Joint modeling gravity, geoidal and geothermal of the Lithosphere in Sergipano Belt and tectonic implications, NE Brazil.

In this research paper were applied methods which integrate geophysical and petrological data to the crustal and lithospheric structure in the Sergipano Belt, in the Southern Borborema Province, NE Brazil. These methods allowed us to unveil the 10 main crustal features and their tectonic relationships with the crust-mantle interface (Moho), as well as the lithosphere-asthenosphere boundary (LAB) of this region. The crust structure was constrained by available geological, gravimetric and seismological data, and chemical composition for mantle bodies. Gravity, geoid and heat flow data integrated with existent geological and geophysical data provides to understand the evolution of the Southern Borborema Province. The modeling method was based on local isostasy in which the first phase of the work consisted on the filtering of geoid anomaly and 15 topography data using low-pass in the harmonic and frequency domains, respectively, to estimate the thickness of the crust and the lithosphere, and the second phase is characterized by geothermal modeling. The crustal structure was constrained by geological information, thermo-physical parameters, and gravity and magnetic data inversion results, while the lithospheric mantle was constrained by chemical composition of xenoliths. The results show that maximum depths of Moho and thermal lithosphere (LAB) along the modeled profile are 34 - 42 km and 194-202 km observed, and temperatures are recorded with 20 an average of 600° C in the Moho temperature at Sergipano belt. The results presented suggest a thinning lithospheric mantle below the Sergipano belt, with values varying between 168 to 180 km, excepting in Girau do Ponciano whose is 194 km and a moderate lithospheric thickening below the Rio Coruripe, and thicker lithosphere at Pernambuco-Alagoas terrains reached 202 km. approximate geometries for the LAB. The SEB domain and the Sergipe belt are characterized by LAB ranging 169 to 180 km, an abrupt change is observed in Girau do Ponciano. These results are in agreement obtained by gravity data. Based on lateral variation of the chemical composition of the lithospheric mantle, was possible to obtain the best fit of the seismic velocities waves P and S, and the density 375 distribution. Our results show that the PEAL domain has a different composition that extends to the Marancó, Poço-Redondo, Canindé and Rio Coruripe domains, from the rest of the region due to collision with the São Francisco Cráton. The crustal thickness results show that the Salvador-Esplanada-Boquím domain (SEB), Sergipano belt and Alagoas Sub-basin are characterized by thinning crust with values varying 32-36 km, thicker crust for Pernambuco-Alagoas terrain varying 38-40 km, in agreement with seismic and gravimetric data. Results of lithosphere-asthenosphere boundary (LAB) in all two 380 profiles show maximum values in Pernambuco-Alagoas domain reaching ~201 km. profiles show approximate geometries for the LAB. The SEB domain and the Sergipano belt are characterized by LAB ranging 169 to 180 km. These results are in agreement obtained from gravity data. The chemical mantle composition and its lateral variation show that the chemical composition of PEAL extends to the first three domains of the Sergipano Belt, and their respective adjustments in density


