Heat Loss of the Earth and Energy Budget of the Mantle
Determination of the rate of Earth's energy loss is based a very large number of heat flux measurements in a variety of geological settings. Difficulties in integrating the flux over the Earth surface stem from two facts. One is that heat flux varies on a wide range of spatial scales and, in continents, is not a function of a single variable such as geological age, for example. The other difficulty is that the data exhibit large scatter. Advances in the interpretation of oceanic heat flux data are due to a thorough understanding of hydrothermal circulation through oceanic crust and sediments. In continents, the total heat loss has been constrained by sampling of old cratons is now adequate and systematic studies of heat flux and heat production have provided robust constraints on the crustal contribution to the surface heat flux. Heat loss through the ocean floor cannot be determined from the raw data because they are affected by hydrothermal circulation and irregularities in sediment cover. Predictions of the "half-space" model for the conductive cooling of oceanic lithosphere are consistent with heat flux measurements in selected "noise-free" environments as well as with the bathymetry of the sea floor. They are also consistent with values of the mantle temperature beneath oceanic ridges derived from petrology. This cooling model is also consistent with numerical calculations of mantle convection with plates. Using an accurate determination of the area extent of oceanic sea floor including marginal basins and accounting for enhanced heat flux over hot spots, we estimated the rate of heat loss through the oceans to be 32±2 TW (1012 Watts). This result is valid only for the present-day age distribution of sea floor and heat loss may have been different in the past when the distribution of sea floor ages was different from the present. For continents, bias due to the very uneven sampling of the surface heat flux is removed by area- weighting the average. The average value is independent of the size of the sampling window. Estimates of Moho heat flux beneath stable cratons obtained by several independent methods are within a range of 12-18 mW m-2. We estimate the continental heat loss to be 14±1 TW. Therefore, we obtain that the total heat loss of the Earth is 46±3 TW. Because continental heat production accounts for about half of the continental heat loss, i.e. 7 TW, the total heat loss from the mantle is 39±3 TW. It must be accounted for by radio-active heat production in the mantle, heat flux from the core, and secular cooling of the mantle. All these components are poorly constrained. The bulk silicate earth model suggests that the mantle heat production is 13±4 TW, and it is estimated that the core must lose 8±4 TW to maintain the geodynamo. The resulting estimates for the mantle Urey number are in the range 0.2-0.5.
Testing the Importance of Hydrothermal Circulation in Oceanic Lithosphere
Determining Earth's energy budget and the sources and mechanisms for heat transfer within it depends
largely on assumptions about the magnitude of the heat loss from the formation and cooling of oceanic
lithosphere, which covers about 60% of Earth's surface. Hofmeister and Criss (2005) have suggested that
hydrothermal circulation is not a significant factor in oceanic crust, and so they estimate that total global heat
flow is about 30 TW, about 25% less than previous estimates. The debate as to which approach to take
revolves about the following issues. Standard reference models for the cooling of oceanic lithosphere predict
the average variation in observed temperature-dependent properties, such as depth and heat flow with age, but
a significant discrepancy exists between observed and expected heat flow for ages less than about 65 million
years. Traditionally the difference is explained by significant hydrothermal circulation through the uppermost
oceanic crust removing some heat transferred by conduction from lithospheric depths. The fluid exits at bare or
poorly sedimented seafloor. In the absence of significant hydrothermal circulation using the observed heat
flow values results in about one third less heat flux and implies that conventional thermal models for the
oceanic lithosphere are inadequate. One test of this hypothesis is to examine whether marine heat flow
measurement sites are biased towards locations such as sediment ponds where hydrothermal circulation is
more likely to occur and convective heat loss should be large. Such sites may have been preferentially
selected before the issue of hydrothermal circulation was recognized, because most heat flow equipment
requires sufficient sediment for penetration. Most of the global geographic coverage is from old
measurements. We developed simple categories to characterize the local site environment and expected
large-scale variations in heat flow assuming hydrothermal circulation. When available, we use seismic
reflection records to assign categories for measurement sites in the old heat flow data set. We then examine
