Tectonic Rift-Related Manganese Mineralization System and Its Geophysical Signature in the Nanpanjiang Basin

Abstract

The southeastern Yunnan region in the southwestern Nanpanjiang Basin is one of the most important manganese enrichment zones in China. Manganese mineralization is mainly confined to marine mud–sand–carbonate interbeds of the Middle Triassic Ladinian Falang Formation (T₂f), which contains several medium to large deposits such as Dounan, Baixian, and Yanzijiao. However, the geological processes that control manganese mineralization in this region remain insufficiently understood. Understanding the tectonic evolution of the basin is therefore essential to unravel the mechanisms of Middle Triassic metallogenesis. This study investigates how rift-related tectonic activity influences manganese ore formation. This study integrates global gravity and magnetic field models (WGM2012, EMAG2v3), audio-frequency magnetotelluric (AMT) profiles, and regional geological data to investigate ore-controlling structures. A distinct gravity low–magnetic high belt is delineated along the basin axis, indicating lithospheric thinning and enhanced mantle-derived heat flow. Structural interpretation reveals a rift system with a checkerboard pattern formed by intersecting NE-trending major faults and NW-trending secondary faults. Four hydrothermal plume centers are identified at these fault intersections. AMT profiles show that manganese ore bodies correspond to stable low-resistivity zones, suggesting fluid-rich, hydrothermally altered horizons. These findings demonstrate a strong spatial coupling between hydrothermal activity and mineralization. This study provides the first identification of the internal rift architecture within the Nanpanjiang Basin. The basin-scale rift–graben system exerts first-order control on sedimentation and manganese metallogenesis, supporting a trinity model of tectonic control, hydrothermal fluid transport, and sedimentary enrichment. These insights not only improve our understanding of rift-related manganese formation in southeastern Yunnan but also offer a methodological framework applicable to similar rift basins worldwide.

1. Introduction

The Nanpanjiang Basin (also referred to as the Youjiang or Dian–Qian–Gui Basin) developed as an intraplate rift on the Yangtze Block during the Hercynian–Indosinian tectonic cycle, influenced by the breakup of the Paleo-Tethys Ocean. It exhibits typical extensional rift characteristics. Previous studies have shown that mineralization within this rift basin is closely related to the migration and precipitation of hydrothermal fluids under an extensional tectonic regime, with ore-forming fluids concentrating along faults to form both fault-controlled and stratabound deposits [1,2,3]. Located in the southwestern part of the basin, southeastern Yunnan represents the core area of manganese enrichment, with substantial resource potential and promising exploration prospects. Manganese mineralization in this region is primarily hosted in marine mudstone–sandstone–carbonate sequences of the Middle Triassic Falang Formation (T₂f) and is strongly influenced by NE-trending synsedimentary faults and slope-facies deposition during the Ladinian stage. These factors jointly formed a sedimentary–hydrothermal metallogenic system, exemplified by major deposits such as Dounan, Baixian, and Yanzijiao [4,5,6]. However, despite the proposed existence of a rift system in the Nanpanjiang Basin, the detailed spatial configuration and its coupling relationship with the metallogenic system remain unclear. This knowledge gap hampers the development of robust metallogenic prediction models. Moreover, as shallow manganese resources in southeastern Yunnan are progressively depleted, the challenge of deep mineral exploration is becoming increasingly critical. Therefore, clarifying the structural framework and evolutionary processes controlling mineralization is essential for refining regional metallogenic theories and guiding future deep exploration.

Deep density and magnetic anomalies in rift zones reflect fundamental geodynamic processes, including lithospheric thinning, crustal uplift, and mafic magmatism [7,8,9]. These anomalies typically manifest as gravity lows and magnetic highs along rift axes, with steep gravity–magnetic gradients delineating major fault zones [10,11]. Audio-Magnetotellurics (AMT), with its deep penetration, high vertical resolution, and resistance to cultural noise, is increasingly applied to resolve deeply buried ore bodies and complex fault systems. Complementarily, satellite-derived gravity and magnetic datasets offer cost-effective, large-scale coverage. Satellite missions such as GRACE and SWARM have significantly advanced lithosphere-scale studies, as documented in previous research [12,13]. However, their spatial resolution (~150 km) remains insufficient for resolving regional fault networks [14]. Global models such as EGM2008 [15], EMAG2 [16], WMM2015 [17,18], and WDMAM [19,20] have improved resolution to approximately 10 km, yet this remains inadequate for detailed delineation of buried faults and basement structures. Although AMT provides detailed local-scale structural information, its application is often limited by spatial coverage. To address these limitations, this study integrates the high-resolution WGM2012 [21] gravity model (~3.7 km, 2 arc-minute resolution) and the EMAG2-V3 [22] magnetic anomaly model (~3.7 km, 2 arc-minute resolution), combined with structural boundary analysis, to achieve a more precise mapping of the Nanpanjiang rift basin’s geometry. Moreover, AMT surveys in key ore districts validate the interpreted subsurface structures and metallogenic centers. This multi-method approach balances broad spatial coverage with detailed local resolution, providing a robust framework for refining the deep structural controls on manganese mineralization. Such an approach is critical given the depletion of surface manganese deposits and the pressing need for improved exploration strategies.

To clarify the paleotectonic evolution and manganese metallogenesis in southeastern Yunnan, this study integrates WGM2012 gravity data and EMAG2-V3 magnetic anomaly data with edge detection and depth estimation techniques to delineate the structural geometry of the Nanpanjiang rift basin and identify potential hydrothermal vent centers. Audio-frequency magnetotelluric (AMT) validation in key ore districts confirms the presence and metallogenic significance of these centers. The results improve our understanding of basin-scale structural controls on manganese mineralization and offer a geophysical framework for evaluating resource potential in analogous rift-related systems worldwide.

