Variations of the magnetic field near Mars caused by magnetic crustal anomalies

The possible avenues for photoelectron transport were determined during southern hemisphere winter at Mars by using a mapping analysis of the theoretical magnetic field. Magnetic field line tracing was performed by superposing two magnetic field models: (1) magnetic field derived from a three-dimensional (3D) self-consistent quasi-neutral hybrid model which does not contain the Martian crustal magnetic anomalies and (2) a 3D map of the magnetic field associated with the magnetic anomalies based on Mars Global Surveyor magnetic field measurements. It was found that magnetic field lines connected to the nightside of the planet are mainly channeled within the optical shadow of the magnetotail whereas magnetic field lines connected to the dayside of the planet are observed to form the remainder of the magnetosphere. The simulation suggests that the crustal anomalies create “a magnetic shield” by decreasing the region near Mars which is magnetically connected to the Martian magnetosphere. The rotation of Mars causes periodic changes in magnetic connectivity, but not to qualitative changes in the overall magnetic field draping around Mars.     

The composition and thickness of the crust of Mars estimated from rare earth elements and neodymium-isotopic compositions of Martian meteorites

Abstract— Isotopic and trace element compositions of Martian meteorites show that early differentiation of Mars produced complementary crustal and mantle reservoirs that were sampled by later magmatic events. This paper describes a mass balance model that estimates the rare earth element (REE) content and thickness of the crust of Mars from the compositions of shergottites. The diverse REE and Nd isotopic compositions of shergottites are most easily explained by variable addition of light rare earth element (LREE)–enriched crust to basaltic magmas derived from LREE-depleted mantle source regions. Antarctic shergottites EET 79001, ALH 77005, LEW 88516, and QUE 94201 all have strongly LREE-depleted patterns and positive initial ¹⁴³Nd isotopic compositions, which is consistent with the generation of these magmas from depleted mantle sources and little or no interaction with enriched crust. In contrast, Shergotty and Zagami have negative initial ¹⁴³Nd isotopic compositions and less pronounced depletions of the LREE, which have been explained by incorporation of enriched crustal components into mantle-derived magmas (Jones, 1989; Longhi, 1991; Borg et al., 1997). The mass balance model presented here derives the REE composition of the crustal component in Shergotty by assuming it represents a mixture between a mantle-derived magma similar in composition to EET 79001A and a LREE-enriched crustal component. The amount of crust in Shergotty is constrained by mixing relations based on Nd-isotopic compositions, which allows the REE pattern of the crustal component to be calculated by mass balance. The effectiveness of this model is demonstrated by the successful recovery of important characteristics of the Earth's continental crust from terrestrial Columbia River basalts. Self-consistent results for Nd-isotopic compositions and REE abundances are obtained if Shergotty contains ～10–30% of LREE-enriched crust with >10 ppm Nd. This crustal component would have moderately enriched LREE (Sm/Nd = 0.25–0.27; ¹⁴⁷Sm/¹⁴⁴Nd = 0.15–0.17; La/Yb = 2.7–3.8), relatively unfractionated heavy rare earth elements (HREE), and no Eu anomaly. Crust with these characteristics can be produced from a primitive lherzolitic Martian mantle by modest amounts (2–8%) of partial melting, and it would have a globally averaged thickness of <45 km, which is consistent with geophysical estimates. Mars may serve as a laboratory to investigate planetary differentiation by extraction of a primary basaltic crust.     

On the properties of O and O2 ions in a hybrid model and in Mars Express IMA/ASPERA-3 data: A case study

We have performed a numerical simulation to analyze the energy spectra of escaping planetary O⁺ and O₂⁺ ions at Mars. The simulated time-energy spectrograms were generated along orbit no. 555 (June 27, 2004) of Mars Express when its Ion Mass Analyzer (IMA)/ASPERA-3 ion instrument detected escaping planetary ions. The simulated time-energy spectrograms are in general agreement with the hypothesis that planetary O⁺ and O₂⁺ ions far from Mars are accelerated by the convective electric field. The HYB-Mars hybrid model simulation also shows that O⁺ ions originating from the ionized hot oxygen corona result in a high-energy (E>1 keV) O⁺ ion population that exists very close to Mars. In addition, the simulation also results in a low-energy (E<0.1 keV) planetary ion population near the pericenter. In the analyzed orbit, IMA did not observe a clear high-energy planetary ion or a clear low-energy planetary ion population near Mars. One possible source for this discrepancy may be the Martian magnetic crustal anomalies because MEX passed over a strong crustal field region near the pericenter, but the hybrid model does not include the magnetic crustal anomalies.     

