Magsat Anomalies Research Articles

Satellite magnetic anomalies of the Antarctic crust

Spatially and temporally static crustal magnetic anomalies are contaminated by static core field effects above spherical harmonic degree 12 and dynamic, large-amplitude external fields. To extract crustal magnetic anomalies from the measurements of NASA's Magsat mission, we separate crustal signals from both core and external field effects. In particular, we define Magsat anomalies relative to the degree 11 field and use spectral correlation theory to reduce them for external field effects. We obtain a model of Antarctic crustal thickness by comparing the region's terrain gravity effects to free-air gravity anomalies derived from the Earth Gravity Model 1996 (EGM96). To separate core and crustal magnetic effects, we obtain the pseudo-magnetic effect of the crustal thickness variations from their gravity effect via Poisson's theorem for correlative potentials. We compare the pseudo-magnetic effect of the crustal thickness variations to field differences between degrees 11 and 13 by spectral correlation analysis. We thus identify and remove possible residual core field effects in the Magsat anomalies relative to the degree 11 core field. The resultant anomalies reflect possible Antarctic contrasts due both to crustal thickness and intracrustal variations of magnetization. In addition, they provide important constraints on the geologic interpretation of aeromagnetic survey data, such as are available for the Weddell Province. These crustal anomalies also may be used to correct for long wavelength errors in regional compilations of near-surface magnetic survey data. However, the validity of these applications is limited by the poor quality of the Antarctic Magsat data that were obtained during austral Summer and Fall when south polar external field activity was maximum. Hence an important test and supplement for the Antarctic crustal Magsat anomaly map will be provided by the data from the recently launched Ørsted mission, which will yield coverage over austral Winter and Spring periods when external field activity is minimal.

Resolving source parameters through pseudo potential anomalies

Ambiguity is inherent in the interpretation of all geopotential anomalies. The ambiguity can possibly be reduced by generating pseudo potential anomalies and comparing them with the observed anomalies. A 0.25° digitized gravity anomaly data set of the Indian subcontinent and adjoining regions was used to generate a pseudo magnetic anomaly map at an altitude of 400 km. This map was compared with the MAGSAT scalar anomaly map and there seems to be a good reproduction of MAGSAT map over the Himalayas. However, anomalies over east and west coasts show latitudinal shifts of about 2° when compared to MAGSAT anomalies. The southern peninsular of India is characterized by strong negative Bouguer anomalies, possibly due to the bias of Indian ocean geoid low. The pattern of pseudo magnetic map as obtained from gravity data differs from that of MAGSAT suggesting different sources of gravity and magnetic anomalies over southern India.

Determination of depths to centroids of three-dimensional sources of potential-field anomalies with examples from environmental and geologic applications

A method is developed for determining the depth to the centroid (the geometric center) of `semi-compact' sources. The method, called the anomaly attenuation rate (AAR) method, involves computing radial averages of AARs with increasing distances from a range of assumed source centers. For well-isolated magnetic anomalies from `semi-compact' sources, the theoretical AARs range from ∼2 (close to the sources) to ∼3 (in the far-field region); the corresponding theoretical range of AARs for gravity anomalies is ∼1 to ∼2. When the estimated source centroid is incorrect, the AARs either exceed or fall short of the theoretical values. The levelling-off of the far-field AARs near their theoretical maximum values indicates the upper (deeper) bound of the centroid location. Similarly, near-field AARs lower than the theoretical minimum indicate the lower (shallower) bound of the centroid location. It is not always possible to determine usable upper and lower bounds of the centroids because the method depends on characteristics of sources/anomalies and the noise level of the data. For the environmental magnetic examples considered in this study, the determined deeper bounds were within 4% of the true centroid-to-observation distance. For the case of the gravity anomaly from the Bloomfield Pluton, Missouri, USA, determination of only the shallower bound of the centroid location (∼7 km) was possible. This estimate agrees closely with the centroid of a previously determined three-dimensional model of the Bloomfield Pluton. For satellite magnetic anomalies, the method is appropriate only for high-amplitude, near-circular anomalies due to the inherent low signal-to-noise ratio of satellite magnetic anomalies. Model studies indicate that the AAR method is able to place depths within ±20–30 km of actual center locations from a 400-km observation altitude. Thus, the method may be able to discriminate between upper crustal, lower crustal, and mantle magnetic sources. The results from the prominent Kentucky anomaly are relatively well-resolved (centroid depth ∼30 km below the Earth's surface). For the Kiruna Magsat anomaly, the deleterious effects from neighboring anomalies make a determination difficult (possible depth could be between 20 and 30 km). The centroid depths are deeper for the Kursk anomaly (∼40–50 km). These depths may indicate that magnetic anomalies from the near-surface Kursk iron formations (a known contributor) and deep crustal magnetic sources could combine to form the Kursk Magsat anomaly.

