Dusk Over Dawn O{}_{2} Asymmetry in Europa’s Near-Surface Atmosphere

Dusk Over Dawn O Asymmetry in Europa’s Near-Surface Atmosphere

Apurva V. Oza , Francois Leblanc , Robert E. Johnson, Carl Schmidt , Ludivine Leclercq , Timothy A. Cassidy , Jean-Yves Chaufray Physikalisches Institut, Universität Bern, Bern, Switzerland
LATMOS/IPSL, UPMC Univ. Paris 06 Sorbonne Universités, UVSQ, CNRS, Paris, France
Engineering Physics, University of Virginia, Charlottesville, Virginia, USA
Physics, New York University, New York, NY, USA
Center for Space Physics, Boston University, Boston, USA
Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, USA
Resumen

The evolution of Europa’s water-product exosphere over its 85-hour day, based on current models, has not been shown to exhibit any diurnal asymmetries. Here we simulate Europa’s exosphere using a 3-D Monte Carlo routine including, for the first time, the role of Europa’s rotation on the evolution of exospheric molecules tracked throughout the orbit. In this work we focus on understanding the behavior of a single atmospheric constituent, O, sputtered by a trailing hemisphere source with a temperature-dependence under isotropic plasma conditions as also modeled by previous works. Under rotation, the O is also subject to the centrifugal and Coriolis forces in addition to the standard gravitational forces by Jupiter and Europa in our model. We find that the O component, while global, is not homogeneous in Europa local time. Rather, the O consistently accumulates along the direction of Europa’s rotation at the dusk hemisphere. When rotation is explicitly excluded in our simulations, no diurnal asymmetries exist, and any accumulation is due to the prescribed geometry of the sputtering source. We find that the assumed thermal-dependence on the O source is critical for a diurnal asymmetry: the diurnal surface temperature profile is imprinted on to the near-surface O atmosphere, due to small hop times for the non-adsorbing O, which then effectively rotates with Europa. Simulation tests demonstrate that the diurnal asymmetry is not driven by the thermal inertia of the ice, found to have only a weak dependence (). Altogether, the various test cases presented in this work conclude that the dusk-over-dawn asymmetry is driven by Europa’s day-night O cycle synchronized with Europa’s orbital period based on our model assumptions on O production and loss. This conclusion is in agreement with the recent understanding that a non-adsorbing, rotating O source peaking at noon will naturally accumulate from dawn-to-dusk, should the O lifetime be sufficiently long compared to the orbital period. Lastly we compare hemispherically-averaged dusk-over-dawn ratios to the recently observed oxygen emission data by the Hubble Space Telescope. We find that while the simulations are globally consistent with brighter oxygen emission at dusk than at dawn, the orbital evolution of the asymmetries in our simulations can be improved by ameliorating the O source & loss rates, and possibly adsorption onto the regolith.

keywords:
Exospheres, O, Atmospheres, Aurorae
myfootnotemyfootnotefootnotetext: apurva.oza@space.unibe.ch

1 Introduction

Europa’s molecular oxygen is of keen interest as it has been suggested to be a possible source of 0.1- 100 kg/s of O (Johnson et al., 2003 ; Hand et al., 2007; Greenberg, 2010) to Europa’s putative saltwater ocean, thought to have a chemistry analogous to Earth’s oceans (Vance et al., 2016). This range of oceanic source rates can in principle be connected to the production and dynamics of the tenuous oxygen atmosphere, or exosphere, which we simulate in this work.

The origin of Europa’s water-product exosphere is ultimately linked to its neighboring satellite Io. Since the Voyager spacecraft first confirmed the presence of energetic ions trapped in the Jovian magnetosphere, the exosphere has been modeled to be predominantly generated by the bombardment of oxygen and sulfur ions ultimately sourced by Io’s extreme volcanism (Peale et al., 1979). These energetic ions both electronically excite and impart significant momentum to the water molecules in Europa’s surface triggering dissociation and chemistry (Johnson et al., 2013). The energy released by these processes leads to the ejection, or sputtering, of water-products: primarily HO, H, O, and trace amounts of OH, H, O, and HO, which then populate the exosphere. We focus here on a single component: the molecular oxygen, to isolate our discussion to the orbital evolution of the neutral exosphere. In particular, we focus on the observed near-surface component (Hall et al. 1998) derived to have a variable column density between O   cm at an altitude below 400 km. The origin and fate of this O2 (Johnson et al. 2018), as well as a comprehensive overview of Europa’s global exosphere (Plainaki et al. 2018) has been recently reviewed.

Figura 1: Top-down geometry of the simulation domain and associated physical processes at a sub-observer longitude of deg, corresponding to the sunlit leading hemisphere. The O ejection flux due to sputtering, highlighted by the red vector , dominates ejection at the sub-plasma point against the ram direction of the plasma, following Equation 1. The ejected O is assumed to be thermalized to the local surface temperature and is subject to dissociation and ionization. Adsorption, or sticking, is indicated in violet.

In order to focus primarily on the influence of rotation on Europa’s near-surface atmosphere, we employ a ballistic, 3-D Monte Carlo Exosphere Global Model (EGM: §2 ). The EGM simulates Europa’s neutral exosphere as it orbits Jupiter in Europa’s rotating frame (EphiO). The neutral O exosphere is governed by various physical processes illustrated in Fig. 1. The simulations demonstrate how Europa’s near-surface O exosphere behaves in the absence of detailed knowledge of influences such as the putative HO plumes (Roth et al. (2014b); Roth et al. (2014a); Sparks et al. (2016); Teolis and Waite (2016) ; Sparks et al. (2017)) and the near-surface plasma conditions. Although such uncertain effects could be adjusted to fit the observations, here we focus on the neutral dynamics, the effect of the source distribution, and the rotation. Our work includes the influence of Europa’s rotation on the near-surface O exosphere for the first time. The simulations are carried out in Europa’s rotating reference frame, where the Coriolis and centripetal forces act on the exospheric molecules under a gravitational potential dictated by both Europa and Jupiter’s gravitational fields.