how the average site characteristics change with crustal age. Also, we examine how the fraction of the
measured to expected heat flow with age varies within each category to explore the importance of hydrothermal
Heat flux and crustal radio-activity near the Sudbury neutrino observatory, Ontario, Canada
During its next phase, the Sudbury neutrino observatory (SNO) will detect geoneutrinos, antineutrinos produced by the decay of U and Th in the Earth. These observations will provide direct constraints on the contribution of radiogenic heat production in the crust and mantle to the energy budget of the Earth. The geoneutrino flux at SNO depends on the local level of crustal radio-activity. Surface heat flux data record average crustal radio-activity unaffected by small scale heterogeneities. We review all available heat flux data measurements in the Sudbury structure as well as measurements of U, Th, and K concentrations in the main geological units of the area. With all available data, the average heat flux in the Sudbury basin is ~53mW m-2, higher than the mean value of 42mW m-2 for the entire Canadian Shield. The elevated heat flux is due to high heat production in the shallow crust. We estimate that the average heat production of the upper crust near Sudbury is >1.5μ W m-3 compared to an average of 0.95μ W m-3 for the Superior Province. The high crustal radio-activity near Sudbury results in an about 50% increase of the local crustal component of the geoneutrino flux. Crustal radio-activity is highest in the southern part of the structure, near the Creighton mine where SNO is located. High heat flux and heat production values are also found in the Southern Province, on the margin of the Superior Province. An azimuthal variation in the geoneutrino flux with a higher flux from the south than from the north is expected on the basis on the present information. However, we shall need better estimates of the contribution of the rocks in the Superior Province to the North to assess the extent of azimuthal effects. The many available exploration drill holes and core samples provide an opportunity to determine the spatial variations in crustal radioactivity near SNO and improve the interpretation of future measurements of the geoneutrino flux.
Potential and Limitations of Geoneutrino Detection Technique for Studying HPE in Earth's Deep Interior
The basic physics principles will be addressed briefly to identify what can be achieved with the geoneutrinos emitted by the heat producing elements (HPE) for constraning the concentrations in the Earth's interior. The potential and limitations of using the geoneutrinos will be examined to define goals for detector designs for assaying the HPE concentrations in the core and mantle. Results from kiloton neutrino detectors built already will be discussed.
U Solubility in Planetary Cores: Evidence from High Pressure and Temperature Experiments
Uranium is the most important heat producing element in the Earth and other terrestrial planets. The presence of an appreciable amount of U in planetary cores would have an important influence on Earth and planetary dynamics. In this study, the solubility of U in pure Fe, Fe-10wt% S and in Fe-35wt% S was measured by partitioning experiments with starting mixtures of peridotite, uraninite, Fe and FeS powder at pressure (P) of 0- 14.5 GPa and temperature (T) of 1500-2500 ° C. We found that in all run products, the solubility and partitioning of U in the pure metal or metal-sulfide phase relative to the silicate phase (DU) increases with increasing P and T. With a molten silicate phase, DU is generally 3-6 times larger than with a solid silicate phase, reaching a maximum of 0.15 for 10wt%S at 7 GPa. While DU has a positive dependence on S concentration of the metal-sulfide phase, there is a negative correlation between Ca and U. In addition, the experimental results indicate a P dependent decrease in oxygen fugacity, which is consistent with a depth dependent decrease in oxygen fugacity in the Earth's mantle as revealed in several previous studies. This is also supported by the positive correlation between Si concentration in liquid Fe and experimental pressure. Results from experiments and from mantle studies also imply that the oxygen fugacity at the core-mantle boundary may be much lower than previously expected. From our study, low oxygen fugacity is a favorable condition for U entry the core. Therefore, the experimental results that Si and U in the metal phase are positively correlated with pressure and negatively correlated with oxygen fugacity support U and Si inclusion in planetary cores at the time of core formation. According to our calculations based on these experimental results, if the core has formed from a magma ocean at a P of 26 GPa at its base and the core contained 10wt% S, then it could have incorporated at least 10 ppb U. Alternatively, if the core formed by percolation and contained 10wt% S, then it could have incorporated 5-22 ppb U. The geophysical implications of U in the cores of Earth and Mercury are discussed.
Geochemical Constraints On the Core-Mantle System?