A new interpretation of the MAGSAT anomalies of Central Europe

The MAGSAT map for Europe shows two significant sign-reversed magnetic anomalies. The negative one is located over Palaeozoic crust, whereas the positive one is over Precambrian crust. As an alternative to a previously published interpretation a three-dimensional model body is presented situated only over Palaeozoic crust showing a magnetization of constant Permian direction but reversed polarity. Varying magnetization intensities are associated with different parts of the body. The predominant remanent magnetization requires a Q ratio clearly above 1. The possible carriers of the magnetization could be pyrrhotite and fine-dispersed magnetite.

Magnetic petrology of arc xenoliths from Japan and Aleutian Islands

Xenoliths from Japan and the Aleutian Islands were studied for the purpose of identifying source rocks with sufficient magnetization to account for positive subduction zone Magsat anomalies. Two types of magnetic lithologies are identified: (1) amphibolites and hornblende gabbros containing up to 8.6% titanomagnetite, mostly as large, unoxidized, multidomain crystals, and (2) pyroxene gabbros and metagabbros with up to 20.3% titanomagnetite, usually in association with primary ilmenite. The first type is represented by xenoliths from Ichinomegata, Japan and Adak Island in the Aleutians. The second, restricted to Iki Island, Japan and Kanaga Island, Aleutians, has abundant coarse‐grained titanomagnetite which in most samples has been oxidized, producing secondary, trellis‐type ilmenite lamellae; a few Iki Island xenoliths show more advanced oxidation. Arc xenoliths show 3 orders of magnitude range in natural remanent magnetization and saturation remanence and about 2 orders of magnitude range in magnetic susceptibility. Nearly all Iki Island rocks and many from Ichinomegata and the Aleutians possess induced magnetization 2 A m−1 and greater, with the expectation of additional magnetization due to viscous remanence at elevated temperature. The TiO2 content of titanomagnetite, together with the presence of appreciable Al2O3 and MgO, ensure that most xenoliths have Curie points between 400° and 550°C. Remanence acquisition curves exhibit a range reflecting varying degrees of magnetic hardening due to oxidation. Correspondingly, each xenolith site groups distinctly on a plot of magnetic hysteresis ratios. This paper indicates that lithologies with total magnetizations of many A m−1 exist in lower crust arc settings. Their volumetric distribution is uncertain based on the limited xenolith population reported herein.

Comment on “An interpretation of the MAGSAT anomalies of Central Europe” by Taylor and Ravat

Comment on “An interpretation of the MAGSAT anomalies of Central Europe” by Taylor and Ravat

An interpretation of the Magsat anomalies of central Europe

Magsat studies of central Europe have revealed correspondence between tectonic elements /geological structures and long-wavelength satellite-altitude anomalies. The most striking correlation is the two Magsat anomalies (one positive and the other negative) associated with either side of the Tornquist-Teisseyre Zone (TTZ). The TTZ, which extends from the Black Sea to the Baltic Sea, represents the structural boundary (suture) between the younger, thinner, and hotter crust of the Hercynide or Paleozoic block (negative anomaly) and the older, thicker, and colder crust of the East European Precambrian Platform (positive anomaly). This produces a gradient of approximately 18 nT (peak-to-trough) and a half-wavelength of 720 km along a Magsat orbit at 325 km altitude. This anomaly pair is the result of the juxtaposition of these two significantly different crustal blocks. To model this field, two irregular three-dimensional bodies were used to represent the Precambrian and Paleozoic crust. The former was modeled with a normal magnetization vector while the latter was modeled with a reverse vector. Aeromagnetic anomalies in western Germany and rock magnetic measurements from Germany and Scandinavia support a model in which reversely magnetic geologic units, extending from upper-crustal metasediments to mid-crustal sources produce this negative anomaly.

