospheric temperature.The measurements performed during the flyby are modeled by performing Monte Carlo radiative transfer calculations of solar Lyman-α radiation, which is resonantly scattered on atomic hydrogen in Titan’s exosphere. Two different atomic hydrogen distribution models are applied to determine the best fitting density profile. One model is a static model that uses the Chamberlain formalism to calculate the distribution of atomic hydrogen throughout the exosphere, whereas the second model is a Particle model, which can also be applied to non-Maxwellian velocity distributions.The density distributions provided by both models are able to fit the measurements although both models differ at the exobase: best fitting exobase atomic hydrogen densities of n_H = (1.5 ± 0.5) × 10⁴ cm⁻³ and n_H = (7 ± 1) × 10⁴ cm⁻³ were found using the density distribution provided by both models, respectively. This is based on the fact that during the encounter, HDAC was sensitive to altitudes above about 3000 km, hence well above the exobase at about 1500 km. Above 3000 km, both models produce densities which are comparable, when taking into account the measurement uncertainty.The inferred exobase density using the Chamberlain profile is a factor of about 2.6 lower than the density obtained from Voyager 1 measurements and much lower than the values inferred from current photochemical models. However, when taking into account the higher solar activity during the Voyager flyby, this is consistent with the Voyager measurements. When using the density profile provided by the particle model, the best fitting exobase density is in perfect agreement with the densities inferred by current photochemical models.Furthermore, a best fitting exospheric temperature of atomic hydrogen in the range of T_H = (150-175) ± 25 K was obtained when assuming an isothermal exosphere for the calculations. The required exospheric temperature depends on the density distribution chosen. This result is within the temperature range determined by different instruments aboard Cassini. The inferred temperature is close to the critical temperature for atomic hydrogen, above which it can escape hydrodynamically after it diffused through the heavier background gas.     

Observations of Mercury's exosphere: Spatial distributions and variations of its Na component during August 8, 9 and 10, 2003

Following the observations of August 2002 [Barbieri, C., Verani, S., Cremonese, G., Sprague, A., Mendillo, M., Cosentino, R., Hunten, D., 2004. Planet. Space Sci. 52, 1169-1175], the high resolution spectrograph of the 3.5-m Galileo National Telescope (TNG) has been used to obtain several spatially resolved spectra of Mercury's Na-D on the evenings of 8, 9 and 10 August 2003. The resolution of the spectrograph was 115,000, the slit dimensions were ^0.4″×^27″. With respect to Paper I, the planet was in a fairly similar orbital configuration, being at a geocentric distance of approximately 1 AU, and having a True Anomaly Angle (TAA) from 163°-168° instead of 171°-174°. We present here a significantly larger set of observations and discuss several important features regarding the formation of Mercury's sodium exosphere, in particular the role of photon stimulated and thermal desorptions, as well as of the solar wind sputtering and micro-meteoroid vaporization. Thanks to the very good seeing of these observations, we also report and discuss the origins and variations of equatorial structures in Mercury's early morning sodium exosphere.     

Constraints on Mercury’s surface composition from MESSENGER and ground-based spectroscopy