2. Geological Background and Mineralization of the Nanpanjiang Basin

2.1. Regional Geological Setting

The Nanpanjiang Basin lies at the southwestern margin of the South China Block (Figure 1a,b), southwest of the Jiangnan Orogenic Belt. It occupies a triple junction among the Cathaysia Block to the southeast, the Indochina Block to the west, and the Yangtze Block to the north. Its southern boundary comprises a complex tectonic zone formed by subduction and closure of the Paleo-Tethys branch oceans, including (from north to south) the Dian–Qiong, Zhaijiang, Majiang suture zones and the Red River–Ailao fault zone. The northwestern margin is defined by the Shizong–Mile Fault, adjacent to the Kangdian Upland and the Emeishan Large Igneous Province (ELIP). To the northeast, it is bounded by the Ziyun–Nandan–Hechi and Nandan–Du’an–Mashan faults along the Jiangnan Belt. The southeastern margin contacts the Yunkai Terrane [4,5,23,24,25,26,27].

Since the Devonian, the Nanpanjiang Basin has evolved through multiple stages, transitioning from a deep-water basin to a shallow-water platform. It exhibits diverse sedimentary architectures and a relatively complete stratigraphic record, except for Precambrian sequences (Figure 1c). Cambrian–Ordovician strata, dominated by argillaceous dolostones and shales, are widely exposed. Devonian strata are well-preserved, while Carboniferous units mainly comprise light-colored carbonates typical of shallow marine platforms. Permian exposures are sparse. The Triassic presents the most extensive distribution, predominantly north of Mengzi–Yanshan–Guangnan, with scattered occurrences in Dongma and Xichou. It features thick sequences with significant variability in lithofacies. The lower unit consists of fine clastics and marl, with basal conglomerates near paleoshorelines, ranging from 280 to 1800 m thick. The middle unit is subdivided into a lower carbonate with interbedded shale and an upper manganiferous carbonate intercalated with siltstone, 126–3659 m thick. The upper unit comprises clastics interbedded with carbonaceous shale, 215–1736 m thick. The Middle Triassic Falang Formation (T₂f), a key manganiferous horizon, is divided into three members. The lower member (T₂f₁) contains yellow shale, mudstone, slate, with interbeds of siliceous shale, carbonaceous shale, and argillaceous limestone. Manganiferous mudstone, occurring locally, is also present, with thicknesses ranging from 400 to 700 m. The middle member (T₂f₂) mainly comprises argillaceous limestone with limestone interbeds, 400–800 m thick. The upper member (T₂f₃) predominantly consists of variegated shale, tuffaceous phyllite, and slate, with lenticular argillaceous limestone, approximately 1000 m thick.

Four manganese mineralization episodes—Late Devonian, Early Carboniferous, Middle Permian, and Early to Middle Triassic—align with key Paleo-Tethys Ocean geodynamic stages: initial rifting, oceanic spreading, early subduction, and final closure. These tectonic transitions induced marginal subsidence and submarine hydrothermal activity, driving manganese enrichment under restricted marine conditions [5,25,26,28,29]. In the Late Devonian, rift-related basin-platform alternations dominated the Nanpanjiang Basin. Deep-water facies—mainly argillaceous limestone, chert, and shale—were deposited, with initial manganese precipitation along rift margins influenced by subsidence and hydrothermal inputs. During the Early Carboniferous, continued oceanic spreading created NW-trending open carbonate platforms in the Nanpanjiang Basin, localizing manganese mineralization at platform margins and basin centers within mixed carbonate-siliciclastic successions. The Middle Permian Dongwu Orogeny triggered synsedimentary faulting and formed narrow foreslope carbonate belts along platform edges. These faults enhanced hydrothermal circulation and remobilized manganese, concentrating ore along basin-foreslope boundaries. In the Early Triassic, Paleo-Tethys closure isolated platforms flanked by mixed siliciclastic-carbonate foreslopes, sustaining manganese mineralization. The Middle Triassic featured turbidite-dominated basins and extensional settings along the western margin, hosting manganese-bearing units. By the Late Triassic, the Indosinian Orogeny caused intense folding and uplift, closing Paleo-Tethys, terminating marine sedimentation, and ending manganese metallogenesis (Figure 2).

2.2. Deposit Geology

Manganese deposits in the southeastern Nanpanjiang Basin are predominantly located in the Jianshui, Baixian, Dounan, and Heqing areas, dated to the Ladinian stage (Middle Triassic). The ore-hosting Falang Formation consists of marine mudstone, siltstone, sandstone, and interbedded carbonates. Two principal deposit types occur: marine sedimentary and residual weathering manganese deposits. To date, 107 manganese occurrences have been documented, including 19 medium-sized and 63 small deposits, plus 28 mineralized sites. Major deposits such as Dounan and Yanzijiao collectively contain an estimated 13.74 million tonnes of proven manganese resources. The region exhibits considerable metallogenic potential [30].

The Dounan manganese belt is a major high-grade manganese concentration in southeastern Yunnan, encompassing deposits such as Dounan, Laowu, and Dajing, along with numerous mineralized sites (Figure 3a). The Dounan deposit is the largest and richest, characterized by a distinctive Mn-bearing sedimentary assemblage of carbonate and terrigenous clastics, serving as a key stratigraphic marker for manganese metallogenesis in the basin [23]. The Shiwang manganese deposit at Xiaoguanzhai, Qiubei County, typifies a marine sedimentary oxidized manganese deposit (Figure 3b). Its stratiform to stratabound ore bodies occur within the upper Falang Formation (T₃f^b), predominantly hosted by interbedded siltstone and mudstone. The deposit shows strong stratigraphic control and represents a typical manganese occurrence in the Nanpanjiang rift basin.

3. Data and Methods

3.1. Gravity and Magnetic Data

This study employs high-resolution geophysical datasets, the World Gravity Model WGM2012 [31] and the global magnetic anomaly dataset EMAG2v3 [22], as primary geophysical inputs. Gravity anomalies are sourced from the WGM2012 global gravity grid at 2′ × 2′ resolution, developed through international collaboration. Magnetic data are derived from EMAG2v3, referenced to the WGS84 datum, with a 2 arc-minute spatial resolution and a reference elevation of 4 km above the geoid.

