A method and system for reconstructing p-t-t trajectory of metamorphic rock based on multi-source data fusion

Abstract

The application discloses a metamorphic rock P-T-t trajectory reconstruction method and system based on multi-source data fusion, relates to the field of metamorphic rock P-T-t trajectory reconstruction, and comprises the following steps: acquiring electronic probe micro-area mapping data and internal standard sample data of a metamorphic rock sample, and performing multi-stage data correction; performing mineral phase identification and chemical component quantitative interpretation, generating a mineral classification mapping, an element concentration distribution diagram and a whole rock major component data set; performing thermodynamic phase equilibrium simulation, generating a P-T apparent profile diagram, and extracting a P-T trajectory; performing micro-area in-situ chronology analysis on selected minerals in the mineral classification mapping to obtain chronology data; coupling the P-T trajectory and the chronology data in time and space, constructing a P-T-t three-dimensional trajectory model of the metamorphic rock, and outputting a visualized reconstruction report. The application realizes full-process automation and standardization of the metamorphic rock P-T-t trajectory from data acquisition and processing to model reconstruction, and improves the precision, efficiency and reliability of trajectory analysis.

Description

Technical Field

[0001] This invention relates to the field of metamorphic rock PTt trajectory reconstruction, and more specifically to a method and system for metamorphic rock PTt trajectory reconstruction based on multi-source data fusion. Background Technology

[0002] Pressure-temperature-time (PTt) trajectory reconstruction of metamorphic rocks is a core technical means to understand the history of crustal evolution. Its research results are widely used in fields such as the analysis of orogenic belt tectonic processes, the diagnosis of mineralization environment of ore deposits, and the inversion of deep Earth processes.

[0003] Traditional metamorphic rock PTT studies typically rely on two technical approaches. One is the traditional thermobarometry method based on single-mineral chemical composition. Although this method is widely used, it has significant limitations: its applicability is highly dependent on specific mineral assemblages, and in many rocks that have undergone deep melting, it is mostly a product of retrograde metamorphism, making it unable to effectively capture peak metamorphic conditions. At the same time, this method is highly dependent on manually selecting mineral pairs and manually inputting composition data, which is highly subjective and easily leads to a step-by-step amplification of errors due to misinterpretation of mineral balance or inappropriate selection of activity models, ultimately resulting in systematic biases in geological interpretation.

[0004] Another approach combines whole-rock composition and thermodynamic phase equilibrium simulation, using calculated PT profiles to define metamorphic conditions, which is theoretically more accurate. However, this approach still faces significant challenges in practice: the electron probe microanalysis it relies on cannot fully represent the prevalent compositional zoning and heterogeneity within minerals, resulting in unavoidable statistical errors; although electron probe surface scanning technology can qualitatively display compositional distribution, traditional images cannot be directly interpreted quantitatively, and compositional differences between minerals are prone to interference, causing subtle zoning information to be obscured or ignored. Furthermore, thermodynamic simulation software is complex to operate, and the quality and reliability of its results are highly dependent on the operator's professional knowledge and experience, making it difficult to standardize the process and objectify the results.

[0005] What's even more challenging is that both of these traditional technical approaches share a common core flaw: the key steps of mineral chemical composition analysis, whole-rock composition acquisition, temperature and pressure calculation, and geochronological determination are usually separated and conducted independently. This fragmentation of the data chain not only increases the complexity of the work but also causes information distortion and accuracy loss during data transmission and integration, making it difficult for the final PTT trajectory to truly and continuously reflect the evolutionary history of the rock.

[0006] Therefore, how to design a metamorphic rock PTt trajectory reconstruction method based on multi-source data fusion that can integrate multi-source data, has a high degree of automation, minimizes human intervention and experience dependence, and achieves accurate and efficient reconstruction of metamorphic rock PTt trajectories is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0007] In view of this, the present invention provides a method for reconstructing the Pt trajectory of metamorphic rocks based on multi-source data fusion, aiming to solve the key technical problems existing in the current method for reconstructing the Pt trajectory of metamorphic rocks, such as strong subjectivity, cumbersome process, data fragmentation in different technical links, large error and poor repeatability caused by reliance on human experience, so as to obtain more accurate and reliable information on the metamorphic evolution history of rocks.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] In a first aspect, the present invention provides a method for reconstructing the PTT trajectory of metamorphic rocks based on multi-source data fusion, comprising the following steps:

[0010] S1. Obtain electron probe microanalysis data and internal standard data of metamorphic rock samples, and perform multi-level data correction;

[0011] S2. Mineral phase identification and quantitative interpretation of chemical composition are performed on the corrected micro-area map data to generate mineral classification map, element concentration distribution map and whole rock major component dataset;

[0012] S3. Based on the whole-rock major component dataset and mineral classification map, perform thermodynamic phase equilibrium simulation to generate PT profile; and based on the element concentration distribution map and multiphase equilibrium constraints, extract the PT trajectory from the PT profile.

[0013] S4. Perform micro-area in-situ geochronological analysis on the selected minerals in the mineral classification map to obtain geochronological data;

[0014] S5. Couple the PT trajectory with the geochronological data in time and space to construct a three-dimensional PT trajectory model of the metamorphic rock and output a visual reconstruction report.

[0015] Preferably, S1 includes:

[0016] Multi-element X-ray intensity spectra and BSE or SEI images for mineral phase identification and localization were acquired simultaneously using an electron probe microanalyzer.

[0017] High-precision single-point analysis was performed on 10-20 representative sites in each mineral phase, and the oxide weight percentage data was exported after ZAF correction as internal standard data.

[0018] Based on the internal standard data, the multi-element X-ray intensity spectrum is sequentially corrected for dead time, topography, time drift, mixed pixel, and standard position to obtain the corrected micro-area spectrum data.

[0019] Preferably, S2 includes:

[0020] S21. A multi-scale convolutional neural network based on an attention mechanism is used to identify mineral phases in the corrected micro-area map data, generating a mineral classification map that reflects the distribution of each mineral phase.

[0021] S22. Based on internal standard sample data, the X-ray intensity in the micro-area spectrum is quantitatively converted into oxide concentration using an empirical correction algorithm to generate an elemental concentration distribution map.

[0022] S23. Integrate the mineral classification map and element concentration distribution map, and calculate and output the whole rock principal component dataset at the micro-area scale after weighted averaging and density correction.

[0023] Preferably, S21 includes:

[0024] The encoder extracts features from the input corrected micro-area map data through multiple convolutional layers, and outputs a feature map containing mineral features at different scales.

[0025] The feature map is processed by the channel and spatial attention module. By calculating the channel and spatial weights, a weighted feature map is generated to highlight key mineral areas.

[0026] The decoder upsamples and fuses the weighted feature map through a transposed convolutional layer, and outputs a mineral category probability map with the same size as the input map.

[0027] By combining the classification output layer, pixel-by-pixel argmax operation is performed on the mineral category probability map to generate a mineral classification map.

[0028] Preferably, in step S22, the empirical α-factor correction algorithm is expressed as:

[0029]

[0030] Among them, C i k represents the weight percentage concentration of the oxide of element i. i I represents the correction factor. i,sample I is represented by the characteristic X-ray intensity of element i in the sample. i,standard It is represented by the characteristic X-ray intensity of element i in the standard sample.

[0031] Preferably, S23 includes:

[0032] The mineral classification map and the elemental concentration distribution map are spatially registered at the pixel level to establish the correspondence between mineral phase categories and chemical compositions;

[0033] Based on the mineral classification map, the percentage of area of ​​each mineral phase within the micro-region is calculated;

[0034] The chemical composition data of all pixels in each mineral phase region are extracted from the element concentration distribution map, and the average value is calculated as the representative chemical composition of the mineral phase.

[0035] The principal components of the whole rock are calculated using an area-weighted average algorithm. The results are then volume-corrected by combining the density values ​​of each mineral phase, and a dataset of principal components of the whole rock at the micro-area scale is output.

[0036] Preferably, S3 includes:

[0037] S31. Based on the whole rock major component dataset, select a thermodynamic database and define a solid solution model, and set the pressure-temperature simulation range and step size;

[0038] S32. Using the Gibbs free energy minimization algorithm, the stable mineral assemblage at each PT point is initially solved by linear programming, and then the composition and proportion of each phase are refined by nonlinear iterative optimization algorithm to generate PT view profile.