Contributions of Cretaceous Quiet Zone natural remanent magnetization to Magsat anomalies in the Southwest Indian Ocean

The Magsat magnetic anomalies over the Southwest Indian Ocean are modeled using a combination of induced plus viscous remanent magnetization (IM/VRM) and natural remanent magnetization (NRM). Two broad, roughly parallel, SW to NE trending triple‐peaked positive anomalies dominate the region, one lying south of Africa and the other north of Antarctica. Although these anomaly peaks generally correspond with the Agulhas Plateau/Maud Rise, Mozambique Plateau/Astrid Ridge, and Madagascar Ridge/Conrad Rise conjugate pairs, the IM/VRM contribution from structural characteristics (i.e., crustal thickness) accounts for only about 20% of the anomaly amplitudes. A spatially variable but observationally constrained NRM contribution in Cretaceous Quiet Zone (KQZ) crust is required to account for the location, shape, and amplitude contrast of these anomalies. Many crustal features in the Southwest Indian Ocean near Antarctica have little geophysical data to constrain their structure but do have tectonic conjugates near Africa for which much more geophysical data are generally available. Using geophysical and geological constraints from one member to model the magnetization structure of its conjugate reproduces the observed Magsat reduced‐to‐pole anomalies over both structures very well. This suggests that no significant alteration in their magnetization structure has occurred since the features split. Models of these conjugate structures show that IM/VRM reproduces the Magsat anomalies associated with non‐KQZ crust but that both IM/VRM and a dominant NRM component are required to explain the anomalies associated with KQZ crust.

Induced magnetization of the oceanic lithosphere and ocean‐continent magnetization contrast inferred from Magsat anomalies

A petromagnetic forward model of induced magnetization of Iceland Plateau is presented. Magnetic susceptibilities of Icelandic rocks from the literature are placed in a vertical stratigraphic context according to a model of igneous and metamorphic layering of the Icelandic crust that is compatible with seismic and other geophysical data. Curie isotherm depths for the magnetic forward model are based on the geothermal gradients measured in drill holes and those derived from magnetotelluric surveys. The petromagnetic model is used to predict the satellite altitude (400 km) magnetic field of Iceland Plateau. Comparison of calculated and Magsat anomalies shows that the total induced magnetization (the vertically integrated induced magnetization) in the surrounding oceanic lithosphere is ≈2300 A. This value is very similar to independent estimates based on rock magnetic models of the oceanic crust Rockall Plateau is a Precambrian microcontinent largely surrounded by oceanic crust a few hundred kilometers south and east of Iceland. The size, age, and location of this plateau are ideally suited to characterize the ocean‐continent magnetization contrast The total induced magnetization contrast between the plateau and its surroundings estimated from the Magsat anomaly of Rockall Plateau is ≈14,000 A.

European tectonic features observed by Magsat

Regional three-dimensional magnetic models have been developed to characterize the principal European long-wavelength magnetic anomalies represented on the improved Magsat magnetic anomaly map of Europe. The magnetic models were constrained by regional variations in geology and geophysical parameters (e.g., geologic boundaries, crustal thickness, heat flow). Because only limited measurements of magnetization are available on lower crustal and uppermost mantle rock samples, our results are useful in constraining and understanding the overall magnetization of these regions. Illustrations of these include: (1) geologic provinces across the Tornquist-Teisseyre tectonic zone; (2) regions of thin crust and high mantle heat flow in south-central Europe; (3) the Kursk-Voronezh magnetic anomaly; and (4) the Ladoga-Gulf of Bothnia zone. The region of the Tornquist-Teisseyre tectonic zone, that marks the boundary between the Fennoscandian-Baltic Shield and metastable Europe, is a major magnetic discontinuity. In south-central Europe, the regional magnetic variations appear to be directly related to variations in the lower crustal thickness and possibly also to heat flow. In addition, the famous Kursk (Ukraine) iron-ore deposit produces a prominent bullseye anomaly at satellite altitude. The Kiruna Magsat anomaly is modelled as having a large, deep body as its source. The high P-wave velocity, basal crustal layers encountered in rift (e.g., the Tornquist-Teisseyre tectonic zone itself) and continental arc (e.g., the Ladoga-Gulf of Bothnia zone) settings of Europe appear to be nearly non-magnetic.