Introduction
The understanding the tectonic evolution of Precambrian Terrans is complex due to their tectonic dynamics of their structures, either by erosion or overlapping of tectonic events. This difficulty resides in the reconstruction of the evolutionary process that generated these terrains, as well as in the characterization of its lithosphere, its dynamics, heat production on the 30 earth's surface and the location of the suture zones between the different blocks. Geothermal models have gained prominence these days, as they allow the interpretation of the current lithospheric characteristics of these terrains and for being an important tool to answer several questions arising from tectonic dynamism, in space and time, thus helping to provide the geological history in a region.
Gravity and geoid data, as well as magnetic, seismological and heat flow data have been recently successfully used to better 35 understand and delineate geological terrains (eg, Kaban et al., 2014;Carballo et al., 2015;Afonso et al., 2016;Qashqaiet al., 2016;Sampaio et al., 2017;Abedi et al., 2018;Souza et al., 2019;Brown et al., 2020, Kumar et al., 2020. In this paper, we use gravimetric, geoid and topographic data associated with thermal properties, density distribution and seismological velocities to characterize the thermal structure of the lithosphere of the Southern Borborema Province, NE Brazil. Therefore, this modeling allowed us to understand the geodynamic processes involved in the formation of the crust, crust-mantle 40 interface and lithosphere-asthenosphere interface of the region. The conception Southern Borborema Province (SBP) is related to the role of neoproterozoic orogen between South America and West Africa during the Gondwana supercontinent development, his conFiguretion was due to the convergence of the Amazonian Cráton, São Francisco Congo and São Luís -West Africa during the orogeny Brito Neves et al., 2000). This is an area that attracts a lot of attention due to its complex geotectonic evolution, the collision generated a 1000 45 km long continental orogen on the north margin of the São Francisco Craton composed of the Rio Preto, Riacho do Pontal and Sergipano belts (Caxito et al., 2016 and references), in South America, extending across northwestern Africa, in the Oubanguides Orogen, on the northern margin of the Congo Craton (Trompette, 1997).
In the last few years, several geological and geophysical studies have been developed in this area in order to unveil the crustal structure as Argollo et al. (2011) who studied the crustal model and heat flow in the Estância, Vaza-Barris and 50 Macururé domains, based on the measurement of density, conductivity and estimation of the volumetric rate of heat production; Dutra et al. (2018), who developed a crust structure model and geothermal flow map for the adjacent region to the Sergipe-Alagoas basin; Araújo (2019) who studied the size of sources in the transition zone between of the Segipano belt and the Pernambuco-Alagoas domain; Dutra et al. (2017) which estimated the basement thickness and lateral variation of the density contrast in the Estância, Canudos-Vaza-Barris, Macururé domains and Pernambuco-Alagoas Terrain; Fianco et. 55 (2019) and Dutra et al. (2019) which estimated the moho thickness for Pernambuco-Alagoas terrain and also for the Sergipano belt; and Oliveira (2009)  in most regions of the Sergipano belt, obtained through integrated 3D inversion of data from the Bouguer and geoid anomalies. On the other hand, studies about lithospheric structure, LAB depths, thermal modelling, mantle density and seismic velocity distribution still need to be done. 60 The modeling of geophysical and petrological data, associated with its relationship with plate tectonics in the different geological domains, allowed to determine the structure in muti-layers of the Southern Borborema Province. This study integrates gravity anomaly (free-air and Bouguer anomalies), geoid anomaly, topography and heat flux, thermal conductivity and radiogenic heat rate data to produce a geothermal, density distribution and velocity model of the seismic waves (S and P) for the crustal and lithosphere-asthenosphere structure up to 400 km depth. 65

Geological Settings
The SBP comprises one of the main tectonic segments of the Borborema Province. Many researchers infer the subdivision of the Borborema Province into three subprovinces: Northern, Central or Transversal and Southern , Pereira et al., 2020. The Borborema Province is resulted from continental blocks that collided during the amalgamation of Western Gondwana at the end of the Neoproterozoic. Therefore, the SBP results on the collision between 70 the Pernambuco-Alagoas Massif and São Francisco Cráton to the north, this region is divided by the Sergipano belt (SB), Riacho do Pontal (RPD), Rio Preto and Pernambuco-Alagoas domains which show the lateral continuity and related geological features Brito Neves et al., 2000, Oliveira et al., 2010Santos et al., 2000).