[0039] S33. Based on the mineral composition zonation structure or composition gradient information provided by the element concentration distribution map, and combined with the multiphase equilibrium reaction relationship, track the stability domain boundary changes of a specific mineral assemblage in the PT view profile to extract the metamorphic PT trajectory.

[0040] Preferably, S4 includes:

[0041] S41. Identify the zircon and monazite dating mineral regions in the mineral classification map;

[0042] S42. Micro-area in-situ ablation is performed using a laser ablation system with a spot diameter of 10-30 μm, and instrument drift is corrected based on a combination of external and internal standards.

[0043] S43. Collect U, Th, and Pb isotope signals, calculate U-Pb age using the isotope dilution algorithm, and obtain chronological data.

[0044] Preferably, in step S5, constructing the PTt three-dimensional trajectory model of the metamorphic rock includes:

[0045] Based on the chronological data, an absolute time scale is provided for the PT trajectory. A continuous functional relationship of temperature-pressure-time is established through an interpolation algorithm, and the entire path evolution process of rock metamorphism is reconstructed in a three-dimensional coordinate system.

[0046] Secondly, the present invention provides a metamorphic rock PTt trajectory reconstruction system based on multi-source data fusion, comprising:

[0047] Data acquisition and correction module: used to acquire electron probe microanalytical data and internal standard data of metamorphic rock samples, and to perform multi-level data correction;

[0048] The identification and quantitative interpretation module is used to identify mineral phases and quantitatively interpret chemical components in the corrected micro-area map data, generating mineral classification maps, elemental concentration distribution maps, and whole-rock major component datasets.

[0049] PT trajectory extraction module: used to perform thermodynamic phase equilibrium simulation based on the whole rock major component dataset and mineral classification map to generate PT view profile; and to extract PT trajectory from the PT view profile based on the element concentration distribution map and multiphase equilibrium constraints;

[0050] Chronology module: used to perform micro-area in-situ chronological analysis on selected minerals in the mineral classification map to obtain chronological data;

[0051] PTt Coupled Modeling Module: This module is used to spatiotemporally couple the PT trajectory with geochronological data to construct a three-dimensional PTt trajectory model of metamorphic rocks and output a visual reconstruction report.

[0052] As can be seen from the above technical solution, compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0053] 1) This method integrates electron probe microanalysis, computer vision recognition, thermodynamic simulation and micro-area geochronology to construct a complete strategy from data acquisition, intelligent mineral identification, composition interpretation, phase equilibrium simulation to spatiotemporal coupling. This reduces the subjective errors and operational inconsistencies introduced by human intervention and experience dependence in traditional methods, and improves the reliability and repeatability of data results.

[0054] 2) Based on the attention mechanism, multi-scale convolutional neural network and empirical α factor correction algorithm, accurate mineral phase identification and quantitative interpretation of chemical composition of micro-area maps are achieved. It can effectively capture micro-area composition variation information such as mineral zoning and reaction edges, and obtain high-reliability micro-area scale whole-rock composition data through weighted fusion and density correction, laying a reliable data foundation for subsequent thermodynamic simulation.

[0055] 3) Deep fusion and three-dimensional dynamic visualization of PTT information were achieved. By coupling high-precision PT trajectory with absolute chronology data in three-dimensional spatiotemporal dimensions, a dynamic PTT model that can reflect the metamorphic evolution process of rocks was constructed, and an integrated visualization report was generated, thereby revealing the time series of metamorphic events, changes in temperature and pressure conditions, and tectonic evolution process more intuitively and comprehensively. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0057] Figure 1 A flowchart of a method for reconstructing the PTT trajectory of metamorphic rocks based on multi-source data fusion is provided in an embodiment of the present invention.