The composition and chemistry of Mercury’s regolith has been calculated from MESSENGER MASCS 0.3-1.3 μm spectra from the first flyby, using an implementation of Hapke’s radiative transfer-based photometric model for light scattering in semi-transparent porous media, and a linear spectral mixing algorithm. We combine this investigation with linear spectral fitting results from mid-infrared spectra and compare derived oxide abundances with mercurian formation models and lunar samples. Hapke modeling results indicate a regolith that is optically dominated by finely comminuted particles with average area weighted grain size near 20 μm. Mercury shows lunar-style space weathering, with maturation-produced microphase iron present at ∼0.065 wt.% abundance, with only small variations between mature and immature sites, the amount of which is unable to explain Mercury’s low brightness relative to the Moon. The average modal mineralogies for the flyby 1 spectra derived from Hapke modeling are 35-70% Na-rich plagioclase or orthoclase, up to 30% Mg-rich clinopyroxene, <5% Mg-rich orthopyroxene, minute olivine, ∼20-45% low-Fe, low-Ti agglutinitic glass, and <10% of one or more lunar-like opaque minerals. Mercurian average oxide abundances derived from Hapke models and mid-infrared linear fitting include 40-50 wt.% SiO₂, 10-35 wt.% Al₂O₃, 1-8 wt.% FeO, and <25 wt.% TiO₂; the inferred rock type is basalt. Lunar-like opaques or glasses with high Fe and/or Ti abundances cannot on their own, or in combination, explain Mercury’s low brightness. The linear mixing results indicate the presence of clinopyroxenes that contain up to 21 wt.% MnO and the presence of a Mn-rich hedenbergite. Mn in M1 crystalline lattice sites of hedenbergite suppresses the strong 1 and 2 μm crystal field absorption bands and may thus act as a strong darkening agent on Mercury. Also, one or more of thermally darkened silicates, Fe-poor opaques and matured glasses, or Mercury-unique Ostwald-ripened microphase iron nickel may lower the albedo. A major part of the total microphase iron present in Mercury’s regolith is likely derived from FeO that is not intrinsic to the crust but has been subsequently delivered by exogenic sources.     

Neutral particle release from Europa’s surface

In this paper, we look at space weathering processes on the icy surface of Jupiter’s moon Europa. The heavy energetic ions of the jovian plasma (H⁺, O⁺, S⁺, C⁺) can erode the surface of Europa via ion sputtering (IS), ejecting up to 1000 H₂O molecules per ion. UV photons impinging the Europa’s surface can also result in neutral atom release via photon-stimulated desorption (PSD) and chemical change (photolysis). In this work, we study the efficiency of the IS and PSD processes for ejecting water molecules, simulating the resulting neutral H₂O density. We also estimate the contribution to the total neutral atom release by the Ion Backscattering (IBS) process. Moreover, we estimate the possibility of detecting the sputtered high energy atoms, in order to distinguish the action of the IS process from other surface release mechanisms. Our main results are: (1) The most significant sputtered-particle flux and the largest contribution to the neutral H₂O density come from the incident S⁺ ions; (2) the H₂O density produced via PSD is lower than that due to sputtering by ∼1.5 orders of magnitude; (3) in the energy range below 1 keV, the IBS can be considered negligible for the production of neutrals, whereas in the higher energy range it becomes the dominant neutral emission mechanism; (4) the total sputtering rate for Europa is 2.0 × 10²⁷ H₂O s⁻¹; and (5) the fraction of escaping H₂O via IS is 22% of the total sputtered population, while the escape fraction for H₂O produced by PSD is 30% of the total PSD population. Since the PSD exosphere is lower than the IS one, the major agent for Europa’s surface total net erosion is IS on both the non-illuminated and illuminated side. Lastly, the exospheric neutral density, estimated from the Galileo electron density measurements appears to be higher than that calculated for H₂O alone; this favors the scenario of the presence of O₂ produced by radiolysis and photolysis.     

Sodium winds on Mercury

Solar radiation acceleration imparts anti-sunward velocities to sodium atoms in the Mercury exosphere. The Earthward-directed vectors of the Sun-accelerated atom velocities can be observed from Earth as small Doppler shifts, either added to, or subtracted from the Earth-Mercury Doppler shifts. We measured these small Doppler shifts using high resolution spectrographs capable of detecting sodium velocity differences as small as 0.1 km/s. We report here four sets of observations performed at different Mercury true anomaly angles. For these measurements, the spectrograph slit was oriented first east-west, and then north-south on the planet so as to get east-west and north-south transects of the velocities. The velocity patterns in east-west transects could be explained in terms of sodium flows outwards from the subsolar point, except for unexpectedly large Earthward velocities observed above the dawn terminator, which we interpreted to be the result of evaporation of sodium as the cold surface is heated by the rising Sun. North-south transects also showed a general pattern consistent with sodium flows outwards from the subsolar point. However, in all cases, the velocities were higher in one hemisphere relative to the other. For two cases, excess sodium emission was observed in the same hemisphere as the velocity excess. We interpreted these results to mean that there existed sources of sodium at high latitudes, which could appear in either hemisphere.     

