T43B-1318 1340h
Numerical modeling of mantle plume diffusion
To clarify the influence of the heat diffusion on the mantle plume evolution, we develop a two-dimensional numerical model of the plume diffusion and relevant efficient numerical algorithm and code to compute the model. The numerical approach is based on the finite-difference method and modified splitting algorithm. We consider both von Neumann and Direchlet conditions at the model boundaries. The thermal diffusivity depends on pressure in the model. Our results show that the plume is disappearing from the bottom up - the plume tail at first and its head later - because of the mantle plume geometry (a thin tail and wide head) and higher heat conductivity in the lower mantle. We study also an effect of a lateral mantle flow associated with the plate motion on the distortion of the diffusing mantle plume. A number of mantle plumes recently identified by seismic tomography seem to disappear in the mid-mantle. We explain this disappearance as the effect of heat diffusion on the evolution of mantle plume.
T43B-1319 1340h
Scale and Amplitude of Heat Flow Variations at the Base of the Continental Lithosphere
Thermal conditions at the base of the continental lithosphere provide a constraint that models of mantle convection must satisfy. The scale of the surface heat flow variations controls how they affect the thermal structure and thickness of the lithosphere. In order to downward continue the heat flow to the base of the lithosphere, one must determine the average heat flow over a scale which is sufficiently large: the thicker the lithosphere, the larger the horizontal averaging scale. Thus, the determination of lithospheric thickness and average heat flow cannot be treated independently. Within a single craton, such as the North American craton for example, there are significant variations of surface heat flow at short spatial scales which bear no consequence for the deep lithosphere. Data in the Canadian Shield and the Appalachians are now extensive enough to address problems of scale and relationship between average heat flow and heat production. We have analyzed the entire data set as well as data from five compositionally distinctive subprovinces ranging in age from Archean ($>$2.5Ga) to Mid-Proterozoic (1.1Ga). Within each province, on scales $<$ 500 km, observed heat flow variations are linked to changes of local crustal structure. For the five subprovinces, the average values of heat flow ($\overline{Q}$) and heat production ($\overline{A}$) conform to the simple relationship $\overline{Q} = Q_o + H \overline{A}$, where $H \approx$ 9km and $Q_o $\approx 33 \mathrm{mW~m^{-2}}$. This shows that, on scales larger than the dimensions of these provinces ($>$ 500 km), variations in crustal heat production dominate and hence that variations of mantle (Moho) heat flow must be small. For instance, the large heat flow step at the Grenville-Appalachian boundary ($\approx 16 \mathrm{mW~m^{-2}}$) may be accounted for by a change in crustal heat generation only. In that case, the lithosphere is $\approx$ 40 km thinner in the Appalachians than in the Shield. At wavelengths of 500 km or more, mantle (Moho) heat flow variations are constrained to be smaller than the detection limit of heat flow studies, or about $\pm 2 \mathrm{mW~m^{-2}}$ , and may not be correlated with surface geology. Downward continued to the base of the lithosphere, the amplitude of these variations depends on wavelength and must be smaller than $\approx 7 \mathrm{mW~m^{-2}}$. Such variations imply that temperature differences must be smaller than 400 K at 150 km depth. These bounds are consistent with seismic shear wave velocity variations and geothermobarometry studies on mantle xenoliths. Because of the thickness of the lithosphere, these low amplitude long wavelength variations do not include transient or short period (i.e. $<$ 500 Myears) thermal fluctuations at the lithosphere-asthenosphere boundary which have no effect on surface heat flow. Without these fluctuations, the heat flow over large cratonic areas remains low on very long time scale.