[0058] Figure 2 This is a schematic diagram of the mineral phase identification and quantitative interpretation process of chemical composition provided in an embodiment of the present invention;

[0059] Figure 3 This is a schematic diagram of the PT trajectory extraction process provided in an embodiment of the present invention;

[0060] Figure 4 A PT cross-sectional view provided for an embodiment of the present invention;

[0061] Figure 5 This is a schematic diagram of the micro-area in-situ geochronological analysis process provided in an embodiment of the present invention;

[0062] Figure 6 This is a structural framework diagram of a metamorphic rock PTt trajectory reconstruction system based on multi-source data fusion, provided for an embodiment of the present invention. Detailed Implementation

[0063] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0064] Example 1;

[0065] like Figure 1 As shown, this embodiment provides a method for reconstructing the PTT trajectory of metamorphic rocks based on multi-source data fusion, including the following steps:

[0066] S1. Obtain electron probe microanalysis data and internal standard data of metamorphic rock samples, and perform multi-level data correction;

[0067] S2. Mineral phase identification and quantitative interpretation of chemical composition are performed on the corrected micro-area map data to generate mineral classification map, element concentration distribution map and whole rock major component dataset;

[0068] S3. Based on the whole-rock major component dataset and mineral classification map, perform thermodynamic phase equilibrium simulation to generate PT profile; and based on the element concentration distribution map and multiphase equilibrium constraints, extract the PT trajectory from the PT profile.

[0069] S4. Perform micro-area in-situ geochronological analysis on the selected minerals in the mineral classification map to obtain geochronological data;

[0070] S5. Couple the PT trajectory with the geochronological data in time and space to construct a three-dimensional PT trajectory model of the metamorphic rock and output a visual reconstruction report.

[0071] This method effectively overcomes the shortcomings of traditional methods, such as strong subjectivity, fragmented process, and significant error propagation, by integrating electron probe microanalysis, deep learning-based intelligent mineral identification, thermodynamic phase equilibrium simulation, and micro-area in-situ geochronology. It achieves quantitative, accurate, and three-dimensional dynamic reconstruction of the PTT trajectory of metamorphic rocks from microscopic mineral composition to macroscopic tectonic evolution history.

[0072] The following provides a further detailed explanation of each step in the above method;

[0073] In this embodiment, S1, electron probe microanalysis data and internal standard data of metamorphic rock samples are acquired, and multi-level data correction is performed; including:

[0074] Multi-element X-ray intensity spectra and BSE or SEI images for mineral phase identification and localization are acquired simultaneously using an electron probe microanalysis system. Specifically, the acquisition of multi-element X-ray intensity spectra must ensure sufficient statistical counts for each element, including major elements such as manganese (Mn), sodium (Na), calcium (Ca), potassium (K), iron (Fe), magnesium (Mg), aluminum (Al), silicon (Si), and titanium (Ti). The peak counts of the major elements must be no less than 10. 4 To reduce statistical errors; BSE or SEI images are used to identify major minerals such as garnet, sillimanite, biotite, muscovite, plagioclase, potassium feldspar, and quartz. They should maintain a strict spatial correspondence with the X-ray spectrum and have a consistent pixel resolution, usually above 1024×1024, to ensure the accuracy of subsequent registration and analysis.

[0075] High-precision single-point analysis was performed on 10-20 representative sites in each mineral phase. After ZAF correction, the oxide weight percentage data was exported as internal standard data. The ZAF correction was a combination of atomic number correction, absorption correction and fluorescence correction. The selection of internal standards should cover all major mineral phases and compositional variation ranges. For example, in garnet, points should be systematically selected from the core to the edge to establish a reliable intensity-concentration conversion relationship.

[0076] Based on the internal standard data, the multi-element X-ray intensity map is sequentially corrected by dead time, topography, time drift, mixed pixel, and standard position to obtain the corrected micro-area map data.

[0077] The dead time correction uses the following formula:

[0078]

[0079] Among them, I i This represents the measurement count rate of element i. τ represents the count rate after correction for element i, and τ represents the detector dead time.

[0080] Specifically, in dead time correction, the detector dead time τ needs to be calibrated in advance using standard samples, typically 100–200 ns. After calibration, it can effectively eliminate signal omissions at high count rates. Topographic correction uses surface morphology data retrieved from BSE images to perform cosine correction on the X-ray emission angle, which is especially crucial for inclined surfaces or rough areas. Time drift correction is for long-term testing, requiring the insertion of standard samples every 4–6 hours to monitor intensity changes, and correction is performed using linear or quadratic function fitting to ensure data temporal stability. Mixed pixel correction uses image morphology methods to identify mixed regions at phase boundaries and replaces them with interpolated neighboring pure-phase pixels to ensure the purity of compositional data. Standard sample position correction calculates sub-pixel-level offset vectors by cross-correlating the coordinates of standard sample points with the positions of spectral pixels and applies spatial transformation to eliminate systematic errors caused by mechanical displacement.

