The structural framework of the European continent is the result of several geodynamic events that produced a complex “collage” of crustal blocks differing in geological and geophysical properties (Fig. 2). Central Europe formed while the supercontinent Rodinia dismantled from Neoproterozoic events (McCann, 2008), yielding the tectonic structures of the Alpine system, the Carpathians chain, the Pannonian Basin, the Bohemian Massif and the Trans-European Suture Zone (TESZ). The TESZ is the main crustal boundary in central Europe as a result of the movement and juxtaposition between the stable and older northeastern East European Craton (EEC), or Precambrian platform, and the younger southwestern Paleozoic Platform (PP) during the Caledonian and Variscan orogenic events (e.g., Pharaoh, 1999; Thybo, 2001; Banka et al., 2002). The TESZ includes two main structural segments: the Sorgenfrei–Tornquist Zone (STZ) in the northwestern part and the Teisseyre–Tornquist Zone (TTZ) in the southeastern part. It extends from the central North Sea via the Baltic Sea (northwest) to the Black Sea (southeast) (Pharaoh, 1999). In northeastern Europe, the Baltica paleocontinent evolved during several Precambrian episodes of crustal accretion and reworking. It comprises the Fennoscandian and Ukrainian shields and the East European Craton. The EEC extends from Denmark into Russia on the eastern side and includes the segments of Sarmatia, Volgo–Uralia and Fennoscandia, which differ in forming dynamics and lithological and tectonic features (Bogdanova et al., 2006). Generally, the EEC is characterized by a thick and cold crust composed mainly by Precambrian rocks evolved during Archean–Paleoproterozoic episodes of accretion and reworking (Gaàl and Gorbatschev, 1987) and covered by thin series of Phanerozoic rocks (Plant et al., 1998).

The structural setting of the European crust has been modeled thanks to large amounts of seismic and potential field data that allowed for an investigation of the lithosphere down to 60 km of depth (e.g., Guterch et al., 1999; Pharaoh, 1999; Mushayandebvu et al., 2001; Banka et al., 2002; Majorowicz and Wybraniec, 2011). Several seismic reflection surveys, such as BIRPS (UK), CROP (ltaly), DEKORP (Germany), ECORS (France), NFP 20 (Switzerland) and the European Geotraverse (EGT) project aimed at developing a three-dimensional representation of the geological structures and properties of the European lithosphere (Freeman and Mueller, 1990; Blundell et al., 1992; DEKORP Research Group, 1988). Further deep seismic studies were carried out in Europe such as POLONAISE'97, CELEBRATION 2000, ALP 2002 and SUDETES 2003 (Guterch et al., 1999, 2003; Grad et al., 2003; Brückl et al., 2003) in Poland, as well as the KTB deep-drilling surveys in Germany (Gebrande et al., 1989), revealing the strong contrast between the PP, the EEC and the structural features of the TESZ area. Seismic data interpretation revealed that the cratonic part of Europe is characterized by anomalous thickness, wherein the roots may reach a maximum depth of 55–60 km (Artemieva and Meissner, 2012), such as in the Ukrainian, Uralides and Baltic shields. The Variscan terranes in central Europe instead show a uniform crust thickness of around 28–32 km (e.g., EUROBRIDGE Seismic Working Group, 1999; Guterch and Grad, 2006). These variations in crustal thicknesses are shown in Fig. 3a using the Moho depth map of Europe produced by Grad et al. (2009). A clear depth variation is observed in the TESZ area, passing from the thin Paleozoic Platform to the EEC, with a 50 km increase in crustal thickness in southern Sweden, Denmark, the Baltic Sea, Poland and Slovakia (BABEL Working Group, 1993; Giese and Pavlenkova, 1988; Thybo, 1990, 2001; Korja and Heikkinen, 2005; Guterch and Grad, 2006).

Figure 3(a) Depth to the Moho beneath the European continent (data from Grad et al., 2009). (b) Heat-flow map of Europe (data from Davies, 2013).