T43B-1320 1340h
The Thermal Structure of Subduction Zone Backarcs
A remarkable feature of subduction zones is that the mantle beneath the volcanic arc and well into the backarc is hot, even in areas that have not undergone recent extension. The occurrence of active volcanism at subduction zones suggests that temperatures in the mantle wedge below the arc are $>$ 1200 \deg C, despite the cooling effects of the subducting slab. Such high temperatures are inferred to result from mantle flow above the slab, which carries heat into the subduction zone from below. A key constraint on backarc mantle dynamics is the thermal structure of the mantle wedge, especially well behind the arc. We determine the detailed thermal structure of several backarcs using a number of independent indicators of temperature (e.g., surface heat flow, mantle seismic velocity, xenoliths, effective elastic thickness, lithosphere thickness, thermal isostasy studies, mantle viscosity). We focus our analysis on six backarcs that have not undergone recent ($<$ 50 my) extension: 1) northern Cascadia, 2) Mexico/Central America, 3) South America, 4) Alaska/eastern Aleutians, 5) Kamchatka, and 6) Sunda. For each region, the observations show that high temperatures in the mantle wedge (1200\deg C at 60 km depth) extend well into the backarc regions, with little variation for 100's to 1000's of km behind the arc. Similar high temperatures are inferred for extensional backarcs in the western Pacific and southern Europe/Asia, but their thermal structure is complicated by extension and spreading. Although there are site-specific explanations for the high temperatures in each backarc, we propose that a hot mantle wedge and backarc is a fundamental feature of subduction zones. The only evidence for a cool backarc is the Peru flat slab region of South America, where the flat subducting slab may exclude a hot mantle wedge. Observations of a hot and nearly isothermal backarc mantle place a very restrictive constraint on backarc mantle dynamics that has not been widely recognized. In particular, such high temperatures are inconsistent with the thermal structure predicted by numerical models of slab-driven corner flow, which can produce high temperatures beneath the arc but low temperatures in the shallow backarc mantle further landward. We conclude that vigourous small-scale convection is required in the backarc upper mantle. High temperatures and hydration of the mantle wedge by the subducting slab may reduce the backarc mantle viscosity, allowing convection, which rapidly carries heat upwards from depth.
T43B-1321 1340h
Numerical models of the Earth's thermal history
A number of different scenarios have emerged for the thermal evolution of the Earth which is controlled by the slow convective motion in the mantle in order that it meet the constraint imposed by the modern-day surface heat flow. These include models with high internal heating rates in the mantle, models with significant degrees of internal heating in the core, models with large and/or time-varying degrees of layering at 660-km depth and models with weak feedback between surface heat flow and mantle Rayleigh number. Most investigations of the Earth's thermal evolution have been carried out using parameterized convection studies in which the governing equations are significantly simplified. In this contribution, we compare the results of more complex numerical models with those of parameterized models for the various thermal history scenarios described above. Although these types of models are in general agreement in a time-averaged sense, only the former can assess short-timescale and lateral variability and these allow for new constraints to be brought to bear.
T43B-1322 1340h
Interpretation of Heat Flow and Seismic Data from Yellowstone Lake, Yellowstone National Park
The heat flow map of Yellowstone Lake shows a complex thermal regime with variations of more than two orders of magnitude. The major geothermal systems mapped are within the caldera boundary, in the northern part of the lake and in West Thumb. In the northern part of the lake two areas of extremely high heat flow are identified: an area east of Stevenson Island (heat flow more than 10000 mW/m$^{2}$), not previously known to have high heat flow and Mary Bay-Sedge Bay geothermal system (up to 40000 mW/m$^{2}$). These are separated by an area with background heat flow of 1500 to 2000 mW/m$^{2}$ which correlates with large sediment thickness. Therefore the whole area from east of Stevenson Island to the Mary Bay is interpreted as a single geothermal system variably attenuated by the conductive effects of the sediments. Seismic data show that the western boundary of this high heat area is the complex graben which extends from the Lake Hotel in an approximately north-south direction east of Stevenson Island. We present evidence (heat flow, seismic and geologic features observed from underwater video) that the Hotel Fault, one of the graben faults, extends on the east side of the Stevenson's Island, not on the west. West of the graben the heat flow could be as low as 150 mW/m$^{2}$. Although several topographic notches similar to the hydrothermal vents east of Stevenson Island were identified west of the graben structure, the nearby heat flow values are low, suggesting the presence of cold, not hot springs. The decrease of the gradients to the south is more gradual and is not associated with major faulting or with the caldera boundary. The other important geothermal system is West Thumb, but no extremely high heat flow values were found. The areas east of Stevenson Island and Mary Bay - Sedge Bay are identified as potential geothermal hazards. However, in Mary Bay, at the hottest point located in the lake, the boiling point temperatures are reached at more than 5 m below the bottom of the lake, in a water depth of 40 m. If a geothermal explosion were to occur it would require a sudden drop of water level with at least 6 m. By connecting several discontinuous features it has been argued that the maximum possible length of the Hotel Fault, the largest fault in the lake, is 25 km. However, the fault was mapped only north of Stevenson Island (for about 3 km). In the south it could be masked by deformed area in the east of Stevenson Island (up to 8 km). No link to the Eagle Bay fault to the south was observed in sparker records and thus the maximum fault length is only 12 km. We argue that a fault 12 km long could not offset the bottom of the lake by 6 m. The overall low stress condition in the caldera (weak crust) support the presence of a high number of small events rather than large single fault events.