[0081] like Figure 2 As shown, mineral facies identification and quantitative interpretation of chemical composition are performed on the corrected micro-area spectral data to generate mineral classification maps, elemental concentration distribution maps, and whole-rock major component datasets; including:

[0082] S21. A multi-scale convolutional neural network based on an attention mechanism is used to identify mineral phases in the corrected micro-area map data, generating a mineral classification map that reflects the distribution of each mineral phase.

[0083] S22. Based on internal standard sample data, the X-ray intensity in the micro-area spectrum is quantitatively converted into oxide concentration using an empirical correction algorithm to generate an elemental concentration distribution map.

[0084] S23. Integrate the mineral classification map and element concentration distribution map, and calculate and output the whole rock principal component dataset at the micro-area scale after weighted averaging and density correction.

[0085] Furthermore, S21 includes: the encoder extracts features from the input corrected micro-area map data through multi-level convolutional layers, and outputs a feature map containing mineral features at different scales; the feature map is processed by a channel and spatial attention module, and a weighted feature map is generated by calculating channel and spatial weights to highlight key mineral regions; the decoder upsamples and fuses the weighted feature map through a transposed convolutional layer, and outputs a mineral category probability map with the same size as the input map; combined with the classification output layer, a pixel-by-pixel argmax operation is performed on the mineral category probability map to generate a mineral classification map;

[0086] This section employs an encoder-decoder network based on an attention mechanism. Through multi-scale feature extraction and channel and spatial attention weighting, it enhances the identification capability of key mineral regions. After upsampling and feature fusion, it generates pixel-level classification results and finally outputs a high-precision mineral classification map.

[0087] Furthermore, the empirical α-factor correction algorithm in S22 is expressed as:

[0088]

[0089] Among them, C i k represents the weight percentage concentration of the oxide of element i. i I represents the correction factor. i,sample I is represented by the characteristic X-ray intensity of element i in the sample. i,standard The characteristic X-ray intensity of element i in the standard sample is represented by k; the empirical α-factor correction coefficient k is used at this location. i The coefficient of determination R is obtained by linear regression of internal standard data. 2 Typically, the value needs to be greater than 0.99 to ensure the linear reliability of the intensity-concentration conversion; for overlapping peaks or interfering elements, peak stripping algorithms or multivariate correction models should be used for deconvolution.

[0090] Furthermore, S23 includes:

[0091] The mineral classification map and the elemental concentration distribution map are spatially registered at the pixel level to establish the correspondence between mineral phase categories and chemical compositions. The pixel-level spatial registration adopts mutual information maximization or feature point matching algorithm to ensure that the mineral phase boundaries and composition gradients are strictly aligned.

[0092] Based on the mineral classification map, the percentage of area of ​​each mineral phase within the micro-region is calculated;

[0093] The chemical composition data of all pixels in each mineral phase region are extracted from the element concentration distribution map, and the average value is calculated as the representative chemical composition of the mineral phase.

[0094] The principal components of the whole rock are calculated using an area-weighted average algorithm, and the results are then volume-corrected by combining the density values ​​of each mineral phase, resulting in a micro-scale dataset of the principal components of the whole rock.

[0095] The area-weighted average algorithm is expressed as follows:

[0096]

[0097] in, This represents the calculated concentration of the p-th oxide in the whole rock. Let A be the average concentration of the p-th oxide in the q-th mineral phase. q This represents the area percentage of the q-th mineral phase, and n represents the total number of mineral phases.

[0098] The volume correction is expressed as:

[0099]

[0100] in, This represents the concentration of the p-th oxide in the whole rock after volume correction, and is the density of the q-th mineral phase.

[0101] This step, through the deep integration of computer vision and chemometrics methods, completes the entire process from intelligently identifying mineral phases and accurately quantifying chemical composition from calibrated micro-area spectral data to efficiently calculating the whole-rock composition. This process not only achieves automated and high-precision segmentation of mineral boundaries, but also obtains reliable compositional information with spatial resolution down to the micrometer level through multi-source data registration and fusion, providing a complete input data foundation for subsequent thermodynamic simulations and improving the objectivity and repeatability of the entire process.