All datasets were processed and interpreted qualitatively and quantitatively using Oasis Montaj 8.4 [32] and ArcMap 10.8.

3.2. Data Analysis Methods

To correct magnetic asymmetry and improve structural trend and source interpretation, total magnetic intensity (TMI) data were reduced to the pole (RTP) using the IGRF model (inclination 36.01°N, declination −1.51°E). Gravity data were upward continued by 3 km to suppress high-frequency noise. All subsequent analyses were conducted on the RTP magnetic and upward-continued gravity datasets. Edge detection methods—including horizontal gradient magnitude (HGM), tilt derivative (TDR), and regional–residual separation—were applied to RTP magnetic and gravity data to delineate rift boundaries. Depths were estimated via 3D Euler deconvolution, source parameter imaging (SPI), and spectral analysis. AMT data were inverted using Gauss–Newton least-squares (GNLSQ) and Modified Regularized Inversion (MRI) methods (Figure 4).

3.2.1. Reduction to the Pole (RTP) of Aeromagnetic Data

In polar regions, where the geomagnetic field is nearly vertical, magnetic anomalies directly reflect the surface projection of their sources. In contrast, low- to mid-latitude regions exhibit oblique magnetization due to inclined magnetic fields, causing spatial offsets between anomalies and their sources. Reduction to the pole (RTP) corrects this distortion by reprojecting anomalies as if measured at the magnetic pole, improving spatial alignment between anomalies and causative bodies [33].

$$L(\theta )={\displaystyle \frac{{\left[sin(I)-icos(I)cos(D-\theta )\right]}^2}{\left[si{n}^2(I_a)+co{s}^2(I_a)\cdot co{s}^2(D-\theta )\right]\left[si{n}^2(I)+co{s}^2(I)\cdot c{o}^2(D-\theta )\right]}},if(|I_a|<|I|),I_a=I$$

(1)

Here, I denotes geomagnetic inclination; D geomagnetic declination (azimuth); $I_a$ amplitude correction inclination; and L(θ) the RTP operator at latitude θ. The geomagnetic inclination (36.01°) and declination (−1.51°) were obtained from the 11th-generation International Geomagnetic Reference Field (IGRF-11) model (Finlay et al., 2010) at the study area’s center (23.5°N, 104.5°E) [34].

3.2.2. Upward Continuation Processing

Upward continuation extrapolates potential field data, such as gravity and magnetic anomalies, from the observation plane to a higher synthetic observation surface. This frequency-domain operation is performed via two-dimensional Fourier transform [35].

Mathematically, the upward continuation is expressed as follows:

$$F_{z+h}(k_x,k_y)=F_z(k_x,k_y)\cdot e^{-h\sqrt{k_x^2+k_y^2}}$$

(2)

where $F_z(k_x,k_y)$ represents the Fourier transform of the original field and h denotes the continuation height.

The Bouguer gravity dataset from WGM2012 incorporates terrain corrections based on the 1′ × 1′ ETOPO1 model, which may introduce high-frequency noise [36]. To suppress shallow-source noise and observational errors while enhancing anomalies associated with mid-to-deep crustal structures, a 3 km upward continuation filter was applied. This effectively attenuates short-wavelength components, improving the clarity of geologically significant features. All subsequent gravity data processing was performed on the upward-continued field.

3.2.3. Edge Detection Techniques

Horizontal Gradient Method (HGM)

The horizontal gradient magnitude (HGM) is defined as the vector sum of the horizontal derivatives of total magnetic intensity (TMI). It is widely used to highlight the boundaries of magnetic sources. Local maxima of HGM aligned in consistent directions often mark linear contacts or edges of magnetic bodies [37,38].

If M(x, y) denotes TMI at position (x, y), HGM is calculated as follows:

$$HGM=\sqrt{({\displaystyle \frac{\partial M}{\partial x}}{)}^2+({\displaystyle \frac{\partial M}{\partial y}}{)}^2}$$

(3)

The horizontal derivatives, $\partial M/\partial x$ and $\partial M/\partial y$, represent the east–west and north–south gradients of the aeromagnetic data. Horizontal gradient magnitude (HGM) was computed from both the reduced-to-pole (RTP) magnetic and gravity datasets. HGM peaks were extracted using a 3 × 3 moving window to identify local maxima within the RTP grid.

Tilt Derivative (TDR) Method

The tilt derivative (TDR) is a commonly used edge-enhancement method in gravity and magnetic data interpretation, effective for delineating subsurface structural boundaries. It enhances edge detection by calculating the arctangent of the ratio between the vertical and horizontal gradients of the potential field [39].

The tilt angle (θ) is defined as follows:

$$TDR=\arctan \left({\displaystyle \frac{\frac{\partial F}{\partial z}}{(\frac{\partial F}{\partial x}{)}^2+(\frac{\partial F}{\partial y}{)}^2}}\right)$$

(4)

where F represents the potential field (gravity or magnetic), while $\partial F/\partial x$, $\partial F/\partial y$, and $\partial F/\partial z$ denote its first-order derivatives along the x, y, and z directions, respectively [40].

Tilt derivative (TDR) contour maps are effective in delineating source boundaries: positive values correspond to the interior of the source, negative values indicate the exterior, and the zero-contour marks the structural edge. To enhance structural interpretation, in this study, local maxima of the horizontal gradient along the TDR zero-contour were superimposed on the HGM map to highlight boundary orientations.

Residual/Regional Separation

The residual–regional separation method improves near-surface structural resolution by attenuating long-wavelength signals and emphasizing shallow anomalies. This approach employs radial power spectrum analysis of gravity data, where linear segments correspond to the depths of equivalent density source layers. Each segment represents a subsurface level containing gravity sources, under the assumption that gravity anomalies can be modeled by an equivalent surface distribution of mass at depth.