2 Europa Exosphere Global Model

Europa’s water product exosphere is simulated here by tracking representative particles (HO, H, O, OH, H, and O) in a rotating, non-inertial reference frame, in spherical coordinates , centered on Europa and extending up to . We employ a parallelized Monte Carlo routine which is the core of the Exosphere Global Model (EGM). The numerical routines are similar to those described in detail in Turc et al. (2014), as well as in a companion paper on Ganymede’s orbital evolution (Leblanc et al., 2017), where the collisionless Boltzmann equation is solved for an imposed function distribution at the surface , for a species , a velocity , and the acceleration due to gravity , where r is the radial distance from Europa’s center . For a more detailed description of the EGM we refer the readers to Turc et al. (2014) and Section 2.2 of Leclercq (2015). In the EGM we eject test particles from Europa’s surface following known energy and angle distributions, as described later (§2.1). These test particles are on ballistic (collisionless) trajectories, and can escape, stick, and be re-emitted from the surface depending on their properties as indicated in Fig. 1. The fate of ions is not considered in this work. In the case of O, we track its interaction with the surface and account for electron and photon induced dissociation as well as ionization. Following dissociation, we also track the O formed. This procedure is carried out until a steady state O exosphere is achieved. Steady state in our simulations is achieved when the computed thermodynamic quantities change by less than 10% from one orbit to the next for any orbital position. These thermodynamic quantities are calculated based on the statistical weights of each test particle: . Here is the flux of ejected test particles, the area of the grid-cell, is the number of time steps between each test particle ejection, and is the time step between two successive positions of the test-particles, chosen so that these two positions are within one cell or two nearby cells. After a sufficient number of time-steps to achieve approximate steady state, the average number density , temperature , and velocity are calculated in each cell on a spherical grid, and at seven orbital positions, the sub-observer longitude of the satellite defined as when Europa is in eclipse behind Jupiter. In our analysis, we display the calculated densities at seven sub-observer longitudes: , , (sunlit leading), , , (sunlit trailing), and . To improve statistics, these quantities are reconstructed on intervals spanning six degrees in orbital longitude, around each orbital position, corresponding to a total of 1.4 hours.

2.1 Sputtering.

The O exosphere is thought to be produced primarily by ions originating in the Io-plasma torus. The torus, composed predominantly of sulfur and oxygen ions trapped in Jupiter’s magnetic field rotates, on average, seven times as fast as Europa, with the plasma primarily impacting Europa’s trailing hemisphere in the corotation direction (). Evidence for preferential bombardment of the trailing hemisphere by heavy ions has been observed by the Galileo Near-Infrared Mapping Spectrometer (NIMS) (Carlson et al., 1999). Following (Leblanc et al., 2002), we model the effective ion flux , onto Europa’s trailing hemisphere as diminishing with the corotation longitude :

(1)

Where the planetary longitude is 90 out of phase with the corotation: . Molecules are ejected from the trailing surface by the energy and momentum deposited by the incident ion flux in the water ice regolith. Here we use sputtering rates summarized in Cassidy et al. (2013) to discuss the fate of the oxygen component. We note that plasma diversion can affect the incoming flux of all ions and electrons, and thus the exact sputtering rates are still uncertain at present (Cassidy, 2016). In Cassidy et al. (2013) both the hot and thermal components of the incident plasma are accounted for as well as the regolith porosity, assuming the surface is primarily pure water ice. Cassidy et al. (2013) construct a temperature-dependent yield as the number of molecules ejected per incident ion based on Fama et al. (2008) and Johnson et al. (2008):

(2)

In Equation 2, is the local surface temperature, Y the yield at low temperature, and q scales the temperature dependence of the yield for a species . For the total yield given, in terms of equivalent water molecules, q= 200. The globally averaged O ejection rate also depends on the surface temperature of the ice, with , empirically derived from extensive laboratory data described in (Teolis et al. (2010); Teolis et al. (2017)), is O s, with an H yield that is twice this rate. (Cassidy et al., 2013) then estimates a globally averaged sputtering rate: HO s, of which the primary ejecta are HO, H, O with trace amounts of O, H, OH.

The energy distribution, , of the ejected particles is modeled according to the species in question. The HO, OH, O, and H are assumed to follow a water-ice sputtering distribution, based on the heavy ion () sputtering of HO ice (Johnson, 1990):

(3)

where x = 0.7, and the effective surface binding energy U = 0.052 eV. The highly volatile molecular species O and H are produced by radiolysis. As the energetic ions penetrate the ice, they break bonds leading to chemistry and the temperature dependent formation of H and O. The initial ejection energy of O (and H) can also be described by Equation 3 using and U = 0.015 eV (Johnson, 1990). We refer to this distribution as in this work. After the first ejection, those molecules returning to the surface can either react or are subsequently thermally desorbed so that their ejection speeds are roughly determined by a Maxwell Boltzmann (MB) distribution at the surface. The Maxwell Boltzmann speed distribution, accurate for a localized point is given by:

(4)

This distribution applies since the desorbed molecules undergo multiple interactions with the surface, which we model to be fast relative to the production of a new O, and are thus in local thermodynamic equilibrium with the surface ice, with an average energy, .

2.2 Surface Temperature

The effective temperature of the surface ice, is defined by radiative equilibrium. In our model, we consider a slab of ice with conductivity = 2.52 10 erg cm K s (Abramov and Spencer, 2008), and calculate the equilibrium temperature across the surface, over latitude and planetary longitude :

(5)

is the total solar irradiance in erg s cm at 1 AU, the orbital distance in AU, and the Stefan-Boltzmann constant. The albedo, , is assumed to vary linearly with between 0.65 on the leading hemisphere and 0.45 on the trailing, approximating the most recent interpretation of Galileo PPR results (Rathbun et al., 2014). The emissivity of the ice, , is set at 0.96. A small correction for the latent heat of sublimation is included (Abramov and Spencer, 2008). Endogenic heatflow, as estimated by Spencer et al. (1999) with an upper limit of 5 10 erg s cm, is neglected in this work.

The temperature gradient at the surface, in the last term of Eqn. 5 is given by the one dimensional equation for heat conduction along a direction that is normal to the surface:

(6)

where is the mass density, taken as 0.92 g cm, and the specific heat at constant pressure, , is 1.96 10 erg g K (Vance and Goodman, 2009). Eq. 6 is solved with the aid of a thermal routine using the slab boundary method (Spencer et al. (1989); Young (2017)). Forcing of solar insolation over a synodic day in late Feb. 2015 is determined in 2 increments of and using the 309.bsp kernel provided by JPL. While this approach simplifies the variation of thermophysical parameters over the surface (Rathbun et al., 2010), it provides a basis for longitudinal asymmetries about the subsolar axis that result from the thermal inertia of the ice.