T43B-1323 1340h
Thermal Conductivity of Complex Materials at High Temperatures and Pressures
Measurements of lattice thermal diffusivity have been obtained on (Mg$_{0.9}$Fe$_{0.1}$)$_{2}$SiO$_{4}$ olivine, its high-pressure polymorphs, and of (Mg$_{0.9}$Fe$_{0.1}$)SiO$_{3}$ enstatite at conditions of temperature and pressure. We used two independent techniques: Impulsive Stimulated Light Scattering to 5 GPa and 1273 K and the $\AA$ngstr\"{o}m method on cylindrical samples in a multianvil apparatus to 20 GPa and 1373 K. Lattice thermal conductivities were calculated from heat capacities and equations of state. Conductivities are consistent with previous results obtained at 1 atm. In olivines conductivity increases at each phase transition from olivine to spinel structures (corresponding to velocity discontinuities at depths of 410 and 520 km in the Earth). Of greatest significance is the fact that for each of these phases, lattice conductivity closely follows a T$^{-1/2}$ dependence on temperature T, a prediction of Klemens (1960) for materials containing abundant lattice defects. If such a dependence applies to other silicates and complex crystals, there should be a useful way to estimate conductivities at high temperatures from room temperature measurements.
T43B-1324 1340h
Using Continental Elevation to Estimate Heat Production Distributions
Elevation is useful constraint on the thermal structure of the lithosphere and is used extensively in studying oceanic lithosphere evolution. On continents, the use of elevation is limited because of the relative complexity of continental lithosphere thermal regimes, and in particular by uncertainties in lithospheric heat production distributions. We are exploring these uncertainties with sensitivity studies linking crustal and subcrustal heat production profiles with elevation. Construction of steady-state geotherms requires knowledge of four parameters: surface temperature; surface heat flow; thermal conductivity, $k$; and heat production, $A$. Surface temperature and heat flow are determined directly at the Earth's surface, but $k$ and $A$ must be estimated as a function of depth. We test the sensitivity of elevation to variations in $k$ and $A$. Sensitivity tests are conducted by computing the elevation differences due to geotherms with heat flow values of 40-120 mW/m$^{2}$ and systematically varying $k$ and $A$. The effect of varying $k$ between 2.1 to 3.9 W/m/K result in elevation adjustments ranging from +1.1 to -0.6 km at 40 mW/m$^2$ and +0.2 to -0.2 km at 120 mW/m$^2$. Surface $A_{0}$ values used are computed by assuming a sub-crustal heat flow of 40 to 100% of the total heat flow. $A(z)$ are then decreased through the crust via an exponential decreasing relationship. The remaing contribution to the surface heat flow are generated internally by HPE in the crust. Radical changes in $A$ produce elevation differences of 3 km at 40 mW/m$^2$ and 0.25 km at 120 mW/m$^2$. Hence, heat production is the dominant uncertainty in interpreting the integrated thermal structure from elevation, particularly in regions of low heat flow. Since shields have had time to reach steady-state, variations in heat flow may come from variations in sub-crustal heat flow and variations in thermophysical parameters. Assuming sub-crustal heat flow is constant and all variations in heat flow and elevation between cratons arise from heat production, an estimate of $A$ may be made by interpreting anomalous elevation. Results from an analysis of 12 cratonic regions reveal seven provinces with elevations more than 500 m above predicted, and one province with an elevation less than 500 m below expected. These elevation anomalies may be accounted for by reasonable local variations in heat production.