Various aspects of the petrogenesis of the Martian shergottite meteorites

Several controversies are associated with the age and origin of the shergottite meteorites, a suite of basaltic samples from Mars. Here, it will be argued that (1) the shergottites have a young igneous age, ≤600 Myr, (2) their parent magmas were relatively dry, (3) the range of initial isotopic compositions in shergottites is most likely due to assimilation of crustal materials by mantle‐derived basaltic magmas, and (4) the intercumulus liquid compositions of shergottites such as Shergotty and Zagami are relatively well constrained.     

Day-side ionospheric conductivities at Mars

We present estimates of the day-side ionospheric conductivities at Mars based on magnetic field measurements by Mars Global Surveyor (MGS) at altitudes down to ∼100 km during aerobraking orbits early in the mission. At Mars, the so-called ionospheric dynamo region, where plasma/neutral collisions permit electric currents perpendicular to the magnetic field, lies between 100 and 250 km altitude. We find that the ionosphere is highly conductive in this region, as expected, with peak Pedersen and Hall conductivities of 0.1-1.5 S/m depending on the solar illumination and induced magnetospheric conditions. Furthermore, we find a consistent double peak pattern in the altitude profile of the day-side Pedersen conductivity, similar to that on Titan found by Rosenqvist et al. (2009). A high altitude peak, located between 180 and 200 km, is equivalent to the terrestrial peak in the lower F-layer. A second and typically much stronger layer of Pedersen conductivity is observed between 120 and 130 km, which is below the Hall conductivity peak at about 130-140 km. In this altitude region, MGS finds a sharp decrease in induced magnetic field strength at the inner magnetospheric boundary, while the day-side electron density is known to remain high as far down as 100 km. We find that such Titan-like behaviour of the Pedersen conductivity is only observed under regions of strongly draped magnetospheric field-lines, and negligible crustal magnetic anomalies below the spacecraft. Above regions of strong crustal magnetic anomalies, the Pedersen conductivity profile becomes more Earth-like with one strong Pedersen peak above the Hall conductivity peak. Here, both conductivities are 1-2 orders of magnitude smaller than the above only weakly magnetised crustal regions, depending on the strength of the crustal anomaly field at ionospheric altitudes. This nature of the Pedersen conductivity together with the structured distribution of crustal anomalies all over the planet should give rise to strong conductivity gradients around such anomalies. Day-side ionospheric conductivities on Mars (in regions away from the crustal magnetic anomalies) and Titan seem to behave in a very similar manner when horizontally draped magnetic field-lines partially magnetise a sunlit ionosphere. Therefore, it appears that a similar double peak structure of strong Pedersen conductivity could be a more general feature of non-magnetised bodies with ionised upper atmospheres, and thus should be expected to occur also at other non-magnetised terrestrial planets like Venus or other planetary bodies within the host planet magnetospheres.     

Large‐ion lithophile element fractionation during the early differentiation of Mars and the composition of the martian primitive mantle

Abstract— Basaltic shergottites display a systematic decrease in K/Th, K/U, and K/La ratios with increasing K content. These trends are interpreted as mixing lines between relatively young martian magmas derived from highly depleted mantle sources and an ancient large‐ion lithophile (LIL) element‐enriched crustal component. One implication of this is that a substantial fractionation of these ratios occurs during the early crustal differentiation on Mars. Isotopic evidence from SNC meteorites and compositional data from Pathfinder and orbital gamma ray spectroscopy suggest that in excess of 50% of the LIL element complement of Mars resides in the crustal reservoir. If so, the primitive mantle of Mars is significantly more volatile‐depleted (i.e., lower K/Th, K/U, K/La) than previously thought but probably (though not necessarily) still less volatile‐depleted than the primitive mantle of the Earth. The La/Th ratios of virtually all SNC meteorites are subchondritic, including those with the most severe LREE‐depletion. Extrapolation of the basaltic shergottite trend suggests that both the depleted mantle end member and the enriched crustal end member have subchondritic La/Th ratios. This is in contrast with the Earth where basalts from LIL element‐depleted sources such as MORB have superchondritic La/Th ratios, complementary to the subchondritic ratios of the continental crust. Accordingly, assuming that the refractory elements are in chondritic proportions for the Mars primitive mantle, an additional major geochemical reservoir must exist on Mars that may not yet have been sampled.     