Mathematically, this filtering is defined as follows:

$$L(k)={\displaystyle \frac{1}{1+{(k/k_c)}^n}}$$

(5)

Here, k is the wavenumber, $k_c$ denotes the cutoff wavenumber, and n is the order of the Butterworth high-pass filter. In this study, Bouguer gravity data were processed using a fourth-order Butterworth high-pass filter with a cutoff wavelength of 333 km to suppress low-frequency signals and enhance high-frequency anomalies.

3.2.4. Depth Estimation

Euler Deconvolution (Euler3D)

Euler deconvolution is a commonly applied inversion method for estimating the location and depth of gravity or magnetic sources [41,42]. It quantitatively relates the observed field and its gradients to the source position, allowing estimation of the depth to causative bodies. The method incorporates the Structural Index (SI), which describes the rate at which field intensity decays with distance from the source and is selected based on the assumed source geometry. An appropriate SI improves the reliability of source localization and depth estimation.

The Euler homogeneity equation for magnetic data is expressed as follows:

$$(x-x_0){\displaystyle \frac{\delta T}{\delta x}}+(y-y_0){\displaystyle \frac{\delta T}{\delta y}}+(z-z_0){\displaystyle \frac{\delta T}{\delta z}}=N(B-T)$$

(6)

where $(x_0,y_0,z_0)$ denote the magnetic source location corresponding to the total field anomaly T observed at (x, y, z), B is the regional magnetic field, and N is the structural index (SI), representing the decay rate of the field gradient. In this study, Euler deconvolution was applied using a 20 × 20 km moving window and an SI of 1.0, with a maximum depth error tolerance of 15%, to delineate subsurface contacts and faults.

Source Parameter Imaging (SPI)

Source Parameter Imaging (SPI), also known as the Local Wavenumber method [43], applicable to two common geological models: two-dimensional inclined thin plates and inclined contact surfaces.

The local wavenumber k of the magnetic field is defined as follows:

$$k=\left({\displaystyle \frac{\frac{{\partial }^{2}F}{\partial x\partial z}\frac{\partial F}{\partial z}+\frac{{\partial }^{2}F}{\partial y\partial z}\frac{\partial F}{\partial y}+\frac{{\partial }^{2}F}{{\partial }^{2}z}\frac{\partial F}{\partial z}}{(\frac{\partial F}{\partial x}{)}^2+(\frac{\partial F}{\partial y}{)}^2+(\frac{\partial F}{\partial z}{)}^2}}\right)$$

(7)

The maximum local wavenumber $k_{\max }$ occurs directly above isolated contact edges and is independent of strike, declination, inclination, dip, and remanent magnetization. The source edge depth can be estimated as the inverse of $k_{\max }$:

$$Dept{h}_{(x=0)}={\displaystyle \frac{1}{k_{\max }}}$$

(8)

The initial step in applying SPI to gridded data involves calculating spatial derivatives at each grid point along the x, y, and z directions.

Spectral Analysis

Spectral analysis-based depth estimation (radial average power spectrum method) models surface magnetic data as the sum of responses from sources at varying depths. Analysis of the power spectrum reveals the spatial distribution and depth variations in magnetic sources [44]. The slope of the power spectrum estimates the average source depth within frequency bands corresponding to regional or residual fields.

The calculation is given by Equation (9):

$$Z(depth)={\displaystyle \frac{-slope}{4\pi }}$$

(9)

3.3. Audio-Frequency Magnetotelluric (AMT) Data Acquisition and Processing

3.3.1. AMT Data Acquisition

Audio-frequency magnetotelluric (AMT) data were collected using the Geode EM3D 3D tensor electromagnetic system. A cross-shaped electrode configuration was employed to acquire four-component tensor data, including two orthogonal horizontal electric field components ($E_x$ and $E_y$) and one horizontal magnetic field component ($H_y$). The acquisition frequency ranged from 0.35 to 10,400 Hz, with a minimum recording time of 30 min per station.

In the Dounan ore district, a 1.5 km survey line was established with a nominal station spacing of 100 m, locally refined to 50 m, yielding a total of 26 sites. The profile traverses the northeastern limb of the Dounan anticline (Figure 3a). In the Shiwang ore district, a 1.45 km profile was acquired with a station spacing of 50 m, locally refined to 30 m, for a total of 46 sites (Figure 3b).

3.3.2. Audio-Frequency Magnetotelluric Data Processing

The Gauss–Newton least-squares (GNLSQ) inversion algorithm is widely used to derive smooth resistivity models from electromagnetic sounding data. It solves a nonlinear least-squares problem with regularization constraints, achieving rapid convergence and high fitting accuracy, making it effective for practical geoelectrical applications [45].

Given M observed apparent resistivity data points $d_1$, $d_2$, … $d_M$ measured at multiple frequencies, each with an associated uncertainty ${\sigma }_j$, synthetic responses are computed using the forward operator $F[m]$. The misfit between observed and modeled data is quantified using a weighted least-squares objective function:

$$X^2={\displaystyle {\sum }_{j=1}^M{(d_j-F_j[m])}^2/{\sigma }_j^2}$$

(10)

where M is the number of observations and ${\sigma }_j$ represents the error of the j-th data point. The goal of the inversion is to identify a model m that minimizes the regularization term $R_1$ or $R_2$, while achieving an acceptable misfit $X^2$ given the dataset d and its associated errors.

The forward problem is solved using the Modified Regularized Inversion (MRI) method, whose solution can be expressed as follows:

$$d_j=F_j[m],j=1,2,\dots M$$

(11)

A dataset containing M observations is denoted as $d\in E^M$, and the model parameters as $m\in E^N$. The forward function corresponding to the j-th datum is $F_j$ and collectively expressed as $d=F[m]$.