[0102] like Figure 3 As shown, based on the whole-rock major component dataset and mineral classification map, thermodynamic phase equilibrium simulation is performed to generate a PT (potentially terminating) profile; and based on the elemental concentration distribution map and multiphase equilibrium constraints, the PT trajectory is extracted from the PT profile; including:

[0103] S31. Based on the whole-rock major component dataset, select a thermodynamic database and define a solid solution model, and set the pressure-temperature simulation range and step size; the thermodynamic database should be selected according to the rock type, and the solid solution model should cover key minerals;

[0104] Specifically, based on the whole-rock major composition dataset, the MnNCKFMASHTO system database was selected, and specific solid solution models of minerals such as garnet, melt, muscovite, staurolite, biotite, cordierite, ilmenite, and feldspar were activated to construct a thermodynamic system that closely approximates real geological conditions. Subsequently, the simulation ranges for system pressure (P) and temperature (T) were set to 4-15 kbar and 650-900℃, respectively.

[0105] S32. The Gibbs free energy minimization algorithm is used to initially solve for the stable mineral assemblage at each PT point through linear programming. Then, a nonlinear iterative optimization algorithm is used to refine the phase composition and proportion, generating the PT view profile. Here, Gibbs free energy minimization uses linear programming to find the initial solution, and nonlinear optimization uses Newton's iteration or simulated annealing algorithm to refine the phase composition. The initial temperature for simulated annealing is set to 1000K, the cooling rate is 0.95, and the number of iterations is >10. 4 To ensure convergence to the global optimum;

[0106] S33. Based on the mineral composition zonation structure or composition gradient information provided by the element concentration distribution map, and combined with the multiphase equilibrium reaction relationship, track the stability domain boundary changes of a specific mineral assemblage in the PT view profile and extract the metamorphic PT trajectory.

[0107] like Figure 4 As shown, the stable domain of the peak mineral assemblage Grt (garnet)-Ky (kyanite)-Bt (biotite)-Kfs (potassium feldspar)-Pl (plagioclase)-Melt (melt)-Ilm (ilmenite)-Qz (quartz) corresponds to a high-pressure, high-temperature environment. This stability is further demonstrated by precisely tracing the grossular (X) rings within the garnet composition. Grs ) and pyrope (X) Prp The intersection of contour lines, combined with X in biotite Mg The contour lines and the variation trends of anorthosite (An) and sodium feldspar (Ab) components in plagioclase can jointly define the peak metamorphic conditions as being around 11 kbar and 810 °C, reflecting that the metamorphic rock sample had undergone metamorphic processes on a middle and lower crustal scale.

[0108] As temperature and pressure decrease, the mineral assemblage transforms into Grt (garnet)-Sill (sillimanite)-Bt (biotite)-Kfs (potassium feldspar)-Pl (plagioclase)-Melt (melt)-Ilm (ilmenite)-Qz (quartz), indicating the beginning of the retrograde metamorphic stage, showing garnet spessartine (X) Sps The contour lines extend towards the low-pressure direction, combined with biotite XMg The decrease in the value and the reduction in An component in plagioclase, these changes in composition gradients together constrain the PT conditions of approximately 7.8 kbar and 760 °C during the retrograde metamorphic stage; this trajectory reveals that the metamorphic rock sample underwent a near-isothermal decompression retrograde metamorphic process after the peak period, reflecting the dynamic background of tectonic uplift and crustal exhumation.

[0109] In this step, a PT profile of the study area was constructed using the Gibbs free energy minimization algorithm and multiphase equilibrium constraints. Furthermore, by combining spatial variation information such as mineral composition zoning, a continuous PT trajectory reflecting the metamorphic evolution history was extracted. This achieved a leap from static composition data to dynamic temperature and pressure condition inversion, providing a reliable model basis for the thermodynamic reconstruction of rock metamorphic processes.

[0110] like Figure 5 As shown, micro-area in-situ geochronological analysis was performed on selected minerals in the mineral classification map to obtain geochronological data; including:

[0111] S41. Identify zircon and monazite dating mineral regions in the mineral classification map; the identification of dating minerals should be combined with the mineral classification map and BSE / CL image, and regions without cracks and inclusions are preferred.