The cause for the north-south orientation of the crustal dichotomy and the equatorial location of Tharsis on Mars

The crustal dichotomy and the Tharsis rise are the most prominent topographic features on Mars. The dichotomy is largely an expression of different crustal thicknesses in the northern and southern hemispheres, while Tharsis is centered near the equator at the dichotomy boundary. However, the cause for the orientation of the dichotomy and the equatorial location of Tharsis remains poorly understood. Here we show that the crustal thickness variations associated with the dichotomy may have driven true polar wander, establishing the north-south orientation of the dichotomy very early in martian history. Such a reorientation that placed the dichotomy boundary near the equator would also have constrained the Tharsis region on the dichotomy boundary to have originated near the equator. We present a scenario for the early generation and subsequent reorientation of the hemispheric dichotomy, although the reorientation is independent of the formation mechanism. Our results also have implications for the sharply different remanent magnetizations between the two hemispheres.     

Hydrogen isotopic composition of the Martian mantle inferred from the newest Martian meteorite fall,Tissint

The hydrogen isotopic composition of planetary reservoirs can provide key constraints on the origin and history of water on planets. The sources of water and the hydrological evolution of Mars may be inferred from the hydrogen isotopic compositions of mineral phases in Martian meteorites, which are currently the only samples of Mars available for Earth‐based laboratory investigations. Previous studies have shown that δD values in minerals in the Martian meteorites span a large range of ?250 to +6000‰. The highest hydrogen isotope ratios likely represent a Martian atmospheric component: either interaction with a reservoir in equilibrium with the Martian atmosphere (such as crustal water), or direct incorporation of the Martian atmosphere due to shock processes. The lowest δD values may represent those of the Martian mantle, but it has also been suggested that these values may represent terrestrial contamination in Martian meteorites. Here we report the hydrogen isotopic compositions and water contents of a variety of phases (merrillites, maskelynites, olivines, and an olivine‐hosted melt inclusion) in Tissint, the latest Martian meteorite fall that was minimally exposed to the terrestrial environment. We compared traditional sample preparation techniques with anhydrous sample preparation methods, to evaluate their effects on hydrogen isotopes, and find that for severely shocked meteorites like Tissint, the traditional sample preparation techniques increase water content and alter the D/H ratios toward more terrestrial‐like values. In the anhydrously prepared Tissint sample, we see a large range of δD values, most likely resulting from a combination of processes including magmatic degassing, secondary alteration by crustal fluids, shock‐related fractionation, and implantation of Martian atmosphere. Based on these data, our best estimate of the δD value for the Martian depleted mantle is ?116 ± 94‰, which is the lowest value measured in a phase in the anhydrously prepared section of Tissint. This value is similar to that of the terrestrial upper mantle, suggesting that water on Mars and Earth was derived from similar sources. The water contents of phases in Tissint are highly variable, and have been affected by secondary processes. Considering the H₂O abundances reported here in the driest phases (most likely representing primary igneous compositions) and appropriate partition coefficients, we estimate the H₂O content of the Tissint parent magma to be ≤0.2 wt%.     

Interpretation of Mars southern highlands high amplitude magnetic field with total gradient and fractal source modeling: New insights into the magnetic mystery of Mars

Using model studies, the total gradient (TG) of the Z-component magnetic field is shown to be a useful quantity for delineating sources of satellite-altitude magnetic anomalies; this field is used to constrain the location and lateral boundaries of sources of high amplitude magnetic anomalies of southern highlands of Mars. The TG field suggests two parallel linear and oppositely magnetized sources of 1000 and 1800 km length separated by 1000 km of region of intervening non-parallel sources. The simplest interpretation of the long, linear features is that they are zones of multitudinous crustal scale dikes formed in separate episodes of rifting, and not features associated with the mechanism of seafloor spreading. Forward modeling with uniformly magnetized sources suggests that magnetizations of the order of 10-50 A/m (40 km thickness) over ∼100 km width in the case of the southern source and of 12.5-27.5 A/m (40 km thickness) and ∼200 km width for the northern source are necessary to explain the Z-component amplitudes and features of the TG field. If the crustal magnetization on Mars were to be distributed fractally as on Earth, magnetizations matching the largest amplitude features on Mars may be spatially correlated from a 50-100 km distance range (β ∼ 3) to approaching nearly uniform magnetization (β ∼ 5) values. To keep magnetization intensity as small as possible, the higher end of β values are preferred, whereas, small amplitude anomaly features could be generated from sources with β ∼ 3. Many of the Mars anomaly features could be coalescence effects similar to the coalescence of anomaly features observed on http://icarus.cornell.edu/information/keywords.html.Earth.     