T43B-1325 1340h
A Reconstruction of Mantle Plume Evolution by Optimal Heat and Mass Transportation: Impact of Mantle Diffusion and Viscosity
The ascent and evolution of mantle plumes depend on the properties of the source region and the viscosity and thermal diffusivity of the ambient mantle. While mantle properties are relatively constant during a hundred Myr lifetime of most plumes, source region properties can vary substantially with time as the thermal boundary layer feeding the plumes is depleted of hot material. Complete local depletion of the boundary layer cuts the plumes off from their source. It is the subsequent evolution of the plumes and the reconstruction of this evolution that interests us here. We study effects of the heat diffusion on the evolution of mantle plumes at lower Rayleigh numbers ($Ra$). Numerical experiments show that a delay or interruption in plume heat and mass transfer results in the diffusive disappearance of plume tails first and plume heads later. This is the most likely explanation for the seismically detected mid-mantle plumes. We restore these diffused plumes to their prominent states in the past and analyze the restoration errors for various $Ra$. Also we study the effect of the lower viscosity of mantle plumes on their evolution and restoration.
T43B-1326 1340h
2-D Thermal Modelling Across the Geothermal Fields of Tuscany, Italy.
The lithospheric extension affecting the Tuscan-Tyrrhenian domain represents one of the most relevant and recent tectonic processes within the entire Alpine-Mediterranean deformation area. The heat input from the mantle is responsible for the presence of large geothermal resources at accessible depths in the crust, as testified by temperature and heat flow anomalies, locally extremely high. Despite intensive exploration and exploitation drilling programs carried out in Tuscany, mainly since the 70s, the nature, physical properties and structure of the intermediate and lower crust and of the upper mantle are still debated. The available dataset for the Tuscan area was significantly improved by the acquisition of the deep crustal seismic reflection profiles (CROP Profiles), in the mid 90s. The profiles CROP 18A and 18B, crossing NW-SE wards the Larderello and Monte Amiata geothermal fields, and the CROP 03 profile, intersecting with W-E direction the CROP 18B, were recently reprocessed to better characterize the crustal and upper mantle structures of the entire Tuscan geothermal area. The results show new remarkable and interesting features, i.e.: the presence of extensional structures below the "K Horizon" regional high-amplitude discontinuous reflector; a second deeper and more continuous similar horizon; mantle intrusions; strong reflectors in the lower crust and a discontinuous crust/mantle transition with possible underplating. Accurate analyses of the seismic attributes suggest the presence of fluids/melts from the "K Horizon" down to about 10 km depth. These new data were put into a simple and conceptual 2-D model, aimed to provide a set of preliminary thermal models, to be compared with the experimental borehole temperature and heat flow data. The 2-D numerical modelling followed a two-steps process: first, modelling the regional conductive heat transfer in the upper 10-12 km of the crust and, secondly, superimposing local advection, in correspondence of the geothermal fields. The 2-D regional conductive model was realized by means of a steady-state forward simulation, under the assumption of a purely conductive heat transfer. The unknowns are the basal heat flow and the thermal properties of the crustal rocks, whereas the results are the temperature distribution with depth and the surface heat flow. To account for the uncertainties in the physical properties of the crustal rocks we produced two sets of models, using the parameters and assumptions which maximise and minimise, respectively, the surface heat flow output to be compared with the experimental data. Local heat transfer by advection was introduced in the upper crustal structures of the geothermal fields, where the CROP seismic profiles were indicating presence of fluids. The temperature, depth and extension of these reservoirs can explain most of the present extremely high surface heat flow anomalies.
T43B-1327 1340h
Relationships between heat flow, thermal and pressure fields in the Gulf of Mexico
The thermal field of the Gulf of Mexico (GoM) is restored from a comprehensive temperature-depth database. A striking feature is the systematic sharp gradient increase between 2500 and 4000 m. The analysis of the pressure (fracturation tests and mud weights) indicates a systematic correlation between the pressure and temperature fields, as well as with the thickness of Plio-Pleistocene sedimentary layer, and is interpreted as the fact of cooling from fluid flow in the upper, almost hydrostatically pressured layer. The Nusselt number, that we characterize by the ratio between the near high-P gradient over low-P gradient varies spatially and is correlated to the structural pattern of the GoM; this observation outlines the complex relationships between heat and fluid flows, structure and sedimentation. The deep thermal signal is restored in terms of gradient and heat flow density from a statistical analysis of the thermal data combined to the thermal modelling of about 175 wells. At a regional scale, although the sedimentary cover is warmer in Texas than in Louisiana in terms of temperature, the steady state basal heat flow is higher in Louisiana. In addition, beneath the Corsair Fault, which lay offshore parallel to the Texan coast, the high heat flow suggests a zone of Tertiary lithospheric thinning.