A preliminary study of satellite-derived gravity and magnetic anomalies of India and contiguous regions

The residual satellite gravity anomaly map in and around the Indian region has been prepared. These residual anomalies are interpreted for crustal inhomogeneities using the MAGSAT anomalies as one of the constraints. While a strong correlation between these anomalies is observed over the Dharwar, Aravalli, Marwar and Singhbhum regions, inverse correlation prevails over the Eastern Ghats and Cambay basin. The positive gravity and magnetisation (strong correlation) over the stable blocks of the peninsula is related to the alternating layers of basic and ultrabasic materials; the positive gravity and low magnetisation (poor correlation) over the Eastern Ghats and Cambay basin is interpreted as thinning of the crust. The residual gravity map identifies the signatures of west-central Indian lineament and the submarine topographic features/ridges in the Indian Ocean such as the Chagos-Laccadive, 85°E and Ninety East ridges, which are not so obvious in the unprocessed anomaly map at ground level. An attempt is also made to explain the anomaly over the Indo-Ganga-Brahmaputra basin using the ground- and 400-km-level satellite gravity anomaly maps.

The Agulhas Magsat anomaly: Implications for continental break-up of Gondwana

A new processing procedure is implemented on Magsat data over the southern African region and surrounding oceans. The oceanic region of the ensuing map is interpreted in this study. The anomaly map is image processed and a composite image of Magsat with Seasat data is also presented. The Agulhas high-amplitude satellite-magnetic anomaly, situated off the southeastern coast of Africa in the southwestern Indian Ocean (31°E, 35°S), is clearly defined. In addition, a north-south oceanic anomaly bordering the west coast of southern Africa, including in part the Walvis Ridge anomaly, is now recognised. The Walvis Ridge and Agulhas oceanic anomalies are interpreted to represent thickened, remanently magnetised (during Cretaceous quiet times) ocean crust. Similarities between the Walvis Ridge and the Agulhas Magsat anomalies suggest a common genesis. The Agulhas anomaly is interpreted to represent the remnant scar of the process that led to fragmentation of Gondwana. It is hypothesised that fragmentation of Gondwana was initiated by an upwelling of hotter than normal asthenosphere centred directly below the Agulhas anomaly.

Magnetic anomaly modeling at the Indo Eurasian collision zone

One of the consistent features of satellite crustal magnetic anomaly maps is the large negative anomaly in the Indo Eurasian collision zone. This region is modeled by the introduction of magnetic blocks representing the Tibetan, Asian and Himalayan mountain regions. The crustal thickness of these blocks is based on the principle of isostasy. It has been shown that the removal of the core field from the Magsat data removes most of the long-wavelength anomalies associated with the ocean-continent susceptibility contrast, and that some of the observed anomalies, after processing of the satellite data, are a manifestation of this contrast. To account for this, the final model is a superposition of the above block model on a global model of typical continent and ocean susceptibility, called a Standard Earth Magnetization Model (SEMM). A combination of the three blocks superposed on the SEMM reproduces the major features of the observed Magsat anomalies. Block model susceptibilities of the Asian and Himalayan regions are the same as for the continental portion of the SEMM, albeit with a greater crustal thickness. Thus, the crust in these regions is typical. By contrast, the satellite magnetic data require an integrated susceptibility in the Tibetan region which is considerably lower than typical. Two possible explanations are suggested. One, that the origin and tectonic history of the Tibetan region differs from the neighboring regions in such a way as to result in lower susceptibility. This is consistent with current thinking that the Tibetan plateau is made up of several fragments or micro-continents which were attached to Asia prior to the collision of India with Asia. The other possible explanation is that the Curie isotherm in the Tibetan region is elevated compared to the Asian and Himalayan regions.