[0112] S42. Micro-area in-situ ablation is performed using a laser ablation system with a spot diameter of 10-30 μm, and instrument drift is corrected based on a combination of external and internal standards. Laser ablation requires optimization of energy density and repetition frequency; the internal standard is optional. 43 Ca or 49 For Ti, the external standard adopts international standards, such as 91500, GJ-1, and a standard sample is inserted for calibration every 10 points.

[0113] S43. Collect U, Th, and Pb isotope signals, calculate U-Pb age using the isotope dilution algorithm, and obtain chronological data. Isotope data need to be corrected for ordinary lead and interference subtraction. Age calculation uses Isoplot or a similar algorithm. The weighted average age needs to be reported with MSWD and 95% confidence interval.

[0114] In this step, the dating mineral regions such as zircon and monazite are accurately located by integrating mineral classification maps and BSE / CL images. Micro-area in-situ sampling is performed using a laser ablation system with optimized parameters, and internal and external standard corrections are combined to eliminate instrument drift and signal interference. Finally, U-Pb age is calculated by isotope dilution method, and after ordinary lead correction, interference subtraction and statistical verification, high-precision and high-reliability geochronological data are obtained, providing absolute time constraints for PTt trajectories.

[0115] In this embodiment, S5, the PT trajectory is spatiotemporally coupled with geochronological data to construct a three-dimensional PT trajectory model of metamorphic rocks and output a visual reconstruction report.

[0116] The construction of the PTt three-dimensional trajectory model of metamorphic rocks includes: providing an absolute time scale for the PT trajectory based on the geochronological data, establishing a continuous functional relationship of temperature-pressure-time through an interpolation algorithm, and reconstructing the entire path evolution process of rock metamorphism in a three-dimensional coordinate system.

[0117] Furthermore, a visual reconstruction report is output; the report integrates and displays PT view profile, PTt three-dimensional trajectory curve, mineral composition distribution heat map and age data in the form of charts, and includes textual explanations of key metamorphic stages, temperature and pressure conditions and duration.

[0118] This embodiment provides a method for reconstructing the PTt trajectory of metamorphic rocks based on multi-source data fusion. It effectively integrates multi-source heterogeneous data and multidisciplinary strategies, breaks through the technical bottlenecks in traditional reconstruction, and achieves high-precision quantitative characterization of the PTt trajectory of metamorphism. Its reconstruction results show significant advantages in the spatiotemporal continuity of temperature and pressure paths and the analysis of dynamic processes, providing a reliable data model and empirical basis for deepening the study of the coupling relationship between metamorphism and tectonic evolution.

[0119] Example 2;

[0120] like Figure 6 As shown, this embodiment provides a metamorphic rock PTt trajectory reconstruction system based on multi-source data fusion, including:

[0121] Data acquisition and correction module: used to acquire electron probe microanalytical data and internal standard data of metamorphic rock samples, and to perform multi-level data correction;

[0122] The identification and quantitative interpretation module is used to identify mineral phases and quantitatively interpret chemical components in the corrected micro-area map data, generating mineral classification maps, elemental concentration distribution maps, and whole-rock major component datasets.

[0123] PT trajectory extraction module: used to perform thermodynamic phase equilibrium simulation based on the whole rock major component dataset and mineral classification map to generate PT view profile; and to extract PT trajectory from the PT view profile based on the element concentration distribution map and multiphase equilibrium constraints;

[0124] Chronology module: used to perform micro-area in-situ chronological analysis on selected minerals in the mineral classification map to obtain chronological data;

[0125] PTt Coupled Modeling Module: This module is used to spatiotemporally couple the PT trajectory with geochronological data to construct a three-dimensional PTt trajectory model of metamorphic rocks and output a visual reconstruction report.

[0126] This system, through multi-source data fusion and multi-technical collaboration, has built an integrated platform that ranges from micro-area data acquisition, intelligent mineral identification, thermodynamic simulation to geochronological analysis. It has achieved automated reconstruction and three-dimensional visualization of metamorphic rock PTt trajectories, significantly improving the accuracy of data processing and the reliability of geological interpretation.