T43B-1328 1340h
Re-arrangements of Global Plate Motion: Role of True Polar Wander (TPW)
Plate-motion models constructed by R.G. Gordon and D.M. Jurdy (1986) and at Harvard by R.J. O'Connell et al. (1991), further developed by O. Cadek and V. Ricard (1992), show that in sum Cenozoic plate motion has been concentrated around the degree 1 harmonic spectral component, representing net lithosphere rotation (NLR) about the contemporary Pole. Participants allowed for uncertainties in using hotspots as benchmarks. In the absence of an alternative to tidal action to account for perennial convection asymmetry, otherwise enigmatic (D. Bercovici 2003), plate motion under the observed degree 1 system is here compared with the regime to be expected under geocentric plus external gravity. Mantle convection takes place under a minute westward tilt in globally-averaged {\it g}, contributed by a tidal component having the value -$2.16\deg$ identified by G.J.R. MacDonald (1964). Previously impossible, R.D. Ray et al. (2001) recently have combined satellite tracking data and Topex/Poseidon altimetry to separate the solid-earth fraction of the associated dissipation, 110 +/- 25 GW, from the obscuring marine signal. Hitherto, in model construction it has been necessary to assume almost perfect elasticity. Surprisingly large in terms of earlier estimates of the dissipation factor 1/Q, the value obtained conforms notably with supposition as per Cadek and Ricard, that the upper mantle remains constantly at the point of failure representing advance, in precisely this mode, of the internally driven convection basically responsible for global tectonics. The regime is self-reinforcing and prone to be stable. How might it be interrupted? Some 25 years ago P.A. Rona and E.S. Richardson (1978) identified the global plate reorganization which took place in the Lower Cenozoic. Essentially, reorganization was directional in nature, characterized by supplantation of basic geotectonic features such as seafloor spreading and subduction having a N-S orientation, by similar features oriented E-W. It may now be significant that with increased understanding of the paleomagnetic record (M. Prevot et al. 2000; J. Besse and V. Courtillot 2002), shift of Earth's pole of rotation (TPW) has become a well demonstrated occurrence, although agreement is far from universal as to its modus operandi. Prior to -130 Ma interpretation still is subject to much uncertainty, the major subsequent event having been the end of the -130 to -60 Ma period of relatively fast polar wander. Standstill may have set in as from about -50 Ma. An earlier period of standstill, currently estimated to extend from -200Ma to about -150 Ma, was co-extensive with the Jurassic, being replaced by the highly TPW-active Cretaceous. Agreement has not been reached as to the rapidity of a $20\deg$ ultra-fast episode or `jump' culminating at around -115 Ma, at the time of most active break-up of Gondwana into its constituent continents and formation of such ranges as the Rockies and European Alps. If the Cadek/Ricard summary of plate motion or one similar is sound, TPW entails equivalent change in the direction 'west'. The liberty is now taken to suggest that rather than being subsidiary to unexplained reorganizations of global tectonics, TPW is causative and primary. Thermodynamics requires that under TPW, junctures must occur at which Earth's heat efflux is effected most efficiently via convection compliant with a greatly different Pole location, compelling the occurrence of a fast episode or `jump' of the ilk reported by Prevot et al. Within the grossly inhomogeneous, convective Earth, the operation of the conjugate factors NLR and TPW seems basic; of which, the latter has constituted the primary geotectonic determinant.