In Figure 2(a) we show the evolution of Europa’s equatorial surface ice temperature along its orbit in panel A, where we include the effect of Jupiter’s shadow. The Figure 2(a) inlet is a zoom of the eclipse region where Europa’s temperature drops by K before rising to at the brighter, leading hemisphere. Figure 2(b) is the corresponding temperature map of the darker, trailing hemisphere mapped in west longitude. As is true for most planetary bodies, on average, the largest thermal gradient is between noon and midnight, or day and night, as indicated in Fig. 2(b) with a slight increase in surface temperature, K , at dusk over dawn due to the thermal inertia of the ice. The surface temperature inhomogeneities due to thermal inertia, as we will explore below, contribute only to the observed dusk-over-dawn asymmetry.

(a) a
(b) b
Figura 2: a: Equatorial subsolar surface temperature (K) as a function of sub-observer longitude, or Europa’s orbital longitude, demonstrating the leading-trailing temperature profile asymmetry due to albedo. The darker, trailing hemisphere (A = 0.45) is seen to be K warmer than the brighter, leading hemisphere (A = 0.65). b: Surface temperature map of Europa’s sunlit trailing hemisphere at . The surface temperature rapidly increases from dawn, until the subsolar point at . The surface then begins to cool slightly more slowly, passing through dusk at 18h Europa local time. The largest thermal gradients driving migration are therefore between 6h - 12h, and 12h-18h. Due to the thermal inertia of the ice, the dusk hemisphere is warmer than the dawn hemisphere.

3 Results

Here we present two results of our 3-D simulations of Europa’s water product exosphere. One in which Europa was held static, and one where Europa was able to rotate about Jupiter’s axis. In this work we focus on understanding the behavior of the primary atmospheric constituent: the near surface O.

3.1 Atmospheric Bulges

Figura 3: Europa’s static, O exosphere for a simulation fixed at the sunlit trailing hemisphere with no rotation. Vectors are shown to demonstrate the day-night migration in the absence of rotation. An artificial day-side O asymmetry is highlighted, due to preferential ion bombardment at the sunlit trailing hemisphere in the absence of rotation. The low albedo, sunlit trailing hemisphere has an equilibrium temperature of , further exaggerating the effect of O production on the exosphere due to the strong thermal dependence of the O sputtering yield.

3.1.1 No Rotation: Atmospheric O  Bulge at Noon

A trailing hemisphere O source at Europa may be expected to have a large ’atmospheric bulge’ at plasma ram, where the near-surface column density is larger than the background exosphere. As O is thought to weakly interact with the water ice, the O migrates stochastically due to the day-night temperature gradient. The gray vectors in our simulation depicted in Figure 3, demonstrate this migration. If migration were rapid, a night-side bulge would be expected like the Helium bulges on the Moon and Mercury (Leblanc and Chaufray, 2011). However, migration is slow compared to the O lifetime (c.f. Table 1) such that no asymmetry other than sputtering remains.

In a collisionless exosphere, ejected molecules return to the surface where they stick or are re-emitted with an energy determined by the local, surface ice temperature. On interacting with the surface, the fate of a molecule is sensitive to the surface residence time: where is the effective heat of adsorption. Initially for molecular oxygen, we assume the heat of adsorption is negligible and the surface residence time is extremely short, on the order of the vibration time for a Van der Waals potential () so that O has a negligible sticking coefficient. The effect of O sticking due to a longer residence time will be examined in section 4. Thus one can treat O motion as a series of random walks. Hunten et al. (1988) for example, defines , the ratio of gravitational binding energy, to the thermal energy at the surface, approximately equal to the ratio of planetary radius to the scale height, , so that . , is roughly equivalent to the number of ballistic hops a molecule experiences in traveling an average distance yielding a diffusion timescale: , where is the effective diffusion coefficient. As approximated in Hunten et al. (1988), , so that the migration time for traveling a distance r at the surface is:

(7)
Timescale ()
Orbital Period 1.0
Ballistic O hop
Day-night migration 4.2
O lifetime 1.08
Tabla 1: Exospheric timescales for Europa. Given sec. The average hop time for an O molecule is computed using an average surface temperature of , as . The ballistic hop ,, assumes no interaction with the surface. The day-night migration time is estimated as a random-walk process across one Europa radius assuming no O trapping or reactions with the surface ice. The O lifetime,, is limited by dissociation and ionization, computed by the product of the net reaction rate coefficient and the electron number density: , where cm s is the orbit-integrated reaction rate coefficient including e- impact as well as ionization, and cm the electron number density.

In Table 1, we estimate several timescales scaled to Europa’s orbital period , critical to the evolution of the near-surface O in the static and rotating cases (3.1.2). Under a uniform gravitational potential, the ballistic hop time for an O molecule is on the order of a few minutes, far shorter than the O lifetime. Table 1 also indicates that the O lifetime, , limited by the O production and loss rates is comparable to the orbital period, yielding an average exospheric column of O cm. Thus the exosphere, based on our assumptions, should be actively built and destroyed within one orbital timescale. Lastly, it is seen that the timescale for the day-to-night flow , is not only about four times longer than the O lifetime in the static case, but also four times as long as the orbital period, in the rotating case. In this way, the net spreading of the O source is on average slower than Europa’s orbital speed so we can, to first order, consider the O as coupled to the rotating surface. We will demonstrate that this coupling is the principal mechanism for the time-dependent atmospheric O bulge.

3.1.2 Rotating Case: Atmospheric O  Bulge at Dusk.

Figura 4: Formation map of Europa’s O exosphere, illustrating surface-exosphere coupling of the near-surface O. degrees west longitude corresponds to the subjovian point (SJ), where positive longitudes correspond to the trailing hemisphere (T) and negative longitudes, leading (L). At we provide a high-resolution map of Europa’s surface, for comparison to the near-surface exosphere, taken by Voyager and Galileo as adopted in (Doggett et al., 2009). On each simulation map, the subsolar point is indicated by a white X, whereas the dawn and dusk terminators are indicated by the vertical dashed lines. Inner circle surface maps: Exospheric production as shown by the ejection rate of the O sputtered from the surface with respect to latitude and west longitude at different orbital positions of Europa. The sputtering flux ranges from 2 (blue) - 12 (red) O cm s (right panel colorbar). Sub-observer longitudes(), or phase angles are provided inside the orbit. Outer circle exosphere maps: Log of the radially integrated column density of the O exosphere with respect to latitude and west longitude at different orbital positions of Europa. The radial column density ranges from   O cm (left panel colorbar). Whereas the average radially integrated column density of the O exosphere ranges from   O cm. The reader is encouraged to refer to the web-version of this article for a more enhanced view of the smaller plots.