Features of the Martian Magnetic Field Structure

Based on the single-fluid MHD model of Mars space simulation, this paper has studied the magnetic field structure in the near-Mars space and investigated the influence of Martian crustal magnetic anomalies on the magnetic field structure. In the process of the solar wind interaction with Mars, the bow shock and magnetic pile-up region are produced. The interplanetary magnetic lines are curved and deformed while they are towed toward the two poles by the solar wind. The majority of magnetic lines bypass the two poles, then leave behind a ‘V-shaped’ structure in the magnetotail behind Mars. In the crust of Mars, the local magnetic anomalies have a noticeable influence on the magnetic field structure. The magnetic anomalies at different positions and in different intensities interact with the solar wind to form the mini-magnetospheres of different structures and morphologies, such as the towed mini-magnetosphere and the mini-magnetosphere with open magnetic lines. The local magnetic anomalies have changed the near-Mars magnetic field structure, and probably changed the plasma distribution as well.     

Seismic velocity models for an internally asymmetric Mars

The well-known dichotomy in topography, surface age, and crustal structure between the northern lowlands and the southern uplands of Mars has been explained by both endogenic and exogenic processes. According to the used model this asymmetry might be a result of a certain mechanism of core fommation influencing the following planetary evolution. Therefore it has been assumed that the present internal structure of Mars is characterized by different velocity-depth distributions of the mantle for the northern and southern hemisphere, respectively. For both regions significant differences in travel times of seismic waves were calculated. These results may be important for the future seismic exploration of Mars.     

Pole wandering of Mars: Evidence from paleomagnetic poles

We use the mapping-phase high-altitude magnetic measurements provided by Mars Global Surveyor (MGS) between March 1999 and April 2003 to model nine relatively isolated magnetic anomalies of Mars. Each anomaly is modeled with an elliptical prism. Each component of the observed magnetic field is modeled independently using an elliptical prism in order to assess the reliability of the results and suppress non-crustal and nearby crustal source contaminations. The paleomagnetic pole positions are obtained from the magnetization vectors of the model source bodies. We clean the data by removing the bad tracks and then divide the entire data into two sets that are measured at different times. Applying covariance analysis in the Fourier domain to two maps of the same magnetic component that are derived from the two sets provides a means to extract the most common features of the maps. The quality of a model is evaluated and only good models are used in the final geophysical interpretation. Most poles that come from good models cluster in the Tharsis region, suggesting that Mars experienced polar motion since the magnetic source bodies were magnetized.     

Asymmetry of plasma fluxes at Mars. ASPERA-3 observations and hybrid simulations

The asymmetry of fluxes of solar wind and planetary ions is studied by using the ASPERA-3 observations onboard the Mars Express spacecraft in February 2004 to March 2006. Due to the small scale of the Martian magnetosphere and its induced origin, the flow pattern near Mars is sensitive to the directions of the interplanetary magnetic and electric (-V×B) fields. Asymmetry of the magnetic field draping produces an asymmetry in plasma flows in the plane containing the IMF. The crustal magnetic fields on Mars also influence the flow pattern. Scavenging of planetary ions is less efficient in the regions of strong crustal magnetization and therefore the escape fluxes of planetary ions in the southern hemisphere are smaller. The results of the observations are compared to simulations based on a 3D hybrid model with several ion species.     