T43B-1329 1340h
New insights on the seafloor spreading patterns of the northern South China Sea (16$\deg$N-23$\deg$N)
There are several models which attend to describe the spreading history of the South China Sea and its sea-floor spreading evolution. Based on the magnetic anomaly systematic study in the past, the South China Sea Basin can be divided into the southwestern and eastern parts by the several spreading orientation change stages. The eastern basin is revealed asymmetric between magnetic anomaly isochron 11 (about 32 Ma) and the ceased of the spreading of the South China Sea (magnetic anomaly isochron 5c) (Briais et al., 1993). Recent magnetic lineations showed the existence of magnetic anomaly 15-17 in the northernmost South China Sea area. However, the spreading features of the northern South China Sea (older then 32Ma) and their tectonic correlation with the whole South China Sea are still not well understood. To understand the rifting structures and their tectonic implications in the northern South China Sea, 9 multi-channel seismic profiles (ORI645, ORI654, ORI689, ORI693, ACT, EW9509, MLTW, 97304Aa and 97034Ab), the magnetic data collected off southwestern Taiwan (117$\deg$E~121$\deg$E longitude and 18.5$\deg$N - 23$\deg$N latitude), NGDC magnetic trackline data (National Geophysical Data Center from 1950 to present), CCOP magnetic data (Coordination Committee for Geoscience Programmes in East and Southeast Asia), satellite gravity anomaly data, the bathymetry (NGDC and swath bathymetry) are used. Based on the seismic data, the post-rift oldest sedimentary features were identified to be uplifted in the northeastern South China Sea (117.5$\deg$E-120.5$\deg$E and 19$\deg$N-22$\deg$N). The oldest features can be divided into a N45$\deg$trending part and a N90$\deg$trending part, separated by a possible N-S trending fracture zone. This feature is clearly absent at the north region of the LRTPB (Luzon Ryuku Transform Plate Boundary) and could be prolonged southward based on the geophysical data (gravity data , the swath bathymetric data and the magnetic data). Furthermore, in the whole northern South China Sea (112$\deg$E-121$\deg$E longitude, 16$\deg$N-23$\deg$N latitude), the orientation of the spreading has a major change between magnetic anomaly isochron 9 and 10. The orientation of the magnetic anomaly 11-17 area revealed E-W to ENE-SWS. However, the oldest sedimentation features in this area trend NE-SW. It is interesting that the anomaly 11 to 17 part of the oceanic crust is not existed at the south side of the South China Sea. There is no evidence which the northern anomaly 11-17 part was the extinct portion of the other spreading system(not belonged to South China Sea). Therefore, the anomaly 11-17 south part of the oceanic crust could be reasonably interpreted as subducted portion beneath the northwest Palawan.
T43B-1330 1340h
Hans Pettersson and the First Heat Flow Probe for Deep Sea Studies
Hans Pettersson (HP) organized and led the Swedish Deep Sea Expedition 1946-47, and invented many marine research tools. His study (1910,1914) of heat generation during radioactive decay made him interested in Joly's ideas (1903,1925)of heat generation in the Earth. Joly's work made HP convinced of the geochemical role of the deep sea volcanism. These ideas were bolstered by the Austrian geologist Kirsch, who had a fairly modern view (1928) of how the Earth's inner heat caused convections and moved the crust, ideas that where approved by Wegener. HP therefore decided to study oceanic heat flow. He con-structed a heat-gradient probe, with a kerosene filled bulb and three riser tubes, 2 serving as thermometer capillaries and one as an expansion tube. Two clockworks controlled a multi-port valve, one tube being closed just before the penetration of the sediment, the other 45 minutes later. The first 2 stations gave credible readings but later on the instrument failed. HP believed this was due to poor thermal compensation, but more likely it was caused by sticking of the valves - just as in modern titration work when the valve grease is being squized out. Refrigerater tests of "old" clocks show that the spring drives work well for 9-10 hours or twice the time HP needed to get a heat gradient.
T43B-1331 1340h
Effects of Seamounts and Lithospheric Structure on the Pacific Seafloor Topography
One fundamental observation about seafloor topography is the reduced dependence on lithospheric age on relatively old seafloor (older than 70 Ma), relative to the prediction from the half-space cooling model. Understanding the origin of this deviation has been one of the central issues in geodynamics, as it has important implications for lithospheric thermal evolution and large-scale mantle dynamics. Various explanations have been proposed to be responsible for this deviation, including lithospheric reheating and erosion, mantle plumes, and statistical bias resulting from seamounts and large igneous provinces (LIP). This study focuses on the Pacific seafloor with two objectives: 1) to quantify the effects of LIPs and seamounts published by Wessel [2001] on Pacific seafloor topography and its relation to crustal age; and 2) to determine the contribution of lithospheric thermal structure as documented in a recent surface-wave seismic model to the Pacific seafloor topography. In our analyses, the topography data was analyzed by removing all 8,000 seamounts and LIPs. We find that the ocean depth vs age curve does not change significantly with the removal of seamounts and LIPs. Seamounts removed at 2 and 3 radii have an average deviation from the unfiltered topography averages of 82 and 106 meters, respectively for all ages. For ages greater than 70 Ma the deviation increases to 115 and 145 meters. However, lithospheric structure seems to have relatively large effects on seafloor topography. For the Pacific, we also construct the residual topography maps that use the half-space cooling model, the plate model, and our seismic tomography model as reference models.