Once rotation is activated and the O column density is tracked throughout the orbit, we simulate an atmospheric bulge persisting throughout Europa’s orbit. What is fascinating about this bulge, representing more than 10% of the bulk O column density, is it exhibits a diurnal dependence peaking consistently at dusk. A diurnal dependence is of course linked to the solar cycle, which in our model only influences the surface ice temperature (Sec. 2.1). As described in Sec. 2.2, the O production is strongly dependent on the ice temperature. In this way, one can gain an intuitive understanding of an atmospheric bulge at dusk, by reexamining the timescales in Table 1 alone. Given that the O is effectively coupled to the rotating surface and the near-surface O can be expected to follow the diurnal temperature variations in Fig. 2(b), and one can simply consider the atmospheric lifetime of an O bulge over one Europa day. For instance, the nominal atmospheric bulge at the sunlit trailing hemisphere noon (Fig. 3), will be able to survive for at least in the direction of rotation towards the dusk hemisphere. This 90 shift from noon was derived analytically for a tidally-locked satellite atmosphere in Oza et al. (2018). It was shown that the balance between the satellite’s atmospheric lifetime and orbital period over the orbit can be expressed as a parameter . The rotational shift of an atmospheric column peaking at noon for a thermally-desorbed collisionless exosphere then, can be estimated as: tan. Based on the modeled production and loss timescales, for Europa and for Ganymede (Leblanc et al. 2017), the O lifetime approaches one orbital period, so that , yielding . The 90 longitudinal shift is responsible for a maximum column at dusk and minimum column at dawn, yielding the dusk-over-dawn atmospheric asymmetry. A dusk-over-dawn asymmetry is not expected if due to far slower O loss for example, at Rhea and Dione Teolis and Waite (2016). On the other hand, if , the O bulge would be lost too quickly, before it has time to accumulate towards dusk.

We illustrate this apparent surface-exosphere synergy occurring throughout the orbit by using two sets of simulated 2-D spatial morphology maps in Figure 4. The center image indicates the orbital longitudes of the seven simulated orbital positions. The inner and outer maps represent the orbital synergy between the O surface and O exosphere varying with rotation. The inner circle, maps the O source flux at the surface: radiolytically sputtered ejecta in cm s , and the outer circle maps the O exosphere: vertical column density in cm . On the outer maps, consistent with the inner maps, we have indicated dusk, dawn, and the subsolar position (X), along with the plasma leading (L) and trailing (T) hemispheres. The outer maps indicate O bulges consistently accumulating near the dusk terminator. The inner maps illustrate that the O source is also concentrated as bulges, suggesting that the atmospheric morphology is shaped by the surface ejecta.

3.2 Orbital Evolution of Europa’s O Europa’s Diurnal Cycle

Figura 5: a: Illustration of Europa’s diurnal O cycle for our modeled thermally-dependent trailing hemisphere source. Day-night phases are superimposed in the frame of an observer at the trailing hemisphere. During ’Trailing Day’, the exosphere grows on average ( N production phase). During ’Trailing Night’, the O exosphere slowly diminishes on average ( N loss phase). The production and loss phases corresponding to the diurnal asymmetry therefore results in the orbital evolution of Europa’s O throughout its 85-hour day. b: Monte Carlo simulation results for Europa’s trailing hemisphere while rotation about Jupiter is active. Orbital evolution of the surface-averaged O column density (black) along with the surface-averaged O production flux ( f: blue) as well as the average O loss flux ( f: magenta) versus the sub-observer longitude. estimates the loss of the average O content, throughout the exospheric lifetime of the O, at each sub-observer longitude: N f. X indicates the day-night terminator separating the average production and loss phases.

Ultimately, since our modeled trailing hemisphere O source exhibits a thermal dependence, the surface-exosphere synergies can be examined qualitatively based on Europa’s day & night phases illustrated in Figure 5 (red vector: trailing hemisphere). Our Monte Carlo simulations in Figure 5 show that the day & night phases for the trailing hemisphere correspond to two different orbital phases: a globally increasing O exosphere during the ”Trailing Day”(between 180 and 360 longitude, shaded in red) and a globally diminishing O exosphere during the ”Trailing Night”(shaded in gray). Altogether, Figures 5 & 5 illustrate Europa’s diurnal O cycle, for a thermally-dependent, trailing hemisphere O source. Unlike the static, non-rotating case simulated in Section 3.1.1 and previous works, this diurnal cycle is found to peak at dusk due to atmospheric O accumulation over time.

Second order effects, such as O migration or thermal effects during eclipse, treated in our simulations, may contribute to angular deviations from a 90 longitudinal offset derived analytically for a highly thermal-dependent O source. In our simulations, the O bulges are vastly more apparent post-eclipse, near the sunlit leading hemisphere phases (i.e. 90 & 150) during which the bulk O is being lost as described in figures 5 & 5.

Overall, the 2-D maps of Fig. 4 reveal that the local O column densities vary substantially, by more than an order of magnitude 90%, due to the thermally-dependent O source fluxes also responsible for the shape of the atmospheric dusk bulges. These simulated maps provide a first attempt at a detailed picture of how Europa’s rotating O exosphere could be evolving throughout its orbit in latitude and longitude, due to the diurnal O production likely responsible for the observed dusk-over-dawn oxygen asymmetries.

4 Discussion

Observations of Europa’s O exosphere over the decades have been limited, in that the only evidence of gas phase O has been based on ratios of the far-ultraviolet atomic oxygen emission lines first observed by the Hubble Space Telescope (HST) as aurorae (Hall et al., 1995; Hall et al., 1998). The ratio of the auroral line intensities is suggestive of e impact dissociation of O: . The line-of-sight O column density then can be derived based on the electron density [cm]and electron-impact dissociation reaction rates [cm/s] (e.g. Table 2 Turc et al. (2014)) from: . Remarkably, the first set of HST observations confirmed the predicted O column density within a factor of two, based on sputtering experiments and atmospheric escape estimates (Brown et al., 1982; Johnson et al., 1982). Based on a re-analysis of Voyager, Galileo, and Cassini data Bagenal et al. (2015) demonstrated that one can expect the O to vary by at least a factor of two. Nevertheless, our knowledge on the near-surface plasma conditions influencing the derived column densities is still severely limited—a point recently reviewed in detail by (Plainaki et al., 2018) along with previous models of Europa’s asymmetric O exosphere (Cassidy et al. 2007; Plainaki et al. 2013). Indeed, Plainaki et al. 2013 did reproduce asymmetries in the exosphere by modeling the uncertain temperature-dependence of O, however without rotation the model did not reproduce consistent dusk-over-dawn asymmetries.