Magnetization of the Pacific Ocean lithosphere deduced from Magsat data

Magsat anomalies over the Pacific Ocean are related to volcanic plateaus and the edge effect of the Pacific Cretaceous Quiet Zone (PCQZ). Four prominent plateaus, Shatsky Rise, the Mid‐Pacific Mountains, Hess Rise, and Manihiki Plateau, are considered in this study. The major lateral magnetization contrasts associated with these plateaus are considered to arise from the topographic swells and the crustal roots, with root thicknesses estimated from topography using an Airy isostatic model. Shatsky Rise and the Mid‐Pacific Mountains are older than the Cretaceous Quiet Period and their natural remanent magnetization (NRM) is not a significant component of the magnetization contrasts that cause the Magsat anomalies associated with these plateaus. Their magnetic properties are characterized by effective magnetic susceptibility contrasts of 0.07 (±0.01, S.I., per unit volume) averaged over their topographic swells and crustal roots, equivalent to 0.04 if averaged over the entire crustal thicknesses (swell plus root plus intervening layer). This value agrees well with the estimates of previous investigators for these plateaus and for similar plateaus in other oceans. Two plateaus of Cretaceous Quiet Period age within the PCQZ, Hess Rise and Manihiki Plateau, have origins similar to those of Shatsky Rise and the Mid‐Pacific Mountains, and they probably have similar effective susceptibility contrasts but they may also carry Cretaceous Quiet Period NRM. Magnetic anomalies due to Hess Rise and Manihiki Plateau, calculated from the effective susceptibility contrast deduced for Shatsky Rise and the Mid‐Pacific Mountains, are substantially larger than the observed Magsat anomalies of these plateaus. There are also Magsat anomalies over parts of the Pacific where no plateaus exist Some of these anomalies are demonstrated as probably due to a contrast in NRM between the PCQZ and its surroundings. The total remanent magnetization contrast (magnetization contrast in amperes per meter times thickness in meters, equivalent to the dipole moment per unit area) of the PCQZ is 14,000–20,000 amperes (±30%), which cannot be simply explained by the absolute values of remanent magnetization predicted by conventional models of oceanic crustal magnetic properties. Modelling of the Cretaceous quiet zones in the northern Atlantic indicates a similar total remanent magnetization contrast. These results may be explained by including uppermost mantle remanent magnetization, with an intensity of about 1–2 A m−1 if magnetization persists to a depth of ≈7 below the seismic Moho, i.e, down to the depth of the crustal roots of the plateaus. Including NRM in the uppermost mantle of the PCQZ also serves to reduce the calculated amplitude of the Hess Rise and Manihiki Plateau Magsat anomalies if the NRM of the crustal roots is less than the NRM of the surrounding uppermost mantle. On this basis the estimated intensity of NRM in the uppermost 7 km of the PCQZ mantle is ≈1–3 A m−1, in very good agreement with the intensity estimated from the PCQZ edge effect anomalies. These estimates of uppermost mantle NRM are probably a maximum limit, recognizing that other sources may also contribute to the total remanent magnetization contrasts of the Cretaceous quiet zones.

Magsat magnetic anomaly contrast across Labrador Sea passive margins

Many passive margins not complicated by nearby anomalous crustal structure have satellite elevation crustal magnetic anomaly contrasts across them that are recognizable in reduced‐to‐pole versions of the Magsat and POGO data. In the Labrador Sea region this contrast is particularly well developed with strong positive anomalies overlying the continental crust of Greenland and eastern Canada and prominent negative anomalies situated over the Labrador Sea and Baffin Bay. We use forward modeling of the large‐scale crustal bodies in this region (continental, oceanic, passive margin, several anomalous structures) to show that the Magsat anomaly contrast is due simply to the change in crustal susceptibility and thickness at the continental/oceanic crustal transition. Because the thickness varies more than the average susceptibility from continental to oceanic crust, the strong anomaly contrast is essentially an edge effect due mostly to the change in crustal structure.

Thermoremanent magnetization of the oceanic lithosphere inferred from a thermal evolution model: Implications for the source of marine magnetic anomalies

Thermoremanent magnetization models of the oceanic crust and the uppermost part of the oceanic mantle are derived on the basis of a typical thermal evolution model of the oceanic lithosphere using different magnetic blocking temperature ranges. The magnetization of the basaltic layer 2A has a similar pattern, but with a lower intensity, to that of the standard sea-floor spreading magnetic model; however, the magnetization patterns of the lower crust and upper mantle are skewed, so that in a given polarity period the total magnetization of a lithospheric column which is created early in the period is greater than that of a column which is created later in the period. Magnetic anomalies of the magnetization models are calculated at sea level and at MAGSAT elevation for slow, moderate and fast sea-floor spreading velocities. Although the magnetization models are derived based on purely physical reasoning without using any a- priori information about the observed marine and MAGSAT magnetic anomalies, they do yield magnetic anomalies which are similar to the observed anomalies. The models explain several features of the marine and MAGSAT anomalies. The anomalous skewness of marine magnetic anomalies is explained in terms of the magnetization of the oceanic lower crust. The magnetization of the oceanic upper mantle may account for the anomalous skewness observed over the Cretaceous quiet zones in the Atlantic Ocean. The skewness of the magnetic anomalies calculated at sea level decreases with the increase in the sea-floor spreading velocity.