A plasma flow velocity boundary at Mars from the disappearance of electron plasma oscillations

The Mars Advanced Radar for Subsurface and Ionospheric Sounding (MARSIS) on the Mars Express (MEX) spacecraft is capable of measuring ionospheric electron density by the use of two main methods: remote radar sounding and from the excitation of local plasma oscillations. The frequency of the locally excited electron plasma oscillations is used to measure the local electron density. However, plasma oscillations are not observed when the plasma flow velocity is higher than about 160 km/s, which occurs mainly in the solar wind and magnetosheath. As a consequence, in many passes, there is a sudden disappearance of the plasma oscillations as the spacecraft enters into the magnetosheath. This fact allows us to identify a flow velocity boundary on the dayside, between the ionosphere of Mars and the shocked solar wind. This paper summarizes the results of the measurement of 552 orbits mostly over a period from August 4, 2005 to August 17, 2007. The boundary points found using MARSIS have been verified by measurements from the Analyzer of Space Plasma and Energetic Atoms (ASPERA-3) Electron Spectrometer (ELS) instrument on Mars Express. The average position of the flow velocity boundary is compared to flow velocity simulations computed using hybrid model and other boundaries. The boundary altitude is slightly lower than the magnetic pile-up boundary determined using Phobos 2 and Mars Global Surveyor (MGS) crossings, but it is in good agreement with the induced magnetospheric boundary determined by ASPERA-3. Investigation of the effect of the crustal magnetic field revealed that the flow velocity boundary is raised at the locations with strong crustal magnetic fields.     

On the effect of the martian crustal magnetic field on atmospheric erosion

Without the shielding of a strong intrinsic magnetic field, the martian atmosphere directly interacts with the impacting solar wind. The neutral constituents of the atmospheric corona can be ionized, and then picked up and accelerated by the magnetic field and convection electric field in the solar wind. A significant fraction of pickup ions escape Mars’ gravitational pull and are lost to space. This non-thermal escape process of heavy species is an important mechanism responsible for atmospheric erosion. While there is a perception that the martian magnetic anomalies are significant for the ionospheric density distribution and the bow shock standoff location, little is known about the quantitative influence of the martian crustal magnetic field on the global distribution of escaping pickup ions. In this paper, we apply a newly developed Monte Carlo ion transport model to resolve the crustal field effect on the pickup oxygen ion distribution around Mars. The background magnetic and electric fields, in which test particles are followed, are calculated using an independent three-dimensional multispecies MHD model. The effects of the crustal magnetic field on particle escape are quantified by varying the crustal field orientation in the model setup and comparing the corresponding test particle simulation results. The comparison is made by turning on or off the crustal field or changing the local time of the strongest field from the dayside to the dawnside. It is found that without the protection of the crustal magnetic field, the total amount of atmospheric escape through the tail region would be enhanced by more than a factor of two. It is shown that the crustal magnetic field not only regionally deflects the solar wind around the martian atmosphere, but also has an important global effect on atmospheric erosion and thus on long-term atmospheric evolution.     

Crustal properties of Mercury by morphometric analysis of multi‐ring basins on the Moon and Mars

Abstract— We use satellite altitude free‐air and terrain gravity correlations to differentiate regional variations in crustal viscosity and transient cavity diameters of impact basins on the Moon and Mars that we combine with surface roughness for a rheological assessment of the crust of Mercury. For the Moon and Mars, we separate the free‐air anomalies into terrain‐correlated and terrain‐decorrelated components using the spectral properties of the free‐air and computed terrain gravity effects. Adjusting the terrain effects for the terrain‐correlated anomalies yields compensated terrain effects that we evaluate for crustal thickness variations of the impact basins. We interpret the terrain‐correlated anomalies for uncompensated elements of the crustal thickness variations that we find are strongly correlated with the distribution of basin rings from photogeologic analyses. Hence, we estimate the transient cavity diameter from the innermost diameter of the gravity‐inferred rings. Comparing these diameters with the related crustal thickness estimates clearly differentiates regional variations in the crustal rheologies. For the Moon, the analysis points to a farside crust that was significantly more rigid than the nearside crust during bombardment time. For Mars, the growth in transient cavity diameters with apparent crustal age also reflects increased viscosity due to crustal cooling. These results are also consistent with local estimates of surface roughness developed from the root‐mean‐squared topography over 64 times 64° patches centered on the basins. Hence for Mercury where gravity observations are lacking, rheological inferences on its crust may result from comparing photometric estimates of transient cavity diameter and local surface roughness with the lunar and martian estimates. These results for the Beethoven Basin, a typical multi‐ring impact feature of Mercury, suggest that the viscosity of the mercurian crust was relatively great during bombardment time. This enhanced rigidity, despite crustal temperatures that were probably much hotter than those of the Moon and Mars, may reflect an extremely dry crust for Mercury in its early development.     