Despite the uncertainties Roth et al. (2016) (hereafter R16), equipped with the upgraded Space Telescope Imaging Spectrograph (STIS) with a spatial resolution of 71-95 km, altered the nominal paradigm of a globally-uniform exosphere. The STIS observations were able to distinguish a two-component atmosphere in altitude and oxygen mixing ratios: (1) near-surface O exosphere (km; 95%-99% O) and (2) corona ( km; 85% O). Previous observations did however provide evidence of asymmetric oxygen emission at various longitudes, (McGrath et al. (2004, 2009)) and (Saur et al., 2011) influenced by various external mechanisms such as the plasma. R16 assessed the influence of Jupiter’s plasma torus on the oxygen aurorae, and found that the polar aurorae are periodically more intense depending on the magnetic tilt. Monitoring Europa’s oxygen aurorae at such high spatial resolutions throughout the orbit additionally discerned a longitudinal, diurnal asymmetry of unknown origin. The dusk hemisphere was observed to be consistently brighter on average than the dawn hemisphere by a factor of two.

Figura 6: Orbital evolution of the dusk-over-dawn ratio, . Black circles with error bars: HST near-surface 1356 Å auroral emission data. The average oxygen emission from the dusk hemisphere is divided by the average dawn hemisphere emission. EGM simulations of for an exosphere subject to isotropic plasma conditions, = 70 cm and = 20 eV, are shown for non-adsorbing O (black squares), non-adsorbing with a small thermal dependence (open squares), highly adsorbing O ”sticking”to the surface (Xs), and a non-uniform adsorption (triangles).

In order to better understand the bulk physical processes regulating the dusk-over-dawn asymmetry we compare hemispherically averaged, dusk-over-dawn ratios for HST observations (black circles) with several EGM simulations assuming isotropic plasma conditions for a range of O adsorption conditions (squares, triangles, Xs) in Figure 6. The HST data is confined to the near-surface O region, and independent of photo-excitation as we uniquely evaluate the forbidden line ratios at 1356 Å. All simulated data in the shaded region, correspond to a dusk-over-dawn asymmetry, whereas points in the unshaded region represent a dawn enhancement. In our simulated O case with the least assumptions (no surface interactions; black squares) the dusk hemisphere is brighter than the dawn hemisphere throughout the entire orbit, in the same manner as the HST oxygen observations. Upon varying the initial conditions for our EGM simulations, our results appear to imply that the dusk-over-dawn asymmetry is thermal. That is, the surface temperature regulates the O production and the adsorption probability which when varied, strongly influence the diurnal asymmetry. The open squares are identical to the former, nominal case, except that a low temperature-dependence is applied to the O yield (q’ = in Eqn. 2). The reduced thermal dependence is seen to considerably reduce the dusk-over-dawn asymmetry. Conversely, increasing the O thermal dependence was found to increase the dusk-over-dawn asymmetry in the 1-D atmospheric evolution model by Oza et al. (2018). Including the thermal inertia of Europa’s ice on the other hand only increased the asymmetry by . Furthermore, simply turning on the non-inertial forces in our model (i.e. Coriolis, centripetal) without accounting for Europa’s rotation about Jupiter resulted in no dusk-over-dawn asymmetry. The result is shown in Figure 7(b).

(a)
(b)
Figura 7: Simulated line of sight column densities of Europa’s O without (a) and with (b) rotation at the sunlit leading hemisphere . Both simulations include the non-inertial Coriolis and centripetal forces. The hemispherically averaged dusk-over-dawn column density ratio is indicated in the caption as .

Next, we evaluate how the dusk-over-dawn asymmetry varies with a variable heat of adsorption for O.

4.1 O  Adsorption on Grains in the Water Ice Regolith

As discussed in Section 3.1.1, O on Europa’s surface is thought to undergo a series of unimpeded ballistic hops due to its low sublimation temperature of . Recently, on the far colder Rhea and Dione, evidence of atmospheric evolution due to O adsorption was discussed by Teolis and Waite (2016). In this regard, we simulate two additional cases where we alter the surface heat of adsorption for the returning O: (Xs) and (triangles).

The large heat of adsorption case (X) results in stronger interactions with the regolith (i.e. HO molecules freezing), and in effect longer residence and migration times, which is seen to significantly reduce the dusk-over-dawn asymmetry. The atmospheric bulges described in section 3 then, will not have time to rotate to dusk as the longer timescales will permit destruction near the production region. Lastly, we consider a case where the heat of adsorption is non-uniform between the leading and trailing hemispheres (triangles). To test this effect globally we model the heat of adsorption with a simple function with latitude () and planetary longitude ( ):

(8)

Equation 8 is a simple surface adsorption relation modeled in the EphiO frame where treats as preferential adsorption on the leading hemisphere and as preferential adsorption on the trailing hemisphere. In Fig. 6 the triangles present a simulation where , , and . It can be seen that the non-uniform heats of adsorption also reduce the dusk-over-dawn asymmetry, and additionally mitigate the orbital asymmetries near the leading hemisphere phases. This could imply that the simulated region between possesses stronger O interactions with the regolith. Conversely, the trailing hemisphere could also adsorb more efficiently since the column of O driven towards leading hemisphere phases is due to the ejection roughly half of a rotation earlier. Thus we are currently unable to deduce whether the O adsorbs more efficiently at the leading or trailing surfaces.

The transient O adsorption simulated here may describe the trapped O observed at 5771 Åas a condensed frost (Spencer and Calvin (2002) ; Spencer & Grundy 2017). That is, the O could be trapped in bubbles before it is radiolytically ejected, should the temperature be sufficiently high to allow for vacancies produced by diffusion of the incoming ion (Johnson and Jesser, 1997). Recently, Johnson et al. 2018 (submitted) reviewed the origin and fate of the trapped O in ice. It was suggested that the observed O was trapped at dangling hydrogen bonds, from where it is thermally released.