Download

Deep-borehole temperature measurements also reveal a strong difference in geothermal properties between western and eastern Europe (e.g., Čermák et al., 1989; Hurting et al., 1992; Plewa, 1998; Królikowski, 2006) as illustrated by the heat-flow map (Fig. 3b). Low regional heat-flow values (30–40 mW m⁻²) were measured within the Baltic shield (Precambrian platform), which is typical of a thick lithosphere, whereas relatively high heat-flow values were observed in specific areas in depressions of the EEC basement (e.g., Čermák and Rybach, 1979). Central Europe is instead dominated mainly by extended high heat-flow zones trending from east to west (Chiozzi et al., 2005).

Reverse magnetization in central Europe is evidenced by petrological analysis of Permian quartz porphyries beneath the northern German sedimentary basin, containing pyrrhotite–metasedimentary rocks with susceptibility values up to $\mathrm{2}\times {\mathrm{10}}^{-\mathrm{3}}$ cgs and a Koenigsberger ratio (Q) of about 100 (Henkel, 1994). Further studies of Paleozoic metamorphic rocks revealed reversely magnetized sources in western Germany (Pucher, 1994), and additional analysis of metasedimentary rocks showed a possible overprinting during the Permo-Carboniferous reverse superchron (Thominski et al., 1993).

A satellite magnetic field model (e.g., MF7) has been used successfully to define the crustal magnetic field for wavelengths λ>350 km; however, it is unable to resolve anomalies at shorter wavelengths due to the high altitude of satellite measurements. The CHAMP MF7 crustal field was the best available satellite-derived field at the time of the study (Maus, 2010). Due to the satellite's orbital height the wavelength resolution of this field was ∼150 km. To investigate the nature and cause of the anomaly field associated with the TESZ and CEML, terrestrial magnetic data are needed. This study used the European and Mediterranean Magnetic Project (EMMP) dataset from Getech (Fletcher et al., 2011). The EMMP compilation used all original flight-line data and ground data available from each county. So, the resolution of many surveys allowed an optimum grid to be generated down to 1 km. Technical data concerning these surveys also allowed the International Geomagnetic Reference Field (IGRF) correction to be applied. With any compilation, merging surveys with a range of spatial survey sizes, ages, instruments and processing methods can and did generate small differences, which will not necessarily average out over the compilation. Since the EMMP compilation generally lacked long-wavelength control (i.e., $\mathit{\lambda }\ge \sim \mathrm{150}$ km) due to the limited size of surveys, the final processing step, after gridding at 1 km, was to drape it onto the CHAMP's MF7 crustal field.

As indicated in the Introduction, the CEML has been previously interpreted on the basis of satellite-altitude magnetic field data as the result of coalescence between small-scale magnetic anomalies, with mainly reverse polarization. We performed analyses at different altitudes (100–350 km) to identify the contributions of the various magnetic sources to the CEML at different scales. The maps at these altitudes were produced with the upward continuation technique, a standard procedure to transform field data measured at a given altitude to a higher level. The transformation consists of a convolution, which may be performed in the Fourier domain (Baranov, 1975; Fedi et al., 2012). We first filled small areas of no data with a maximum entropy interpolation (Gibert and Galdeano, 1985) and then performed a draped-to-level upward continuation to a 5 km altitude level using the wavelet-based algorithm by Ridsdill-Smith (2000) and the digital elevation model SRTM30 (Farr et al., 2007) (Fig. 4a). Then, we performed a level-to-level upward continuation using the classical convolution integral:

where $U(x,y,z)$ is the potential field upward continued to the scale z, $U(\mathit{\xi },\mathit{\eta },\mathrm{0})$ is the potential field at level z=0 and $z/\left[\right(x-\mathit{\xi }{)}^{\mathrm{2}}+(y-\mathit{\eta }{)}^{\mathrm{2}}+z^{\mathrm{2}}{]}^{\mathrm{3}/\mathrm{2}}$ is the upward continuation operator. In Fig. 4b we show the map of the upward continuation magnetic field at 100 and 350 km of altitude. Note that, in principle, Eq. (1) is applicable to fields known continuously on the entire flat surface. Our data instead comprise discrete datasets containing equally spaced data distributed on a finite surface. However, as shown in Fedi et al. (2009), accurate upward continuation may be computed if the data are known on a surface larger than the area of interest and/or by extrapolating data beyond their borders with a suitable algorithm. Upward continuation of data may indeed produce a border (or edge) effect near the grid margins due to space-domain aliasing (Oppenheim and Schafer, 1975; Fedi et al., 2012; Mastellone et al., 2014). To reduce this effect, upward continuation has been performed by extrapolating the aeromagnetic dataset onto a larger area using a maximum entropy algorithm.

Figure 4Aeromagnetic field map of Europe at 5 km (a), 100 km (b) and 350 km of altitude (c); map of MF7 magnetic field model at 350 km of altitude (d).

Download

Furthermore, it may be useful to calculate the extent ΔL of the upward continuation operator $z/\left[\right(x-\mathit{\xi }{)}^{\mathrm{2}}+(y-\mathit{\eta }{)}^{\mathrm{2}}+z^{\mathrm{2}}{]}^{\mathrm{3}/\mathrm{2}}$ in order to estimate the portion of the map that may be mostly affected by the border effect. We may simply map the operator at different altitudes and evaluate ΔL from its half-width. We define ΔL as the distance occurring from the maximum to the point at which it is decreased by a factor 1∕e, with e being the Neper number. Thus, we estimated ΔL=96 km at 100 km of altitude, which means that, using ΔL, we may be rather confident in the upward-continued field computed at 100 km of altitude over most of the European continent, excluding a border defined by ΔL. At 350 km of altitude, the border effect is significantly greater with ΔL=347 km, as is the discrepancy with the MF7 model map (Fig. 4c). Therefore, the total gradient analysis at satellite altitude is performed on the map of the MF7 magnetic data at 350 km (Fig. 4d). The European magnetic field map at 5 km of altitude is mostly characterized by short-wavelength anomalies. The first and most important feature of the map is the visible contrast in the distribution and intensity of the anomalies located over the Paleozoic and EEC regions. As revealed by seismic modeling, the TESZ is a sharp geological–tectonic boundary here, which controls the main regional trend of the magnetic field. It suggests a very sharp change in the crust's magnetic properties in addition to geological, compositional and geothermal differences. The anomalies associated with the TESZ define the most important magnetic anomaly zone of central Europe and relate to sources involving the whole crust. The TESZ magnetic anomaly is especially visible in Poland and Romania, and it divides these countries into two opposite magnetic provinces: the southwestern area, almost covered by magnetic lows, and the northeastern region, characterized by numerous magnetic highs. Actually, this anomaly trend may be extended over the European continent, identifying the TESZ as the margin between these two opposite magnetic behaviors (Banka et al., 2002; Williamson et al., 2002; Fairhead, 2015). Such magnetic behavior is imaged even more clearly at 100 km of altitude, at which the map shows a considerably smoothed magnetic field when compared to that at 5 km. Here, the strong difference between the EEC and Paleozoic crusts is well appreciable by the development of an extended magnetic low in central Europe in contrast with an area of magnetic highs to the NE, anticipating the formation of the reverse dipolar anomaly above the TESZ structure and the CEML at 350 km of altitude (Fig. 4c). Following Ravat et al. (1993) the origin of the CEML could be associated with the considerable difference in crustal thickness and heat flow between the EEC and Paleozoic platforms. Taylor and Ravat (1995), however, carried out an alternative interpretation based on the predominance of remanent magnetization within the Paleozoic crust. Their sketched model consists of two prismatic blocks representing, respectively, the reversely magnetized Paleozoic Platform and the normally magnetized Precambrian crust. Therefore, the main origin of the CEML was assigned to a pattern of relatively small reversely magnetized sources in the upper–middle crust of the Paleozoic Platform, coalescing at satellite altitude and forming the extended long-wavelength magnetic low. Pucher and Wonik (1998) proposed a further interpretation based on reversely magnetized sources in the Paleozoic crust and on the magnetic properties of central Europe, while the EEC was not considered to contribute to CEML formation. The effect of coalescence between short-scale anomalies, shared by these magnetic models of central Europe, could be verified if we simultaneously compare the field map at small and large altitudes.