Modeling of major martian magnetic anomalies: Further evidence for polar reorientations during the Noachian

Maps of the vector components of the Mars crustal magnetic field are constructed at the mapping altitude (360 to 410 km) using a selected set of data obtained with the Mars Global Surveyor magnetometer during 2780 orbits of the planet in 1999. Forward modeling calculations are then applied to six relatively strong and isolated, dominantly dipolar, magnetic anomalies for the primary purpose of estimating bulk directions of magnetization. Assuming that the magnetizing field was a (dipolar) core dynamo field centered in the planet, paleomagnetic pole positions are calculated for the six primary source bodies together with that for a seventh anomaly analyzed earlier. In agreement with several previous studies, it is found that six of the seven pole positions are clustered in what is now the northern lowlands in a region centered northwest of Olympus Mons (mean pole position: 34°±10° N, 202°±58° E). Assuming that the dynamo dipole moment vector was approximately parallel to the rotation axis, the modeling results therefore suggest a major reorientation of Mars relative to its rotation axis after magnetization was acquired. Such a reorientation may have been stimulated by internal mass redistributions associated with the formation of the northern lowlands and Tharsis, for example. A comparison of the mean paleo (magnetic) equator to the global distribution of crustal fields shows that magnetic anomalies tend to occur at low paleolatitudes. The same appears to be true for the Noachian-aged valley networks, which exhibit a broad spatial correlation with the magnetic anomalies. A possible interpretation is that the formation of magnetic anomalies and the valley networks was favored in the tropics where melting of water ice and snow was a stronger source of both surface valley erosion and groundwater recharge during the earliest history of the planet. This would be consistent with models in which hydrothermal alteration of crustal rocks played a role in producing the unusually strong martian magnetic anomalies.     

Melting in the martian mantle: Shergottite formation and implications for present‐day mantle convection on Mars

Abstract— Radiometric age dating of the shergottite meteorites and cratering studies of lava flows in Tharsis and Elysium both demonstrate that volcanic activity has occurred on Mars in the geologically recent past. This implies that adiabatic decompression melting and upwelling convective flow in the mantle remains important on Mars at present. I present a series of numerical simulations of mantle convection and magma generation on Mars. These models test the effects of the total radioactive heating budget and of the partitioning of radioactivity between crust and mantle on the production of magma. In these models, melting is restricted to the heads of hot mantle plumes that rise from the core‐mantle boundary, consistent with the spatially localized distribution of recent volcanism on Mars. For magma production to occur on present‐day Mars, the minimum average radioactive heating rate in the martian mantle is 1.6 times 10^?12 W/kg, which corresponds to 39% of the Wanke and Dreibus (1994) radioactivity abundance. If the mantle heating rate is lower than this, the mean mantle temperature is low, and the mantle plumes experience large amounts of cooling as they rise from the base of the mantle to the surface and are, thus, unable to melt. Models with mantle radioactive heating rates of 1.8 to 2.1 times 10 ^?12 W/kg can satisfy both the present‐day volcanic resurfacing rate on Mars and the typical melt fraction observed in the shergottites. This corresponds to 43–50% of the Wanke and Dreibus radioactivity remaining in the mantle, which is geochemically reasonable for a 50 km thick crust formed by about 10% partial melting. Plausible changes to either the assumed solidus temperature or to the assumed core‐mantle boundary temperature would require a larger amount of mantle radioactivity to permit present‐day magmatism. These heating rates are slightly higher than inferred for the nakhlite source region and significantly higher than inferred from depleted shergottites such as QUE 94201. The geophysical estimate of mantle radioactivity inferred here is a global average value, while values inferred from the martian meteorites are for particular points in the martian mantle. Evidently, the martian mantle has several isotopically distinct compositions, possibly including a radioactively enriched source that has not yet been sampled by the martian meteorites. The minimum mantle heating rate corresponds to a minimum thermal Rayleigh number of 2 times 10⁶, implying that mantle convection remains moderately vigorous on present‐day Mars. The basic convective pattern on Mars appears to have been stable for most of martian history, which has prevented the mantle flow from destroying the isotopic heterogeneity.