The data misfit can be formulated as follows:

$$X^2=\left|\right|Wd-WF[m]|{|}^2,w=diag(1/{\sigma }_1,1/{\sigma }_2,\dots ,1/{\sigma }_M)$$

(12)

4. Results

4.1. Gravity and Magnetic Anomaly Characteristics of the Study Area

The Bouguer gravity anomaly and its 3 km upward continuation (Figure 5a,b) reveal a progressive decrease in gravity values from southeast to northwest (−120.0 to 60.0 mGal), reflecting crustal thickening and lateral variations in lithospheric density. The regional field (Figure 5c) displays an SE–NW gradational pattern of alternating moderate-to-high and low anomalies, indicative of a structurally heterogeneous lithosphere. Residual gravity anomalies (Figure 5d) delineate a pronounced NNW–NW-trending high-gravity belt along the Red River Fault, intersected by a secondary SW–NE-trending anomaly, suggesting structural superposition of multiple fault systems.

Integrated analysis reveals the presence of a consistent NE-trending low gravity anomaly zone extending through Qiubei–Wenshan–Gejiu. This zone likely corresponds to an accumulation of low-density materials, potentially representing an ancient rift trough or a region of crustal thinning. Moreover, NE-aligned linear high gravity anomalies are spatially coincident with the Red River Fault, implying a strong structural control and deep lithospheric influence.

The total magnetic intensity (TMI) and reduced-to-the-pole (RTP) maps (Figure 5e,f) exhibit pronounced E–W magnetic zonation, with intensity increasing markedly from west to east. The western sector is characterized by broad, low-amplitude magnetic anomalies locally overprinted by NE-trending highs, attenuated from north to south. In contrast, the eastern zone features an N–NNE-trending high-magnetic belt, forming a continuous NE-oriented magnetic ridge from Qiubei to Hekou. These features reflect deep-seated magnetic sources shaped by multiphase tectonic processes and structural reactivation.

The Bouguer gravity TDR and HGM anomaly maps (Figure 6a,c) reveal concentric alternating high–low anomaly zones across the study area. A closed NE-trending low TDR anomaly in the center likely reflects an ancient rift or crustal thinning. Linear high anomalies trend NE in the north–central region and NW–NNW in the south. The integration of TDR zero-contours and HGM high-gradient zones (Figure 6e) identifies three dominant structural orientations—NE, NNE, and NW/NNW—indicating a multiphase tectonic framework. The central region shows low fault density and structural conditions favorable for fault-controlled basins.

Aeromagnetic TDR and HGM results (Figure 6b,d) delineate a NNW–NS magnetic structural boundary. Central anomalies form alternating NE–NW bands, with localized distortions near TDR zero-contours. The overlap of HGM highs and TDR zeros (Figure 6f) provides strong evidence for NE, NNE, and NNW trends, suggesting possible crust–mantle boundaries or deep-sourced metallogenic zones associated with Pan-African relict structures.

The 3D Euler deconvolution of integrated gravity and aeromagnetic data (structural index SI = 1; Figure 7) reveals a distinctly multi-directional and polyphase deep structural framework across the study area. Gravity-derived Euler solutions identify E–W trending deep-seated density anomalies (>6 km; Figure 7a), concentrated in the southern and western sectors, likely associated with ductile shear zones or deep crustal interfaces. In contrast, Euler depths derived from aeromagnetic magnetization data delineate N–S to NNW-trending magnetic bodies (>8 km; Figure 7b), primarily distributed in the eastern and northern regions, potentially indicating inherited basement structures or deep magmatic conduits. The spatial alignment between Euler-derived anomalies and mapped structural trends demonstrates strong geometric consistency (Figure 7c,d), suggesting these deep structures play a key role in guiding magmatism and associated metallogenic processes.

SPI inversion of Bouguer gravity data reveals density sources at ~1–4.5 km depth, with banded and blocky patterns centered in the study area’s middle (Figure 8a). Aeromagnetic SPI results indicate deeper magnetic sources, extending to 13 km and covering a broader area (Figure 8b). Gravity Euler solutions show E–W and NE-trending structures controlling anomaly distribution, likely linked to deep shear zones or basement faults (Figure 8c). Aeromagnetic data suggest magnetic bodies align along N–S, NNW, and NE trends, reflecting control by inherited structures or deep magmatic conduits (Figure 8d). Integrated interpretation suggests that the intersection of these trends may mark potential paleo-vent centers.

Radial average power spectrum analysis of combined Bouguer gravity and aeromagnetic total magnetization data indicates multi-scale structural layering in southeastern Yunnan. The gravity power spectrum (Figure 9a) identifies a dual-layer structure: a shallow high-density body representing the activated basement under the Indosinian post-arc extensional regime (slope = −207.25; depth = 14.49 km), and a deeper layer corresponding to the stable Precambrian crystalline basement of the pre-Paleo-Pacific subduction craton (slope = −469.14; depth = 30.32 km). The magnetic power spectrum (Figure 9b) delineates the Yanshanian granite magma chamber roof (slope = −207.25; depth = 16.59 km) and the top boundary of the Proterozoic magnetic basement (slope = −469.74; depth = 30.60 km).

4.2. Audio-Frequency Magnetotelluric (AMT) Data

This study conducted audio-frequency magnetotelluric (AMT) surveys at two manganese mining districts: Dounan and Shiwang. The Dounan profile, approximately 1.5 km in length, traverses the northeastern segment of the Dounan anticline (Figure 3a). The Shiwang profile, 1.45 km long, intersects documented manganese ore bodies, providing key constraints for subsurface resistivity interpretation (Figure 3b).

4.2.1. Dounan Mining District

The NW–SE-oriented profile, approximately 1.5 km in length with an investigation depth of ~1000 m, reveals an overall increase in resistivity with depth. A vertically layered resistivity structure is observed, characterized by an “upper low-resistivity—middle low-resistivity anomaly—lower high-resistivity” pattern (Figure 10). Two stable low-resistivity anomalies (<40 Ω·m) occur below the 0.3–0.7 km and 1.3–1.5 km sections at depths of ~80–300 m, interpreted as manganese-rich ore bodies within the Middle Triassic Falang Formation, fourth member (T₂f⁴⁻⁶). Overlying this is the Upper Triassic Niaoge Formation (T₃n), dominated by limestone with resistivity ranging from 100 to 500 Ω·m, exhibiting electrical heterogeneity due to structural fracturing and karstification. The third member (T₂f₃) of the Middle Triassic Falang Formation, located beneath the ore bodies, is a carbonate unit characterized by a high-resistivity background (300–3000 Ω·m).