The dipolar shape of the magnetic field anomalies represents a troublesome limitation in the study and the identification of the magnetic sources due to the non-vertical direction of total magnetization and the inducing field. This issue may be overcome by employing methods able to remove the dipolar behavior of the field almost completely and provide maps of anomalies more directly related to the source position. One such method is the total gradient technique.

The total gradient modulus of the magnetic field T (|∇T|) is defined as

where $\partial T/\partial x$, $\partial T/\partial y$ and $\partial T/\partial z$ are the partial derivatives of the total magnetic field (T) with respect to the directions x, y and z.

The total gradient anomalies have two important features:

1. they are located directly over the source bodies and may help localize the sources edges (e.g., Roest et al., 1992; MacLeod, 1993; Roest and Pilkington, 1993; Salem et al., 2002; Shearer and Li, 2004; Florio et al., 2011; Paoletti et al., 2016); and

2. they are positive anomalies, no matter whether the magnetization is direct or reverse. In fact, |∇T| is completely independent of the directions of both the inducing field and total magnetization in the 2-D case (Nabighian, 1972) and only weakly dependent on the directions of T in the 3-D case (Haney et al., 2003). Salem et al. (2002) pointed out that, in the 3-D case, a maximum shift for inclination (I) of 30^∘ does not cause appreciable changes in the total gradient amplitudes.

In this paper we specifically perform a joint analysis of total gradient and magnetic field anomalies to identify the presence of sources with a strong remanent component of magnetization and to interpret whether large-scale magnetic anomalies can be produced by the coalescence effect. To clarify the problem, we consider a synthetic model composed of three magnetic sources with top depth at 1 km and bottom depth at 5 km: B1 and B2 have normal magnetization ($I=I_{\mathrm{F}}=\mathrm{65}$^∘, D_F=3^∘), while B3 is reversely magnetized with $I=-\mathrm{15}$^∘ and D=230^∘ (Fig. 5), where I_F and D_F indicate the direction of the inducing field, and I and D indicate those of the total magnetization. These parameters were chosen according to the results of the paleomagnetic analysis in central Europe carried out by Pucher (1994) and that were used to interpret the regional magnetic field over Europe by Pucher and Wonik (1998). The map of the magnetic field (T) at 10 km of altitude (Fig. 5a) shows two dipolar anomalies, B1 and B2, with their magnetic lows and highs located to the NE and SW, respectively. The source B3, in contrast, is not clearly identified, since its magnetic effect is covered by the anomalies produced by B1 and B2. The presence of an anomaly due to B3 could perhaps be identified by the slight prolongation of the magnetic high of B2 towards the location of B3 and a by weak increase in the intensity of the magnetic low of B1 to the SW. This example shows that the presence of a reversely magnetized source in a complex magnetic environment is hard to identify and isolate from the map of the magnetic field. The total gradient intensity map of the magnetic field (Fig. 5b) transforms the dipolar shape of the magnetic anomalies into high-amplitude monopole-shaped anomalies, which now provide detailed information on the source locations. The reversely magnetized source B3 is now clearly imaged and distinguished from the other sources. Note that in the map of the magnetic field the field lows (blue) stand for the lows of the dipolar magnetic anomalies, while in the total gradient intensity map the smallest values are around zero and, most importantly, denote the areas in which magnetization is very weak. So, the total gradient intensity highs yield an easy and effective tool to indicate the relative presence or absence of magnetized sources and the location of the sources, even in the case of strong remanent magnetization. We will use both these properties to analyze the magnetic field in Europe.