Figure 1
The Pernambuco-Alagoas domain is limited to the north with the transversal subprovince by the Pernambuco lineament and 75 to the south with the Sergipano belt by shear zones. The Pernambuco-Alagoas Domain is constituted with two large metamorphic complexes, the Belém complex of São Francisco represented essentially by granitic and granodioritic orthogaisse, and Cabrobró Complex, a metavolcanosedimentary unit composed of shales, gneiss, quartzite, marbles, calssilicics and amphibolites . These complexes present a transpressive tectonics showing regions of intense deformation due to the Cariris Velho event (Santos, 1995). In this tectonic domain there are neoproterozoic plutonic 80 intrusions which can be grouped into pre and post-collisional, these granitic intrusions were classified into five batholiths: Buique-Paulo Afonso, Águas Bela-Canindé, Maribondo-Correntes, Ipojucas-Atalaia and Jaboatao-Garanhuns.
The Sergipano belt comprises a sequence of metamorphosed supercustral rocks with characteristics of the passive continental margin and foreland basin (Oliveira et al., 2015). The belt is divisible from north to south into the Canindé, Poço-Redondo-Marancó, Rio Coruripe, Macururé, Vaza-Barris, and Estância geotectonics domains ( fig. 1) (Oliveira et al., 85 2010(Oliveira et al., 85 , 2017Davison e Santos 1989;Dutra et al., 2019). The Marancó domain comprises on metasedimentary pelitic to psamitic rocks and intercalations of basalt, andesite, gabbro and amphibolite rocks. Peridotites and gabbros with varying degrees of serpentinization occur mainly as lenses in metasedimentary rocks or as intrusions to the south; they can be slices of lithospheric mantle under the orogen or fragments of ophiolite (Silva Filho, 2006); The Poço-Redondo subdomain is a https://doi.org/10.5194/se-2021-81 Preprint. Discussion started: 30 July 2021 c Author(s) 2021. CC BY 4.0 License. migmatitic gneiss complex dominated by granodioritic-tonalitic rocks that represent the base of the Marancó subdomain; 90 The Canindé domain is between the Pernambuco-Alagoas domain and the Maraconó unit of the Poço-Redondo domain, morever, it is limited by the Macururé shear zone. It consists of metavolcanic-sedimentary rocks, calcissilicate rocks . In these three domains (Marancó-Poço Redondo and Canindé) are the allochthonous accreted during the Neoproterozoic (Oliveira et al., 2017).
The Estância domain is one of the first domains of the Sergipano belt, from south to north is the Estância, which is basically 95 constituted by sub-horizontal sedimentary rocks, deformed and undeformed (Oliveira et al., 2017;Argollo et al., 2011) . The Vaza-Barris domain is the most deformed when compared to the Estância, it contains several grouped formations in the Miaba, Simão Dias and Vaza-Barris groups (Oliveira, 2006;Silva and McClay, 199)). In the typical locality (Fazenda Capitão), the lower Miaba group comprises a basal quartzite unit (Itabaiana), by dolomite interspersed with stromatolite and limestone (Jacoca Fm). The basal quartzites of the Miaba Group rest unconformable in the gneisses and migmatites of the 100 basement of the Itabaiana and Simão Dias domes; The Macururé domain is basically composed of amphibolite facies, metaturbidites containing garnets, schist mica with smaller intercalations of quartzite, marble and meta-volcanic rocks (Oliveira et al., 2006); The Rio Coruripe domain is located between the Macururé domain in the southeast and Pernambuco-Alagoas in the northeast, moreover, it is limited by the beautiful Monte Jeremoabo shear zone. It is constituted by gneissic metamorphic rocks and migmatite with migmatitic structures. Its basement is essentially formed by amphibolitic and granulitic rocks 105 (Sampaio, 2019;. The basement of the Sergipano belt is represented by the occurrence of gneissic domes of Girau do Ponciano, Simão Dias and Itabaiana, which are involved in the episode of regional metamorphisms and tectonic deformation. These deformations were studied by Oliveira et al. (2006); Oliveira et al. (2010) and Oliveira et al. (2015), which the structural evolution of the belt was synthesized in four deformation phases. The first phase, called remaining structures from the pre-Brazilian 110 deformation event, the second is characterized by nappes and pushes with south convergence, the third associated with the transpressive regime and the fourth expressed by the ductile and brittle character.
The other domain included in our study area is Salvador-Esplanada-Boquim, which belongs to the São Francisco Cráton and is represented by high-grade metamorphic rocks. The São Francisco Cráton represents a tectonic entity of the Brasilian cycle and comprises most of Bahia State and extends to the Minas Gerais, Sergipe, Pernambuco and Goiás. The basement was 115 consolidated at the end of the Transamazonic cycle. Its coverage comprises two complexes (De Almeida, 2017), the Archean and Paleoproterozoic terrains that constitute the basement of the São Francisco Cráton in the Bahia State, these can be grouped into four important Archean crustal segments, the Gavião, Serrinha, Jequié and Itabuna-Salvador-Curaçá Blocks, (Souza, 2013 ). The São Francisco Cráton is represented by high-grade metamorphic rocks -gneissic, migmatitic, granulitic and granitoid rocks, in addition to having some dykes of correlated age. The Cráton is truncated by an aborted N-S oriented 120 rift in which protoliths from the Espinhaço (Mesoproterozoic) and São Francisco (Neoproterozoic) Supergroups have deposited.