5 Conclusion/Summary

Allowing Europa to rotate about Jupiter in our 3-D Monte Carlo simulations of Europa’s water product exosphere, results in a dusk-over-dawn O asymmetry at all orbital phases. The hemispherically-averaged asymmetry is roughly consistent with the near-surface dusk-over-dawn oxygen emission recently observed by HST. We suggest that the near-surface, dusk-over-dawn O asymmetry is more generally part of Europa’s diurnal cycle of O, built and destroyed over one orbit. This O cycle appears to depend strongly on the (1) O production (therefore also the surface temperature), (2) O loss, and (3) Europa’s rotation rate, all of which lead to a peak in column density at dusk.

The diurnal cycle described in this work, can be summarized as follows :

  1. Trailing Day: The trailing hemisphere source is illuminated and rapidly builds an O column rotating with Europa’s surface, ejecting between O cm s throughout the day.

    • After half of an orbit, the rotating O column has accumulated an atmospheric bulge of average density O cm spread across a quarter of a hemisphere from dusk, local time.

  2. Trailing Night: The trailing hemisphere source enters night, at which point the e-impact dissociation begins to slowly overwhelm the O production and acts to isotropically diminish the bulk O exosphere.

A similar dusk-over-dawn asymmetry is predicted at all sub-observer longitudes in Ganymede’s near-surface O atmosphere Leblanc et al. (2017). We note that these O bulges maintained near dusk, hinge on the timescales of tidally-locked atmospheres as described by Oza et al. (2018). The orbital timescale is comparable to the average atmospheric lifetime of the radiolytic O, so that . In other words, the O exosphere survives long enough so that the O can accumulate towards dusk during the day, given its simulated thermal dependence. If the O exosphere survived for too long, for instance , the continued production would average out any asymmetries as the O loss would be negligible over the orbit. If the O exosphere was destroyed too quickly on the other hand, , the diurnal asymmetries would not have time to build an O bulge. Overall, the comparable size of the satellite’s rotation rate and the rate at which the near-surface O is being lost results in a shift towards dusk. The relative sizes of the rotation rate and the estimated loss rate at Europa and Ganymede, result in a periodic O asymmetry, which can be probed in satellite local time.

While we have provided evidence for a dusk-over-dawn asymmetry at Europa, a global understanding of Europa’s exosphere is far from fulfilled. The O cycle depends on the O production and loss: source rates may require a largely thermal source (e.g. Oza et al. (2018), Johnson et al. 2018), and loss rates may need to include plasma interactions (e.g. Dols et al. (2016); Lucchetti et al. (2016)). A self-consistent exosphere model including the plasma interaction is therefore sorely needed. A stronger understanding of the spatial morphologies of O exospheres will be of particular interest for the spacecraft trajectories at the icy Galilean satellites (i.e. NASA’s Europa Clipper; ESA’s JUICE).

Acknowledgments: The authors express their gratitude to Lorenz Roth for insight on better comparing the HST observations and the EGM exosphere simulations, and to Ben Teolis for constructive discussions on O-water ice interactions. AVO and FL acknowledge the support of LabEx/ ESEP . This work was also supported by CNES ”Systéme Solaire”program. REJ acknowledges support from NASA’s Planetary Data Systems Program. This work is also part of HELIOSARES Project supported by the ANR (ANR-09-BLAN-0223) and ANR MARMITE-CNRS (ANR-13-BS05-0012-02). Authors also acknowledge the support of the computational platform CICLAD hosted by the Institut Pierre Simon Laplace.