Figure 5Multisource model with B1 and B2 normally magnetized and B3 reversely magnetized. (a) T map; (b) |∇T| map.

Download

As previously indicated, the two main hypotheses put forward to explain the CEML anomaly are the following:

• A.

  a pattern of reversely magnetized sources in the upper–middle crust of the Paleozoic Platform, coalescing at satellite altitude; and

• B.

  geological contact between a thick magnetized crust (EEC) and a thin low-susceptibility crust (central Europe) with high heat flow.

In order to evaluate these hypotheses, we will now consider two synthetic cases. The first is related to hypothesis A. To this end, we produced a new model with 19 prismatic sources with depths ranging between 2 and 10 km of depth to simulate the magnetic crust of central Europe in a reversely magnetized environment. The magnetization intensity ranged between 1.5 and 2.5 A m⁻¹, and the direction of magnetization was the same as that of the reversely magnetized body B3 in Fig. 5. The inducing field parameters were declination 3^∘ and inclination 65^∘. The magnetic field and total gradient intensity maps were calculated at 1, 100 and 350 km of altitude (Fig. 6). At small altitudes, the short-wavelength anomalies clearly identify the location of the sources both in the T and |∇T| maps. At 100 km of altitude the magnetic field is considerably smoothed, and the coalescence between the magnetic anomalies begins to be relevant. However, at such altitudes it is extremely difficult to identify the source location from the T map.

Figure 6Synthetic model of the multiple shallow sources reversely magnetized. The magnetic field (T) and total gradient |∇T| maps are shown at 1 km (a), 100 km (b) and 350 km (c) altitudes.

Download

On the other hand, the total gradient intensity map allows for the recognition of the location of the magnetic sources, especially those with higher magnetization that produce stronger anomalies at large altitudes. Note that, similarly to the previous model, the magnetic lows in T are mostly substituted by peaks of |∇T| highs, again demonstrating the ability of the total gradient technique to estimate the presence of reversely magnetized sources.

In Fig. 6c the magnetic field at 350 km of altitude is characterized by a large reverse dipolar anomaly. The intensity of such a magnetic anomaly is very weak, meaning that shallow sources have a very low contribution at satellite altitudes, which could be completely negligible in the presence of additional deeper sources. Here the total gradient intensity map shows a large-amplitude high located in proximity to the magnetic low in the southwestern region and suggests the presence of highly magnetized bodies. These results show that, in the case of the coalescence effect, we can even identify areas of strong magnetization by the total gradient intensity regardless of the direction of magnetization.

The second hypothesis B is analyzed by building a new synthetic model of the magnetic crust in order to simulate the magnetic anomaly above the TESZ. We consider two adjacent prismatic bodies representing the Paleozoic and the EEC with crustal thickness of 30 and 40 km, according to a seismic model (e.g., Guterch and Grad, 2006), and induced only magnetization of 1 and 2 A m⁻¹, respectively. Thus, we assume no remanent magnetization for both the sources. The magnetic field and the total gradient intensity calculated at 350 km of altitude are shown in Fig. 7. The map of the magnetic field shows a reverse dipolar magnetic anomaly with a magnetic high above the synthetic red body (Precambrian crust) and a related magnetic low to the SW, for example above the Paleozoic Platform (blue body) (Fig. 7a). The |∇T| map, in contrast, has highs located roughly along the contact surface (Fig. 7b) and a wide gradient zone extending toward central Europe. One more synthetic model was computed considering the same structural setting but assuming the same magnetization of 2 A m⁻¹ for both the prismatic bodies. Therefore, the produced magnetic field is exclusively generated by the difference in thickness between western and eastern Europe. The results are shown in the Supplement. The magnetic field map at 350 km (Fig. S1a in the Supplement) shows a very low-intensity anomaly, this time related to only the variation from the thin southwestern crust to the thicker one to the northeast. Such a structure is also imaged in the total gradient map (Fig. S1b) by an extended maximum amplitude along the contact line.