Material and Methods
We used orbital geophysical data, elevation (topography), gravity anomaly (Bouguer and Free-air), geoid anomaly, and surface heat flow. The elevation data were extracted from ETOPO1, a global relief model, available on the National 125 Geospatial Intelligence Agency -Earth Gravitational Model (NGA, 2008) page, with 1 × 1 min arc resolution, which integrates topography and ocean bathymetry. Figure 2 show the elevation map, the Pernambuco-Alagoas domain is characterized by high elevation values and on the Sergipano belt the map show low elevation.

Figure 2
The gravity anomalies are of data base from International Center for Global Gravity Field Models (ICGEM) and from 130 International Gravimetric Bureau (BGI). The data base ICGEM used the European Improved Gravity model of the Earth by Earth Gravitational Model geopotential model -EGM2008 (Pavlis et al., 2012), (BGI: https: //bgi.obs-mip.fr/). These ICGEM and BGI models are widely used to studies related to Earth internal structure, as well as in lithosphere modeling.
In Figure 3a the Bouguer anomaly map the positive anomalies has amplitudes range from 12 to 60 mGal related to the rocks 135 of the basement with with higher density values, and negatives anomalies with amplitudes ranging −130 to −20 mGal related sedimentary basins (Tucano, Jatobá and Sergipe-Alagoas). The Free-Air was used to interpret anomalies and to model the lithosphere. Figure 3b show the free-air anomaly map and it can be seen that positive anomalies ranging from 22 to 80 mGal on Salvador-Esplanada-Boquim domain and on Sergipano belt (in the Estância, Vaza-Barris) and part of the Macururé domain. Positive anomalies extend to the north of the study area, covering the Rio Coruripe and Pernambuco-140 Alagoas domains, with amplitude 10 to 60 mGal.
Geoid anomaly data were extracted from International Center for Global Gravity Field Models (ICGEM: http://icgem.gfzpotsdam.de) using the GECO model (Gilardoni et al., 2016), this complete model has a spherical harmonic degree and order 2190. Figure 3c show the residual geoid anomaly map (harmonics coefficients up to degree and order 10 removed) and it is observed that maximum anomalies are recorded in the northern region of this study area with amplitude ranging -6 to -3 m, 145 while the southern region is characterized by low amplitude anomalies.
The surface heat flow data were obtained from terrestrial campaigns of the Geoterm project (IF-UFBA), and from National Gravimetry Network through the ANP database (Petrobrás and other oil companies) and used in data modeling, the values are in topography map in Figure 2.

Processing and Lithosphere Modelling
To develop this study, we first searched for a database that allowed us to get the proposed objectives, then, we filtered the topography data by applying low-pass filter in the frequency domain to remove high-frequency (low wavelength) effects associated to the sources. To apply this filter properly, we first developed a power spectrum analyze that allowed us to split the sources into two wavelengths: low wavelength than 66.66km and long wavelength than 66.66 km, which low wavelength 155 is associated with high-frequency while long wavelength is to low-frequency. We also removed the spherical harmonic coefficients up to 10 on the geoid anomaly data to get the residual geoid anomaly, by applying also low-pass filter but in the harmonic domain. These two filtered data were used to estimate automatically the crustal depths (Moho) and lithosphereasthenosphere boundary (LAB) using method that integrates geoid anomaly and topography associate to thermal analyze developed by Fullea et al. (2007). 160 This method is based on local isostatic compensation principle, and considers four layers: crust, sea water, lithosphere and asthenosphere. The density distribution in the lithosphere is dependent on temperature and varies linearly with depth, according to the equation: (1) Where is the linear thermal coefficient expansion, a is the temperature at the lithosphere-asthenosphere limit and 165 is the temperature at depth z in the lithosphere.
The geoid anomaly is calculated selecting the reference column, the one that allow obtaining the best fit and thicknesses. In this work the reference lithospheric column was selected according as described by the method (for details, Fullea et al., 2007). The table 1 show the physical input parameter used in our calculations. We applied improved LitMod2D_2.0 code (Kumar et al., 2020) for lithospheric modeling, which is interactive software that integrates geophysical and petrological data 170 in order to get the lithospherics models up to the 400 km depth. The improved LitMod2D_2.0 is based on the Afonso et al., 2008 code, that allows to determine 2D models, temperature, density and seismic waves velocities (P and S) distributions, in a consistent thermodynamic-geophysical environment, fitting also the geophysical data (Bouguer and free-air anomalies, geoid anomaly, heat flow and topography data). The physical parameters used in our modelling area in table 1.