\biboptions

authoryear

Referencias

  • Abramov and Spencer (2008) Abramov, O., Spencer, J.R., 2008. Numerical modeling of endogenic thermal anomalies on Europa. Icarus 195, 378–385. doi:10.1016/j.icarus.2007.11.027.
  • Bagenal et al. (2015) Bagenal, F., Sidrow, E., Wilson, R.J., Cassidy, T.A., Dols, V., Crary, F.J., Steffl, A.J., Delamere, P.A., Kurth, W.S., Paterson, W.R., 2015. Plasma conditions at Europa’s orbit. Icarus 261, 1–13. doi:10.1016/j.icarus.2015.07.036.
  • Brown et al. (1982) Brown, W.L., Lanzerotti, L.J., Johnson, R.E., 1982. Fast Ion Bombardment of Ices and its Astrophysical Implications. Science 218, 525–531. doi:10.1126/science.218.4572.525.
  • Carlson et al. (1999) Carlson, R.W., Johnson, R.E., Anderson, M.S., 1999. Sulfuric acid on Europa and the radiolytic sulfur cycle. Science 286, 97–99. doi:10.1126/science.286.5437.97.
  • Cassidy (2016) Cassidy, T.A., 2016. Widespread confusion in the modeling of Europa’s magnetospheric interaction: what the potential modeler should consider before getting started. AGU Fall Meeting Abstracts , SM34A–05.
  • Cassidy et al. (2007) Cassidy, T.A., Johnson, R.E., McGrath, M.A., Wong, M.C., Cooper, J.F., 2007. The spatial morphology of Europa’s near-surface O atmosphere. Icarus 191, 755–764. doi:10.1016/j.icarus.2007.04.033.
  • Cassidy et al. (2013) Cassidy, T.A., Paranicas, C.P., Shirley, J.H., Dalton, III, J.B., Teolis, B.D., Johnson, R.E., Kamp, L., Hendrix, A.R., 2013. Magnetospheric ion sputtering and water ice grain size at Europa. Planetary and Space Science 77, 64–73. doi:10.1016/j.pss.2012.07.008.
  • Doggett et al. (2009) Doggett, T., Greeley, R., Figueredo, P., Tanaka, K., 2009. Geologic Stratigraphy and Evolution of Europa’s Surface.
  • Dols et al. (2016) Dols, V.J., Bagenal, F., Cassidy, T.A., Crary, F.J., Delamere, P.A., 2016. Europa’s atmospheric neutral escape: Importance of symmetrical O charge exchange. Icarus 264, 387–397. doi:10.1016/j.icarus.2015.09.026.
  • Fama et al. (2008) Fama, M., Shi, J., Baragiola, R., 2008. Sputtering of ice by low-energy ions. Surface Science 602, 156 – 161. URL: http://www.sciencedirect.com/science/article/pii/S0039602807009879, doi:http://dx.doi.org/10.1016/j.susc.2007.10.002.
  • Greenberg (2010) Greenberg, R., 2010. Transport Rates of Radiolytic Substances into Europa’s Ocean: Implications for the Potential Origin and Maintenance of Life. Astrobiology 10, 275–283. doi:10.1089/ast.2009.0386.
  • Hall et al. (1998) Hall, D.T., Feldman, P.D., McGrath, M.A., Strobel, D.F., 1998. The Far-Ultraviolet Oxygen Airglow of Europa and Ganymede. Astrophysical Journal 499, 475–481. doi:10.1086/305604.
  • Hall et al. (1995) Hall, D.T., Strobel, D.F., Feldman, P.D., McGrath, M.A., Weaver, H.A., 1995. Detection of an oxygen atmosphere on Jupiter’s moon Europa. Nature 373, 677–679. doi:10.1038/373677a0.
  • Hand et al. (2007) Hand, K.P., Carlson, R.W., Chyba, C.F., 2007. Energy, Chemical Disequilibrium, and Geological Constraints on Europa. Astrobiology 7, 1006–1022. doi:10.1089/ast.2007.0156.
  • Hunten et al. (1988) Hunten, D.M., Shemansky, D.E., Morgan, T.H., 1988. The Mercury atmosphere, in: Vilas, F., Chapman, C.R., Matthews, M.S. (Eds.), Mercury, University of Arizona Press, pp. 562–612.
  • Johnson (1990) Johnson, R.E., 1990. Energetic Charged-Particle Interactions with Atmospheres and Surfaces.
  • Johnson et al. (2013) Johnson, R.E., Carlson, R.W., Cassidy, T.A., Fama, M., 2013. Sputtering of Ices, in: Gudipati, M.S., Castillo-Rogez, J. (Eds.), Astrophysics and Space Science Library, p. 551. doi:10.1007/978-1-4614-3076-6_17.
  • Johnson et al. (2008) Johnson, R.E., Famá, M., Liu, M., Baragiola, R.A., Sittler, E.C., Smith, H.T., 2008. Sputtering of ice grains and icy satellites in Saturn’s inner magnetosphere. Planetary and Space Science 56, 1238–1243. doi:10.1016/j.pss.2008.04.003.
  • Johnson and Jesser (1997) Johnson, R.E., Jesser, W.A., 1997. O/O Microatmospheres in the Surface of Ganymede. Astrophysical Journal Letters 480, L79–L82. doi:10.1086/310614.
  • Johnson et al. (1982) Johnson, R.E., Lanzerotti, L.J., Brown, W.L., 1982. Planetary applications of ion induced erosion of condensed-gas frosts. Nuclear Instruments and Methods 198, 147–157. doi:10.1016/0167-5087(82)90066-7.
  • Johnson et al. (2003) Johnson, R.E., Quickenden, T.I., Cooper, P.D., McKinley, A.J., Freeman, C.G., 2003. The Production of Oxidants in Europa’s Surface. Astrobiology 3, 823–850. doi:10.1089/153110703322736123.
  • Leblanc and Chaufray (2011) Leblanc, F., Chaufray, J.Y., 2011. Mercury and Moon He exospheres: Analysis and modeling. Icarus 216, 551–559. doi:10.1016/j.icarus.2011.09.028.
  • Leblanc et al. (2002) Leblanc, F., Johnson, R.E., Brown, M.E., 2002. Europa’s Sodium Atmosphere: An Ocean Source? Icarus 159, 132–144. doi:10.1006/icar.2002.6934.
  • Leblanc et al. (2017) Leblanc, F., Oza, A.V., Leclercq, L., Schmidt, C., Cassidy, T., Modolo, R., Chaufray, J.Y., Johnson, R.E., 2017. On the orbital variability of Ganymede’s atmosphere. Icarus 293, 185–198. doi:10.1016/j.icarus.2017.04.025.
  • Leclercq (2015) Leclercq, L., 2015. From exosphere to magnetosphere of planetary objects : optimization of parallelized modelisations for an application to Ganymede. Theses. Université Paris-Saclay. URL: https://hal-insu.archives-ouvertes.fr/tel-01256191.
  • Lucchetti et al. (2016) Lucchetti, A., Plainaki, C., Cremonese, G., Milillo, A., Cassidy, T., Jia, X., Shematovich, V., 2016. Loss rates of Europa’s tenuous atmosphere. Planetary and Space Science 130, 14–23. doi:10.1016/j.pss.2016.01.009.
  • McGrath et al. (2009) McGrath, M.A., Hansen, C.J., Hendrix, A.R., 2009. Observations of Europa’s Tenuous Atmosphere, in: Pappalardo, R.T., McKinnon, W.B., Khurana, K.K. (Eds.), Europa, Edited by Robert T. Pappalardo, William B. McKinnon, Krishan K. Khurana ; with the assistance of René Dotson with 85 collaborating authors. University of Arizona Press, Tucson, 2009. The University of Arizona space science series ISBN: 9780816528448, p.485, p. 485.
  • McGrath et al. (2004) McGrath, M.A., Lellouch, E., Strobel, D.F., Feldman, P.D., Johnson, R.E., 2004. Satellite atmospheres, in: Bagenal, F., Dowling, T.E., McKinnon, W.B. (Eds.), Jupiter. The Planet, Satellites and Magnetosphere, pp. 457–483.