Discontinuous low-resistivity anomalies along the profile indicate the presence of NW-trending secondary faults, which likely acted as effective conduits for hydrothermal fluid migration. The structural configuration reveals an asymmetrical rift basin characterized by a central graben flanked by uplifted margins. Stratigraphic architecture is governed by major regional faults and associated extensional structures, which provide both migration pathways and depositional space for mineralizing fluids, facilitating manganese ore accumulation.

4.2.2. Shiwang Mine Area

The profile extends 1450 m along an east–west orientation. Resistivity results reveal a “central low-resistivity trough flanked by higher-resistivity zones” pattern (Figure 11), with resistivity generally increasing with depth, interspersed with localized low-resistivity bodies.

Integrated analysis of geological, geomorphological, and physical property data suggests the following: at ~0.75 km along the profile, a large vertically extensive low-resistivity anomaly (chimney-shaped) extends beyond 1 km depth, likely representing a fracture zone or hydrothermal alteration conduit enriched in conductive fluids, indicative of a deep hydrothermal ore pathway. Two elliptical, shallow, closed low-resistivity anomalies occur at ~0.35–0.5 km and 1.25–1.4 km, corresponding to mineralized layers or altered wall rocks related to manganese mineralization. These anomalies overlie the deeper low-resistivity zone, possibly marking the interface where ascending hydrothermal fluids interact with sedimentary layers to form mineralization. Surrounding high-resistivity areas correlate with dense limestone and dolomite bedrock on the terrace, interpreted as unaltered carbonate basement or residual strata at the profile edges.

5. Discussion

5.1. Rift Graben and Hydrothermal Vent Centers

Bouguer gravity data reveal a NE–SW trending low-gravity zone along the Qiubei–Wenshan segment in southeastern Yunnan. This anomaly persists at depths of 2, 4, 8, and 10 km in Bouguer gravity and vertical second derivative maps but vanishes at 15 km depth, indicating a deep basement depression near Qiubei [46]. This gravity signature typifies a rift zone formed by lithospheric extension, crustal thinning, and thick sediment accumulation. Gravity contours along the rift’s western margin (Mile–Jianshui) trend NEE, while localized NW-trending gravity gradient belts suggest complex basement structures, implying proximity to the Kangdian Rift margin. The NE-trending Nanpanjiang Fault defines the rift’s northwestern boundary [47].

Aeromagnetic total intensity data reveal a pronounced high-magnetic anomaly zone along the Qiubei–Wenshan corridor, aligning with the Nanpanjiang Fault system strike. This anomaly primarily reflects Emeishan mantle plume-related basalts, diabases, and buried acidic intrusions [48], indicating a tectono-magmatic regime controlled by NW–SE major faults with alternating Indosinian mafic dike swarms and Yanshanian granitoid intrusions. This pattern underscores strong coupling between deep heat flow pathways and basin structural activity [29,49,50,51]. The Jianshui–Gejiu segment shows low magnetic anomalies beneath a thick Carboniferous–Permian volcanic sequence up to 3600 m, including lava flows and pyroclastic deposits exceeding 3200 m, evidencing large-scale fissure-style volcanism along coeval faults [52].

Integrating lithofacies, volcanic distribution, and gravity-magnetic anomalies, previous studies confirm that the Nanpanjiang Basin exhibited a rift environment since the Early Devonian [5,25,53]. The development of volcanic rocks and deep-water chert deposits further supports a strong extensional tectonic regime during this period. Geodynamically, the rift system originated from the Early Paleozoic collision between the passive Yangtze Plate margin and the South China–Indochina block during the Caledonian orogeny, forming a rift-type basement. Its evolution was controlled by back-arc extension related to the subduction rollback of the Paleo-Tethys oceanic slab [52,54]. In the Middle to Late Early Devonian, NE–SW intraplate extension triggered the rift’s active phase, accompanied by multiple small-scale magmatic events from the Devonian to Early Permian. In the Late Permian, mantle plume activity along coeval faults induced fissure-style mafic magmatism [28,53,55]. During the Middle Triassic, the combined Indosinian orogeny and collision with the Northern Vietnam Craton shifted the basin into a back-arc foreland setting. By the Late Triassic, tectonic convergence further intensified, transforming the Nanpanjiang Basin into a typical foreland basin, marking the rift’s infill and closure stages [56].

The rift’s western margin is influenced by the Ailaoshan–Red River thrust–fold belt, generating a complex stress regime with NW-directed compression and NNE-trending extension. The resultant rift–graben system displays a characteristic “checkerboard” pattern formed by NE-trending primary normal faults intersected by NW-trending secondary faults (Figure 12a,b). Late-stage tectonics were governed by the India–Eurasia collision, left-lateral strike-slip along the Red River Fault, and South China Block rotation, driving erosion, uplift, and subsidence. This led to the development of the modern NE–SW-oriented half-graben structure with pronounced topographic relief.

Hydrothermal vent activity concentrates along the rift’s eastern margin, primarily at intersections of NE–SW major faults and NW–SE secondary transtensional fractures. These zones align with low gravity anomalies, high magnetic intensities, or steep magnetic gradients, delineating four principal deep magma–hydrothermal vent centers. These centers are marked by thermal anomaly convergence, development of high-permeability fracture networks, and intense ore-forming fluid flow. They represent key loci for deep material release and metal enrichment within the rift hydrothermal system. From north to south, these centers are near the Qiubei Basin, Qiubei County, Dounan Basin, and Gejiu Basin, forming a significant manganese metallogenic belt hosting major deposits such as Dounan, Laowu, Dajing, and Shiwang, along with numerous manganese occurrences [57].