Table 1 175
The model consists on set of crust and mantle bodies, each one characterized by its physical properties. The crust is characterized by thermo-physical parameters such as density, conductivity and radiogenic heat, while the mantle bodies are temperature in the lithosphere is calculated assuming stationary regime and solving the heat transport equation, and considering fixed boundary conditions at the surface (0 o C) and at the LAB (1320 o C). The density distribution in the model 180 is calculated using an iterative scheme that allows incorporating the effect of pressure, temperature and composition (Kumar et al., 2020;Afonso et al., 2008). Once the physical properties are determined for each mantle composition, gravity, geoid, elevation, surface heat flow, and P and S seismic velocities are computed and compared with the observed values. Gravity and geoid are calculated by using simple algorithms to each element of the mesh. Elevation is calculated assuming local isostasy referenced to a mid-ocean ridge column with the compensation level at 410-km depth (Jiménez-Munt et al., 2019). 185 The data processing workflow is represented by the Figure 4.

Results
In geophysical direct modelling are supposed an existence of an initial model composed of crustal bodies with physical parameters each one and mantle bodies characterized by chemical composition. The model is modified as necessary to get be 190 best fit of the observable data along the transects. In this section we present our modeling results, crustal and lithosphere structure, temperature and density distribution, as well as seismic wave's distribution.

Moho and LAB depths
The Moho and LAB depths has been estimated according to described on the section 3. The Figure  In relation to the crustal structure, the initial crustal geometry along the transects has been constrained by geological information extracted from different studies (e.g Brito Oliveira et al., 2010;Santos et al., 2000;Barbosa, 2012;Silva and McClay, 1995), that describe the tectonics unities and lithologies, and from seismic, gravimetric and magnetic studies (e.g Araújo, 2019;Fianco et al., 2019;Dutra et al., 2019;Sampaio, 2019). We also used the 205 seismological data from Geoterm project. The first transect, A-A', is oriented from the SW-NE, parallel to the strike of the

Mantle chemical composition
The mantle structure along of the profiles is constrained by the chemical composition and the LAB depth values. The chemical compositions, due to its influence on the density and on the seismic velocity waves, are important for our modelling as well as the LAB depth geometry. In this study we were used two chemical compositions (mantle 1 and mantle 2) as shown in table 2. The initial geometry of the LAB along the transects is inferred from the calculated values in section 215 4.1 through topography and geoid anomaly data associated to the thermal analyze, in a 1D approach.
c In the transects we used two chemical composition (table 2). The mantle 1 has been extracted from Griffin et al. (2009) based on the age of the last thectono-thermal event that occurred in the study area, which is the Cráton do São Francisco collision with Pernambuco-Alagoas domain. Geochronological information indicates that this collision occurred in the 220 neoproterozoic, between 650-530 Ma (Lima et al., 2017). We used this composition to constrain the lithospheric mantle

Crustal and Lithospheric Structure
The results of the crustal models obtained in the modeling along the transects A-A' and B-B' are shown in Figure 6. The lithospheric models were elaborated using multilayer geometry, and correspond to the upper crust, middle crust, lower crust,      . 10a). In addition, the S velocities decrease with depth at the LAB and then increase again to the 400km depth. Laterally, S waves low density are found also in PEAL (4.53 ) domain, 280 increasing to SEB domain and Sergipano belt (4.62 ). It is important to notice that the seismic velocities increase in the opposite direction to the increase in the thickness of the lithospheric mantle ( fig. 10b).  11a and 12a). In addition, it was possible to plote also the density and velocity (P and S waves) distribution curves for each selected point using the density and wave velocities models, and it's important to notice that the density distribution and P waves increase from lithospheric mantle up to 400 km depth, while the 290 S wave, decrease with depth from the lithospheric mantle to the LAB and then increase to 400 km depth (figure 11b and 12b).