Figure 7Contact-like 3-D model of central Europe and related magnetic field (a) and total gradient intensity (b) at 350 km of altitude. This model refers to hypothesis B described in this paper, wherein the main magnetic source is a magnetic contact between the Paleozoic Platform (blue body) and the Precambrian platform (red body).

Download

Therefore, the case of two magnetized crusts differing exclusively in structural features may contribute to the magnetic field but cannot completely explain such a magnetic anomaly. The combination of both structural and magnetic property differences seems to be the best hypothesis to explain the observed magnetic field above Europe.

Intermediate models that consider the magnetic features of both models A and B can also occur. In the next section we will analyze the field and its total gradient intensity by taking into account the results of our synthetic source distributions.

We computed the total gradient intensity maps of Europe through upward continuation of the aeromagnetic dataset for Europe from 5 to 100 km and for the MF7 model data at 350 km of altitude. As discussed previously the simultaneous analysis of the total gradient intensity maps at different altitudes allows for a detailed evaluation of the magnetic properties of the area because it can effectively outline the following: (a) the relative presence or absence of magnetized sources; (b) the location of sources with a strong remanent component of magnetization; and (c) the occurrence of coalescence at large altitudes. The interpretation of the regional magnetic field is indeed improved if the contributions of the shallow sources can be identified in the low-altitude map. We show the maps of the total gradient intensity at 5, 100 and 350 km of altitude in Fig. 8a, b and c, respectively. The lineaments of the main tectonic structures (from Blundell et al., 1992) are shown with white lines.

Figure 8Total gradient intensity maps of the aeromagnetic dataset at 5 km (a) and 100 km of altitude (b); total gradient map of the MF7 magnetic field model at 350 km of altitude (c). White lines represent the main tectonic structures (from Blundell et al., 1992). ABM: Anglo-Brabant Massif; AMA: Adriatic magnetic anomaly; AS: Alpine system; BM: Bohemian Massif; MC: Massif Central; MD: Moldanubian zone; MGCH: Mid-German Crystalline High; MM: Midlands Microcraton; OMB: Osnitsk–Mikashevichi Igneous Belt; PB: Paris Basin; PN: Pannonian Basin.

Download

5.2 Magnetic field at 100 km

At 100 km of altitude (Fig. 8b) we observe mainly long-wavelength anomalies, whereas small-scale anomalies tend to vanish or to coalesce. Here, a strong difference between the EEC and Paleozoic platforms occurs clearly because of the intense |∇T| highs covering most of the EEC and the substantially non-magnetic behavior of the Paleozoic Platform characterized by a few sparse, weak and isolated anomalies. Huge |∇T| highs are located above the TESZ region, which, according to Milano et al. (2016), suggests the TESZ as a magnetic structure involving the whole crust down to the deep lithosphere. This picture seems to favor hypothesis B as the main cause of the CEML. In fact, the |∇T| map clearly shows that the weak and sparse nature of the magnetization in central Europe is not able to characterize the anomaly field at large altitudes through the mechanism of coalescence.

Figure 9Joint interpretation of magnetic and total gradient anomalies in central Europe (a) over the Bohemian Massif (b), the Anglo-Brabant Massif (c) and the Pannonian Basin (d).