5.2. Rift Grabens and Manganese Mineralization

5.2.1. Tectonic Control on Mineralization

During the Middle Permian, intermediate to acidic magmatism along the Jinshajiang–Ailao Shan suture marked progressive closure of the Paleo-Tethys Ocean by the Late Permian [58]. Collision between the Indosinian and South China blocks induced extensive crustal extension, plate fragmentation, and asthenospheric upwelling [5,26], establishing a deep thermal regime and metallogenic material source. By the Early Triassic, continued northwestward convergence of the Indochina and South China blocks activated tectonics along the Red River fault zone on the Nanpanjiang Basin’s western margin, forming a closed foreland basin system. Manganese mineralization is mainly localized at intersections of NE–SW principal faults and NW–SE transtensional secondary fractures.

The Nanpanjiang rift axial zone developed an NE–SW trending half-graben controlled by regional extension and deep-seated faults (e.g., Wenma–Mingsu faults). This structural trend aligns with gravity lows and gravity–magnetic anomaly gradients, highlighting the dominant influence of NE–SW major faults on rift formation and evolution (Figure 12). Elevated gravity and magnetic anomalies along the eastern margin reflect buried dense, highly magnetic basement bodies acting as rigid structural barriers, constraining eastward rift propagation and defining tectonic stability. Conversely, the western margin features a ring-shaped low gravity–magnetic anomaly zone, likely associated with a deep basement depression [46], providing a critical geophysical setting for manganese reductive sedimentation and supergene enrichment via surface runoff.

5.2.2. Hydrothermal Ore Transport

Previous studies reveal a strong spatiotemporal link between magmatism and manganese mineralization, with magmatic–hydrothermal fluids critical for metal supply [28]. The Southeast Yunnan manganese source area hosts volcanic–terrigenous mixed sediments. Hydrothermal fluids ascend along major rift faults and migrate laterally as bottom currents, facilitating fluid redistribution on the seafloor and early manganese ore reorganization and enrichment [5,26]. Carbon isotope (δ¹³C_PDB) values of carbonate manganese ores and Mn-bearing carbonates from Dounan range from −6.1‰ to −7.9‰, with a minimum of −11.6‰ at Baixian, reflecting carbon from seawater bicarbonate, volcanic input, and organic matter decomposition. This indicates mixing of organic-rich, low-oxygen seawater with magmatic hydrothermal fluids altered carbon isotopes, driving Mn precipitation and enrichment on fault-controlled basin slopes. Rare earth element (REE) data further support hydrothermal input: Dounan ores show micro-nodular La/Ce ratios of 0.40–0.86 (mean 0.55), intermediate between hydrothermal (~2.8) and hydrogenous (~0.25) signatures [23]. Y/Ho ratios range from 29.6 to 42.9, exceeding continental crust (25–28) and typical hydrothermal fluids (26–27) [59,60,61], indicating a mixed metallogenic source comprising terrigenous weathering, magmatic degassing, and seawater components.

Integrated gravity and magnetic data indicate that during mineralization, a mid-crustal low-velocity layer regulated anomalous heat flow, mobilizing Mn-rich oxides within the weathering crust, which migrated upward along fault systems. The southern segment of the NE–SW magnetic anomaly ridge (Qiubei–Wenshan) corresponds to Indosinian basic dike swarms, reflecting coupled volcanic–tectonic control over hydrothermal fluid migration. Reduced-to-pole (RTP) aeromagnetic inversion data reveal >8 km-deep fault systems that serve as conduits for hydrothermal fluids sourced from the mantle and mid–lower crust. These structures define the spatial framework of a plume–sediment hydrothermal manganese system. The beaded magnetic anomalies correspond to magnetite mineralization along the tops and margins of Yanshanian intrusions, supporting fault-controlled hydrothermal circulation at intrusion contacts.

5.2.3. Sedimentary Enrichment

During the early mineralization stage, the Nanpanjiang rift basin experienced intense extensional subsidence during the Ladinian, forming a semi-enclosed, weakly redox-stratified shallow marine environment [62]. Mn²⁺ migrated in seawater as ions or colloids, complexing with HCO₃⁻ to form MnHCO₃⁺ or oxidizing to unstable Mn³⁺, which further formed Mn(III)-ligand complexes [63] or hydroxides such as β-MnOOH and γ-MnOOH [64]. These intermediates further transformed into primary manganese oxides such as birnessite, facilitated by dissolved silicate (H₄SiO₄) [23]. Subsequent marine transgression caused upward migration of the redox boundary, subjecting Mn oxides to reductive dissolution and partial remobilization as Mn²⁺, which reprecipitated as rhodochrosite. This produced mixed assemblages of birnessite and rhodochrosite, marking a transitional diagenetic stage [23,65]. Concurrently, high-energy shallow marine conditions deposited abundant detrital material, enhancing Mn adsorption, microbial colonization, and ooid/micronodule nucleation. Biogenic activity further promoted Mn²⁺ oxidation and the co-enrichment of Fe, Co, Ni, and REEs [66]. In the late mineralization stage, NE–SW-trending major faults and associated fractures served as pathways for Mn-rich hydrothermal fluids. These fluids interacted with organic matter or Fe²⁺ under shallow, reducing conditions, resulting in the precipitation of Mn-calcite veins and metasomatic alteration in earlier-formed manganese ore [23,28].

Geophysical anomalies provide essential deep structural constraints on manganese mineralization. In southeastern Yunnan, the sedimentary residual zone of the Wenshan fault-controlled basin coincides with gravity–magnetic lows, indicating that mineralization is spatially inherited from early basin architecture. In the northern rift, a SW–NE-trending high magnetic anomaly belt corresponds to dike swarms and lithological boundaries between mafic and intermediate–felsic intrusions. These low-permeability intrusive bodies impede vertical hydrothermal flow, redirecting fluids laterally along boundary faults and promoting mineral precipitation in overlying permeable sandstone units. Reduced-to-pole (RTP) airborne magnetic inversion delineates shallow normal faults (3–6 km depth) along basin margins, which control secondary depressions (e.g., the Dounan Basin) and localized subsidence, creating structurally favorable environments for hydrothermal metal accumulation [67].