Discussion 295
The present work aimed to study the lithospheric structure, providing temperature distribution models, density and seismic velocities distribution, representative for the Borborema Southern Province. The initial models were constrained using geological information, gravimetric and seismological data at some points. Based on the analysis of available potential data and modeling methods, it was possible to automatically determine the crustal and lithospheric thickness, using elevation and geoid anomaly coupled with temperature and density distribution (Fullea et al., 2007 method), and we also obtain a more 300 accurate 2D model of temperature, density distribution and seismic waves velocities to the lithosphere. Thermophysical parameter and geochemical composition information was coupled as a priori information in the models. We applied the improved LitMod2D_2.0 code (Kumar et al., 2020) to provide the models, and the geophysical data we used are topography, gravity anomalies (bouguer and free-air), and geoid anomaly and heat flow data.

305
The depths for the Rio Coruripe and Pernambuco-Alagoas sources obtained by Araújo (2018)

Tectonic Implications 335
Gravity, geoid and heat flow data interpretation integrated with existent geological and geophysical data provides to understand the evolution of the Southern Borborema Province. Distinctive gravity features suggest that the lithotectonic associations in the area correlate with different fault-bounded terranes that underwent subsequent amalgamation. These features include contrasting gravity and geoidal signature between terranes. Its origin is interpreted as a result of the oblique collision between Pernambuco-Alagoas domain with the São Francisco-Congo Cráton during the Brasilian/Pan-African 340 orogeny. The structural evolution of the Southern Borborema Province is summarized in four deformation phases and the evolution has a better fit with a model that resembles the one proposed by Davison and Santos (1989). The Sergipano Belt was formed through the collage of distinct lithostratigraphic domains. Main shear zones that divide these domains would be thrust zones that juxtaposed geological units of different crustal levels. This is corroborated by Santos et al. (2014),  and our modeling. In this hypothesis, the SFC limit is represented by de Shear zone São Miguel do Aleixo as 345 proposed by Oliveira and Medeiros (2018). The Macururé Domain is an allochthonous terrane accreted to the São Francisco paleoplate margin and the final suture is represented by the collision of the Marancó-Poço Redondo arc (Oliveira et al., 2010) and the closure of an ocean that is represented by the region dominated by high positive gravity (20 to 40 mGal) and geoidal (5.5 to 6.5 m) values located between PEAL and SB. Presents trough gravity relief, and comprises parts of Rio of terranes during the Neoproterozoic Brito Neves et al., 2000). This proposal has been reinforced by recent studies in the Northern and Transversal domains of the province such as those of Santos et al. (2018) and Pereira et al. 355 (2020). Thus, evidence points to an accretionary origin for the Sergipano Belt and highlights the importance of accretion tectonics in the evolution of the Borborema Province and the amalgamation of Western Gondwana.

Conclusion
In this work, we provide lithospheric model for Southern Borborema Province were developed and presented along two transects, by integrating geological, geophysical and petrological data. The observables geophysical data were calculated and 360 compared to the observed along the profiles. The models shown Moho depths in agreement obtained to the other authors using different geophysical data, showing that the crust at SEB domain and Sergipano belt is thinner to the PEAL domain.
The transition zone occurs between the Macururé and Rio Coruripe domains, as can be seen in the moho and LAB maps, also the LAB geometries in the models reveal a thinning a lithospheric mantle in the Sergipano Belt domains and a thickening in the Pernambuco-Alagoas domain, in accordance with proposed by some researchers. The mantle chemical 365 composition has a very important for lithosphere modeling to fit the geophysical data. Its lateral variations, significantly influence to the density distribution and to the seismic velocity waves, it is noticed in our second model in figure

515
Physical Parameter Numerical Value