  • Oza et al. (2018) Oza, A.V., Johnson, R.E., Leblanc, F., 2018. Dusk/dawn atmospheric asymmetries on tidally-locked satellites: O at Europa. Icarus 305, 50–55. doi:10.1016/j.icarus.2017.12.032.
  • Peale et al. (1979) Peale, S.J., Cassen, P., Reynolds, R.T., 1979. Melting of Io by tidal dissipation. Science 203, 892–894. doi:10.1126/science.203.4383.892.
  • Plainaki et al. (2018) Plainaki, C., Cassidy, T.A., Shematovich, V.I., Milillo, A., Wurz, P., Vorburger, A., Roth, L., Galli, A., Rubin, M., Blöcker, A., Brandt, P.C., Crary, F., Dandouras, I., Jia, X., Grassi, D., Hartogh, P., Lucchetti, A., McGrath, M., Mangano, V., Mura, A., Orsini, S., Paranicas, C., Radioti, A., Retherford, K.D., Saur, J., Teolis, B., 2018. Towards a Global Unified Model of Europa’s Tenuous Atmosphere. Space Science Reviews 214, 40. doi:10.1007/s11214-018-0469-6.
  • Plainaki et al. (2013) Plainaki, C., Milillo, A., Mura, A., Saur, J., Orsini, S., Massetti, S., 2013. Exospheric O densities at Europa during different orbital phases. Planetary and Space Science 88, 42–52. doi:10.1016/j.pss.2013.08.011.
  • Rathbun et al. (2010) Rathbun, J.A., Rodriguez, N.J., Spencer, J.R., 2010. Galileo PPR observations of Europa: Hotspot detection limits and surface thermal properties. Icarus 210, 763–769. doi:10.1016/j.icarus.2010.07.017.
  • Rathbun et al. (2014) Rathbun, J.A., Spencer, J.R., Howett, C.J.A., 2014. Galileo PPR Observations of Europa: Correlations of Thermophysical Properties with Exogenic and Endogenic Processes, in: Workshop on the Habitability of Icy Worlds, p. 4045.
  • Roth et al. (2014a) Roth, L., Retherford, K.D., Saur, J., Strobel, D.F., Feldman, P.D., McGrath, M.A., Nimmo, F., 2014a. Orbital apocenter is not a sufficient condition for HST/STIS detection of Europa’s water vapor aurora. Proceedings of the National Academy of Science 111, E5123–E5132. doi:10.1073/pnas.1416671111.
  • Roth et al. (2016) Roth, L., Saur, J., Retherford, K.D., Strobel, D.F., Feldman, P., McGrath, M., Spencer, J.R., Blocker, A., Ivchenko, N., 2016. Europa’s far ultraviolet oxygen aurora from a comprehensive set of HST observations. Journal of Geophysical Research 261, 1–13. doi:10.1016/j.icarus.2015.07.036.
  • Roth et al. (2014b) Roth, L., Saur, J., Retherford, K.D., Strobel, D.F., Feldman, P.D., McGrath, M.A., Nimmo, F., 2014b. Transient Water Vapor at Europa’s South Pole. Science 343, 171–174. doi:10.1126/science.1247051.
  • Saur et al. (2011) Saur, J., Feldman, P.D., Roth, L., Nimmo, F., Strobel, D.F., Retherford, K.D., McGrath, M.A., Schilling, N., Gérard, J.C., Grodent, D., 2011. Hubble Space Telescope/Advanced Camera for Surveys Observations of Europa’s Atmospheric Ultraviolet Emission at Eastern Elongation. Astrophysical Journal 738, 153. doi:10.1088/0004-637X/738/2/153.
  • Sparks et al. (2016) Sparks, W.B., Hand, K.P., McGrath, M.A., Bergeron, E., Cracraft, M., Deustua, S.E., 2016. Probing for Evidence of Plumes on Europa with HST/STIS. Astrophysical Journal 829, 121. doi:10.3847/0004-637X/829/2/121, arXiv:1609.08215.
  • Sparks et al. (2017) Sparks, W.B., Schmidt, B.E., McGrath, M.A., Hand, K.P., Spencer, J.R., Cracraft, M., E Deustua, S., 2017. Active Cryovolcanism on Europa? Astrophysical Journal Letters 839, L18. doi:10.3847/2041-8213/aa67f8, arXiv:1704.04283.
  • Spencer and Calvin (2002) Spencer, J.R., Calvin, W.M., 2002. Condensed O on Europa and Callisto. Astronomical Journal 124, 3400--3403. doi:10.1086/344307.
  • Spencer et al. (1989) Spencer, J.R., Lebofsky, L.A., Sykes, M.V., 1989. Systematic biases in radiometric diameter determinations. Icarus 78, 337--354. doi:10.1016/0019-1035(89)90182-6.
  • Spencer et al. (1999) Spencer, J.R., Tamppari, L.K., Martin, T.Z., Travis, L.D., 1999. Nighttime Thermal Anomalies on Europa, in: Lunar and Planetary Science Conference.
  • Teolis et al. (2010) Teolis, B.D., Jones, G.H., Miles, P.F., Tokar, R.L., Magee, B.A., Waite, J.H., Roussos, E., Young, D.T., Crary, F.J., Coates, A.J., Johnson, R.E., Tseng, W.L., Baragiola, R.A., 2010. Cassini Finds an Oxygen-Carbon Dioxide Atmosphere at Saturn’s Icy Moon Rhea. Science 330, 1813. doi:10.1126/science.1198366.
  • Teolis et al. (2017) Teolis, B.D., Plainaki, C., Cassidy, T.A., Raut, U., 2017. Water Ice Radiolytic O, H, and HO Yields for Any Projectile Species, Energy, or Temperature: A Model for Icy Astrophysical Bodies. Journal of Geophysical Research (Planets) 122, 1996--2012. doi:10.1002/2017JE005285.
  • Teolis and Waite (2016) Teolis, B.D., Waite, J.H., 2016. Dione and Rhea seasonal exospheres revealed by Cassini CAPS and INMS. Icarus 272, 277--289. doi:10.1016/j.icarus.2016.02.031.
  • Turc et al. (2014) Turc, L., Leclercq, L., Leblanc, F., Modolo, R., Chaufray, J.Y., 2014. Modelling Ganymede’s neutral environment: A 3D test-particle simulation. Icarus 229, 157--169. doi:10.1016/j.icarus.2013.11.005.
  • Vance and Goodman (2009) Vance, S., Goodman, J., 2009. Oceanography of an Ice-Covered Moon, in: Pappalardo, R.T., McKinnon, W.B., Khurana, K.K. (Eds.), Europa, Edited by Robert T. Pappalardo, William B. McKinnon, Krishan K. Khurana ; with the assistance of René Dotson with 85 collaborating authors. University of Arizona Press, Tucson, 2009. The University of Arizona space science series ISBN: 9780816528448, p.459, p. 459.
  • Vance et al. (2016) Vance, S.D., Hand, K.P., Pappalardo, R.T., 2016. Geophysical controls of chemical disequilibria in Europa. Geophysical Research Letters 43, 4871--4879. doi:10.1002/2016GL068547.
  • Young (2017) Young, L.A., 2017. Volatile transport on inhomogeneous surfaces: II. Numerical calculations (VT3D). Icarus 284, 443--476. doi:10.1016/j.icarus.2016.07.021.
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
""
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
   
Add comment
Cancel
Loading ...
316173
This is a comment super asjknd jkasnjk adsnkj
Upvote
Downvote
""
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters
Submit
Cancel

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test
Test description