5.3. Case Study of Typical Deposits

The southeastern Yunnan region, located within the Nanpanjiang Basin, constitutes a key segment of the Middle Triassic NE–NEE-trending rift-related manganese metallogenic belt. This belt, extending from Baixian through Laowu to Qiubei, is structurally controlled by regional faults that regulate basin extension, subsidence, sedimentary facies transitions, and provenance input. Stratigraphy shows a west-to-east shift from carbonate-dominated units to clastic sequences with increasing terrigenous input, reflecting the interplay between rift-margin tectonics and depositional environments. Manganese orebodies are primarily hosted in regional fold structures, indicating a structurally controlled mineralization style [23,68].

The Dounan manganese belt, located within the NE-trending Dounan basin, is structurally controlled by the Mingsu and Wenma faults. Bouguer gravity data delineate a NE-oriented low-gravity anomaly zone coinciding with high-gradient airborne magnetic anomalies, suggesting lithospheric thinning and deep heat flow upwelling conducive to hydrothermal plume-related deposition [23]. Ore bodies are concentrated at the intersection of NE-trending gravity gradients and regional magnetic highs, indicating focused fluid migration along syndepositional faults and subsequent precipitation of rhodochrosite and pisolitic Mn carbonates. Integrated tectonic and geophysical analyses place the Dounan deposit on the northwestern margin of Rift Plume Center 2 (TCPL2), reflecting a coupled system of deep thermal input and syn-rift sedimentation.

The Shiwang deposit, situated on the western limb of the Shiwang syncline, lies within the transitional zone between the Yangtze quasi-platform and the South China Fold Belt. It is structurally controlled by NE-trending transpressional faults and subordinate EW faults. Mineralization occurs in the Upper Triassic Falang Formation (T₃f^b), hosted by siltstones and mudstones with pronounced stratiform control. Basin evolution proceeded through non-compensational deposition, intensified tectono-sedimentary transition, and late-stage deep-water sedimentation under “starved basin” conditions, collectively providing heat, metal sources, and reducing environments favorable for Mn mineralization. Regional airborne magnetic data delineate NE-trending magnetic highs corresponding to contacts between deep-seated mafic dikes and granitic intrusions, acting as persistent thermal sources. NE–SW faults serve as primary pathways for hydrothermal fluid migration and ore element enrichment. Integrated geological and geophysical evidence identifies the Shiwang deposit as a plume–sediment metallogenic center within the rift system, designated Rift Plume Center 3 (TCPL3).

Audio-magnetotelluric (AMT) deep sounding profiles provide additional support for the proposed mineralization model. Resistivity inversion sections identify persistent low-resistivity zones associated with fractures, fissures, and karst systems. These conductive belts show strong spatial correlation with ore-controlling faults and fold boundaries, indicating clear structural–electrical coupling. Manganese orebodies are characterized by continuous low-resistivity anomalies, significantly lower than the surrounding bedrock, and are interpreted as hydrothermal alteration halos or mineralized cemented zones. These observations reinforce the role of hydrothermal plume centers as key contributors to manganese mineralization.

In summary, integrated gravity, airborne magnetic, and AMT data, in conjunction with regional geological and mineralogical evidence, demonstrate that manganese deposits in southeastern Yunnan are spatially associated with the intersections of NE-trending gravity gradients and high magnetic anomaly belts. These zones coincide with deep-seated fault systems, reflecting pronounced lithospheric thinning and focused hydrothermal activity. The mineralization is controlled by rift-boundary tectonics and plume–sediment interactions, illustrating a typical tectono–hydrothermal–sedimentary metallogenic model.

5.4. The Directions for Further Research

Global potential field datasets like WGM2012 and EMAG2v3 are valuable for regional geological studies, but their inherent limitations must be acknowledged. WGM2012 (~2 km resolution) effectively reveals large-scale lithospheric structures but lacks the detail needed for local mineral exploration. Similarly, EMAG2v3 (2 arc-minute resolution) struggles to detect subtle magnetic anomalies, weak faults, and minor lithological variations. Both datasets are derived from globally heterogeneous sources, leading to noise, interpolation artifacts, and inconsistent data quality—especially in gravity data—which can obscure shallow structural features [69].

To address these issues, this study applied upward continuation, tilt angle and horizontal gradient filters, and regional–residual separation to suppress long-wavelength trends and enhance shallow structural signals. These methods significantly improved the clarity and continuity of structural features in the study area [70].

Nevertheless, uncertainties remain in complex zones. Thus, interpretations based on global models should be considered preliminary. For accurate identification of ore-controlling faults and mineralization pathways, high-resolution airborne or ground geophysical surveys are necessary. Future work should focus on refining these models, particularly in the structurally complex areas of the Nanpanjiang Basin.

6. Conclusions

NE-trending Bouguer gravity lows, coinciding with magnetic highs or gradients, delineate a NE-trending rift system in the Nanpanjiang Basin. This system is defined by intersecting NE-trending master faults and NW-trending secondary faults, forming a characteristic “checkerboard” extensional pattern that channels deep mantle heat and supports plume development.

Integrated gravity–magnetic data identify four plume centers at fault intersections. These centers represent structurally focused zones for hydrothermal activity and manganese enrichment.

AMT inversion results further support this interpretation, revealing low-resistivity anomalies along major faults within ore-bearing strata—interpreted as hydrothermal fluid pathways.

We propose a trinity metallogenic model involving rift-related structural architecture, deep fluid transport through faults, and sedimentary enrichment. This model offers a geophysical basis for understanding deep manganese mineralization processes and provides a methodological reference for future exploration in the region.