Magnetic anomalies at satellite altitude over continent-ocean boundaries

A pronounced difference in vertically integrated magnetization produces strong magnetic anomalies at satellite altitude over continent-ocean boundaries. However, these anomalies are seldom seen in satellite anomaly maps. Our simulation of the technique normally used to extract anomaly fields from satellite observations of the geomagnetic field suggests that power leaks from short-wavelength crustal anomalies into the longer wavelengths. This phenomenon distorts the anomaly field and adds power to the lower degrees in the anomaly field power spectrum. We will also show that some of the few magnetic anomalies observed in satellite anomaly maps over continent-ocean boundaries in the Atlantic ocean could be caused by thermal destruction of magnetic minerals rather than by a pronounced difference in vertically integrated magnetization. The very low permeability of sediments deposited along continental margins basically precludes convective heat transfer in the sediments. The low thermal conductivity of these sediments results in the underlying lithosphere having a high temperature due to the temperature gradient within the sediment blanket. Our calculations predict that thermal destruction should occur along continental margins where several kilometres of sediments have accumulated. In order to produce magnetic anomalies similar to MAGSAT anomalies, we assume that the upper mantle is magnetized.

The Indian Ocean MAGSAT anomalies and strong geomagnetic field during Cretaceous ‘quiet’ zone

Abstract The palaeomagnetic palaeointensities, natural remanent magnetic intensities of ocean-floor basalts and MAGSAT anomalies over the central Pacific and North Atlantic oceans suggest that the geomagnetic field was appreciably stronger during the ∼ 35 Ma long Cretaceous ‘quiet’ zone of fixed normal polarity. It is shown here that two MAGSAT anomalies over this ‘quiet’ zone in the central and eastern Indian Ocean, one over the Broken Ridge and the other south of Sri Lanka, which have been interpreted differently by other workers, also support this correlation of the geomagnetic field's long-term polarity stability and strength.

Thermoviscous remanent magnetization of oceanic lithosphere inferred from its thermal evolution

Remanent magnetization patterns of the oceanic lithosphere determined on the basis of the thermal evolution of the lithosphere, while taking into account both thermal and viscous magnetizations, give rise to magnetic anomalies similar to those observed by Magsat over the Cretaceous quiet zones of the Atlantic Ocean and by marine magnetic surveys over anomaly 33. Viscous magnetization acquired during a given core field polarity period is not removed completely by that acquired in the next period of opposite polarity because of cooling of the lithosphere and the strong temperature dependence of viscous magnetization. The magnetization pattern of layer 2A resembles that of the standard seafloor spreading magnetic model, consisting of normally and reversely magnetized zones separated by vertical boundaries, and produces symmetric magnetic anomalies. The magnetization pattern of the middle crust is similar to that of layer 2A for high magnetic blocking temperature models, but it is slightly skewed with respect to layer 2A for low magnetic blocking temperature models. The magnetization pattern of the lower crust is strongly skewed relative to that of layer 2A and has diffuse boundaries with pronounced slopes between normally and reversely magnetized zones. It produces magnetic anomalies with appreciable skewness, suggesting that observed skewness of marine magnetic anomaly 33 arises from the magnetization of this layer. The magnetization pattern of the upper mantle is appreciably skewed and almost devoid of small‐scale variations. It produces strongly skewed magnetic anomalies at satellite altitudes which may explain the observed skewness of Magsat anomalies over the Cretaceous quiet zones of the Atlantic Ocean.

Long‐wavelength magnetic anomalies in the Caribbean: Plate boundaries and allochthonous continental blocks

Vector magnetic anomaly maps are derived at satellite altitude from Magsat data over the Caribbean and Middle America. To minimize smoothing of the anomaly field caused by averaging the data over the whole altitude range of Magsat measurements, two maps are derived in two nonoverlapping altitude ranges: 350–420 and 420–550 km. A composite map of the vertical component of the anomaly field is then computed at ground level. This map displays large positive anomalies associated with fast subduction zones. These anomalies define discontinuities in the downgoing slab associated with seismic gaps. Positive anomalies are also consistently observed above allochthonous blocks of continental origin in the Caribbean area and suggest that the average magnetization is larger in the continental lithosphere. Elongated magnetic lows are observed above transform boundaries of the Caribbean plate. The Cayman Trough anomaly is consistent with magnetization of the order of 1 A/m in the upper mantle. Demagnetization of the lithosphere along the Bocono‐El Pilar fault system which bounds the Caribbean plate to the south may be accounted for by frictional heating along shear zones. The pattern of anomalies along the southern margin of the United States is related to its accretionary tectonic history. Various estimates of the vertically integrated magnetization of both oceanic and continental lithospheres are computed using these anomalies and provide constraints on the vertical distribution of lithospheric sources of Magsat anomalies.