Download

However, a coalescence effect can supply an additional contribution to the large-scale magnetic field. Apart from the TESZ, the most noticeable |∇T| high is observed above the Adriatic region, also recognizable in the magnetic field map as the Adriatic magnetic anomaly (AMA). However, in central Europe we can distinguish some slight and isolated |∇T| highs, especially above eastern Germany, the central Czech Republic and southeastern Europe. Rounded and separated highs are also recognized in southern England and Belgium. Comparing the maps of |∇T| at 5 and 100 km, these highs are likely to be produced by coalescence between small-scale anomalies. We indeed observe that most of such |∇T| highs, such as those above PN, BM and ABM, represent the long-wavelength component of sharp and well-defined anomaly patterns that merge with each other at high altitudes, forming a regional-scale magnetic signal. In Fig. 9 we jointly analyze the magnetic and total gradient anomalies in specific areas of central Europe (red boxes in Fig. 9a) in localized regions where reverse magnetization is dominant. In the area of the Bohemian Massif (Fig. 9b) we immediately note that the magnetic field map above BM shows a wide magnetic low (left) in correspondence to the high of the total gradient intensity (right). This feature is interesting because, as discussed above, a reverse magnetization area can be easily identified when the magnetic low of a reverse T anomaly is replaced by a |∇T| high (see Fig. 5). We may so conclude that the joint analysis of T and |∇T| suggests the presence of reversely magnetized sources above BM, even though paleomagnetic analysis did not reveal a significant reverse magnetization here.

In Fig. 9c, we performed the same type of joint analysis in the area of the Brabant Massif, Namur Basin and Ardennes Massif. Magnetic and paleomagnetic studies (e.g., Molina Garza and Zijderveld, 1996) have pointed out Paleozoic carbonate rocks in Belgium characterized by uniform reverse polarity as a consequence of regional remagnetization events that occurred during tectonic and orogenic activity at late Carboniferous times. In the magnetic field map, we observe two intense magnetic lows, with no clear presence of their relative highs, which transform to two highs in the |∇T| map. These results, combined with the paleomagnetic information, allow for an interpretation of these sources as characterized by a strong remanent component of magnetization. Further paleomagnetic studies were carried out in central Europe by Pucher (1994), wherein an analysis of magnetic anomalies in Germany and Belgium pointed out pyrrhotite as the main carrier of remanent magnetization, and a similar magnetic feature was expected in other Paleozoic rocks in central Europe. Other studies have been carried out by Zegers et al. (2003) on Permian remagnetization of Devonian limestones in northern France and Belgium.

The third area considered is the Pannonian Basin (Fig. 9d), which formed contemporary to the Tertiary evolution of the Carpathians and the Eastern Alps, based on SW-to-W-dipping subduction (Bielik et al., 2004). Here an extended magnetic low, visible up to satellite altitudes, lies over Slovakia and Hungary (Fig. 9d, left). Magnetic anomaly modeling was carried out in this area (Taylor et al., 2005; Kis at al., 2011) using magnetic field measurements of the CHAMP satellite. In their model the authors assume a strong remanent magnetization for the source according to exsolution of the hematite–ilmenite minerals found in the upper crust of the Pannonian Basin. Looking at the total gradient intensity map (Fig. 9d, right) we again find a good correlation between the magnetic low anomaly and the high in the |∇T| map.

We conclude that, besides the evidence of a strong contrast in magnetization between SW and NE European regions (hypothesis B), the presence of specific areas where remanent magnetization is dominant represents an additional contribution to the CEML anomaly at 100 km of altitude.

Figure 10Magnetic field (a) and total gradient intensity (b) from a synthetic model of central Europe in agreement with Taylor and Ravat (1995). This model refers to hypothesis A described in this paper. Reverse magnetization is assigned to central European crust and normal magnetization to the Precambrian platform. The magnetic field anomaly is computed at 350 km of altitude (a). The total gradient intensity map (b) shows two highs above both the synthetic bodies.

Download

EMMP high-resolution datasets are available from Getech, UK (info@getech.com), and MF7 model data can be downloaded at http://geomag.org/models/MF7.html (last access: 10 May 2019; Maus et al., 2010).

The supplement related to this article is available online at: https://doi.org/10.5194/se-10-697-2019-supplement.

MM and MF conceived of the presented work and with JDF wrote the paper. MM performed the computations, derived the models and produced the figures. MF and JDF supervised the findings of the work. All authors discussed the results and contributed to the final paper.

The authors declare that they have no conflict of interest.

This paper was edited by Nicolas Gillet and reviewed by Fabio Caratori Tontini and Jörg Ebbing.