Mercury surface composition: Integrating petrologic modeling and remote sensing data to place constraints on FeO abundance

A significant opaque component in Mercury’s crust is inferred based on albedo and spectral observations. Previous workers have favored iron-titanium bearing oxide minerals as the spectrally neutral opaque. A consequence of this hypothesis is that Mercury’s surface would have a high FeO content. An array of remote sensing techniques have not provided definitive constraints on the FeO content of Mercury’s surface. However, spectral observations have not detected a diagnostic 1 μm absorption band and have thus limited the FeO in coexisting silicates to <2 wt.% FeO. In this paper, we assess equilibrium among oxide and silicate minerals to constrain the distribution of iron between opaque oxides and silicates under a variety of environmental conditions. Equilibrium modeling is favored here because the geologic process that produced Mercury’s low-albedo intermediate terrain must have occurred globally, which favors a common widespread igneous process. Based on our modeling, we find that iron-rich ilmenite cannot occur with silicates that do not display a 1 μm absorption feature unless plagioclase abundances are high. However, such high plagioclase abundances are precluded by Mercury’s low albedo. Incorporating equilibrium crystallization modeling with spectral and albedo constraints we find the iron abundance of Mercury’s intermediate terrain is ?10 wt.% FeO. This intermediate iron composition matches constraints provided by visible albedo and total neutron absorption observed by MESSENGER. In fact, the total neutron absorption of mixtures of oxide, plagioclase, olivine and pyroxene for the oxide abundances estimated for Mercury, favor Mg-rich members of the ilmenite-geikielite solid-solution series. This work offers compositional constraints for Fe, Ti, and Mg that will be testable by various MESSENGER instrument data sets after it begins its orbital mission.     

Investigation of local time dependence of Mercury's sodium exosphere based on a numerical simulation

Mercury has a surface-bounded exosphere (SBE) similar to that of the Moon. One of the atmospheric species, sodium, was found by ground-based observations to be the most prominent component. Mercury's sodium SBE is known to be non-uniform with respect to local time (LT) in low-latitude regions: the sodium column density in the dawn-side region is larger than that in the dusk-side region, and the sodium abundance is the largest in the morning-noon region. To reveal the production processes for the exosphere near Mercury's surface, the LT dependence of the exosphere was investigated through a numerical simulation. Three data sets of sodium column densities observed for the dawn-side hemisphere, observed by Sprague et al. [1997. Distribution and abundance of sodium in Mercury's atmosphere, 1985-1988. Icarus 12, 506-527], were compared with results simulated by a 3D Monte Carlo method, and the source rates and density of sodium of the planetary surface were estimated. In the simulation, the photon-stimulated desorption (PSD) and thermal desorption (TD) processes were assumed as the release mechanisms. The sodium source rates for the three data sets, at respective heliocentric distances of about 0.33, 0.42, and 0.44 AU, were estimated as 1-4×10⁸ Na/cm²/s with weak LT dependence. In contrast, the expected sodium surface density showed clear dependence on LT and the heliocentric distance. The sodium surface density decreases from early morning to noon by a few orders, and, particularly for large heliocentric distances, the surface is in a condition of sodium excess and depletion with respect to the surface sodium density assumed by Killen et al. [2004. Source rates and ion recycling rates for Na and K in Mercury's atmosphere. Icarus 171, 1-19] in the early morning and morning-noon regions, respectively. This study implies that the decrease in sodium surface density from the early morning to noon regions might produce the characteristic LT dependence in the low-latitude dawn-side region.