The crystallization history of the inner core, the amount of heat generation in the core via decay of radioactive elements, and the mechanisms of heat transfer from the core to the mantle remain topics of continued debate. Detection of geochemical evidence of core-mantle interactions in mantle materials could help to constrain these parameters, as well as provide a means of tracing heat transport within the mantle. Tests for chemical interaction between the core and the mantle, as recorded in mantle-derived materials, have been previously postulated, based on long- (187Re-187Os, 190Pt-186Os) and short- (182Hf- 187W) lived radiogenic isotope systems, as well as the relative and absolute abundances of elements that are enriched in the core compared to the mantle. Osmium isotopes might even be useful for constraining the timing of inner core crystallization, as substantial crystallization must occur early in Earth history in order to generate an outer core that is isotopically resolvable from the ambient mantle. However, geochemical proof of core-mantle interaction continues to remain elusive. This could mean that: 1) no accessible mantle materials originate at the core-mantle boundary (CMB), 2) necessary assumptions/requirements regarding some of the potential tools (e.g., early inner core crystallization for Os isotopes) are not met by nature, 3) we do not currently have sufficient analytical resolution to detect interactions, or 4) chemical interactions at the CMB are insufficient to transmit signals. For example, some recent experimental work has shown that chemical and isotopic exchange of highly siderophile elements between the core and mantle may be less extensive than previously assumed. Nevertheless, better sample selection and higher resolution capabilities for some geochemical tests may yet yield fruitful results.
Chemical Interactions Between the Core and the Mantle
Gradual changes in temperature and composition in the core continually disturb the chemical equilibrium with minerals at the base of the mantle. The response of the core and/or the mantle to restore equilibrium depends on the details of the chemistry. Recent high-pressure experiments suggest that oxides and silicates in the mantle simply dissolve into liquid iron. It appears that the iron component of (Mg,Fe)O and (Mg,Fe)SiO3 is most soluble, leading to a transfer of O and Si into liquid iron. Once an equilibrium concentration of Si and O in the core is established, cooling and inner-core growth drive the light elements out of solution, under-plating the mantle with (relatively) iron-rich silicates and oxides. Exsolution of light elements may also provide an additional source of buoyancy for powering the geodynamo. To quantify these effects it is important to understand the current state of equilibrium at the core-mantle boundary. Most experiments find that mantle minerals in contact with a representative core composition are substantial depleted in iron. An Mg-rich layer at the base of the mantle is ineffective in isolating the core from the bulk of the mantle, so it is more likely that a light-element rich layer develops at the top of the core. Such a layer would be less than 10 km thick and could have a light-element concentration 50% higher than the bulk of the core. Cooling would still transfer light elements from the layer into the mantle, but the resulting change in composition would be small and the additional buoyancy source would be insignificant for the geodynamo. We examine the origin and evolution of a chemical layer at the top of the core, and explore the possible consequences for core dynamics.
Joint seismic-geodynamic-mineral physical constraints on heat flux across the CMB
The dynamics and thermal evolution of the Earth's interior is strongly dependent on the relative contributions from internal heating in the mantle (due to radioactivity and secular cooling) and from bottom heating across the core-mantle boundary (CMB). The dynamical style of the thermal convective flow, in particular the relative importance of active, thermally buoyant upwellings and mantle cooling due to descending lithospheric plates is also strongly dependent on the amplitude of heat flux across the CMB. We are able to provide new constraints on the convectively maintained heat flux across the CMB thanks to recent progress in mapping the lateral variations in mantle temperature by jointly inverting global seismic and geodynamic data sets, in which mineral physical constraints on mantle thermal heterogeneity are also imposed (Simmons et al. 2009). We present here new models of the present-day global mantle convective flow predicted on the basis of the thermal and non-thermal (compositional) density perturbations derived from the new tomography model and using the inferences of depth-dependent, horizontally averaged mantle viscosity derived from joint inversions of glacial isostatic adjustment and mantle convection data (Forte and Mitrovica 2004). We employ this tomography- geodynamics based mantle convection model to explore the convective transport of mass (buoyancy flux) and heat (advected heat flux) across the lower and upper mantle. We show that the predictions of advected heat flux at the top of the seismic D" layer provide direct constraints on the heat flux across the core-mantle boundary (CMB). Our current best estimates of the present-day CMB heat flux are in excess of 10 TW. We present a sensitivity analysis showing the degree of robustness of this inference, depending on the inferred variation of mantle viscosity in the lower mantle. We also present new predictions of the present-day distribution of secular heating and cooling at different depths in the mantle.