[0127] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0128] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for reconstructing the PTT trajectory of metamorphic rocks based on multi-source data fusion, characterized in that, Includes the following steps: S1. Obtain electron probe microanalysis data and internal standard data of metamorphic rock samples, and perform multi-level data correction; S2. Perform mineral facies identification and quantitative chemical composition interpretation on the corrected micro-area map data to generate mineral classification maps, elemental concentration distribution maps, and whole-rock major component datasets; including: S21. A multi-scale convolutional neural network based on an attention mechanism is used to identify mineral phases in the corrected micro-area map data, generating a mineral classification map reflecting the distribution of each mineral phase; S22. Based on internal standard sample data, the X-ray intensity in the micro-area map is quantitatively converted into oxide concentration using an empirical correction algorithm, generating an elemental concentration distribution map; S23. The mineral classification map and elemental concentration distribution map are fused, and after weighted averaging and density correction, a whole-rock principal component dataset at the micro-area scale is calculated and output; S3. Based on the whole-rock major component dataset, perform thermodynamic phase equilibrium simulation to generate a PT profile; and extract the PT trajectory from the PT profile based on the elemental concentration distribution map and multiphase equilibrium constraints; including: S31. Based on the whole-rock major component dataset, select a thermodynamic database and define a solid solution model, setting the pressure-temperature simulation range and step size; S32. Use the Gibbs free energy minimization algorithm to initially solve the stable mineral assemblage at each PT point through linear programming, and then refine the composition and proportion of each phase through a nonlinear iterative optimization algorithm to generate a PT profile; the nonlinear iterative optimization algorithm uses a simulated annealing algorithm; S33. Based on the mineral composition zoning structure or composition gradient information provided by the elemental concentration distribution map, combined with the multiphase equilibrium reaction relationship, track the stability domain boundary changes of specific mineral assemblages in the PT profile to extract the metamorphic PT trajectory; S4. Perform micro-area in-situ geochronological analysis on the selected minerals in the mineral classification map to obtain geochronological data; S5. Couple the PT trajectory with the geochronological data in time and space to construct a three-dimensional PT trajectory model of the metamorphic rock and output a visual reconstruction report.

2. The method for reconstructing the PTT trajectory of metamorphic rocks based on multi-source data fusion according to claim 1, characterized in that, S1 includes: Multi-element X-ray intensity spectra and BSE or SEI images for mineral phase identification and localization were acquired simultaneously using an electron probe microanalyzer. High-precision single-point analysis was performed on 10–20 representative sites in each mineral phase, and the oxide weight percentage data was exported after ZAF correction as internal standard data. Based on the internal standard data, the multi-element X-ray intensity spectrum is sequentially corrected for dead time, topography, time drift, mixed pixel, and standard position to obtain the corrected micro-area spectrum data.

3. The method for reconstructing the PTT trajectory of metamorphic rocks based on multi-source data fusion according to claim 1, characterized in that, S21 includes: The encoder extracts features from the input corrected micro-area map data through multiple convolutional layers, and outputs a feature map containing mineral features at different scales. The feature map is processed by the channel and spatial attention module. By calculating the channel and spatial weights, a weighted feature map is generated to highlight key mineral areas. The decoder upsamples and fuses the weighted feature map through a transposed convolutional layer, and outputs a mineral category probability map with the same size as the input map. By combining the classification output layer, pixel-by-pixel argmax operation is performed on the mineral category probability map to generate a mineral classification map.

4. The method for reconstructing the PTT trajectory of metamorphic rocks based on multi-source data fusion according to claim 1, characterized in that, In S22, the empirical α-factor correction algorithm is expressed as follows: in, This indicates the weight percentage concentration of the oxide of element i. Indicates the correction factor. The characteristic X-ray intensity of element i in the sample is expressed as... It is represented by the characteristic X-ray intensity of element i in the standard sample.

5. The method for reconstructing the PTT trajectory of metamorphic rocks based on multi-source data fusion according to claim 1, characterized in that, S23 includes: The mineral classification map and the elemental concentration distribution map are spatially registered at the pixel level to establish the correspondence between mineral phase categories and chemical compositions; Based on the mineral classification map, the percentage of area of ​​each mineral phase within the micro-region is calculated; The chemical composition data of all pixels in each mineral phase region are extracted from the element concentration distribution map, and the average value is calculated as the representative chemical composition of the mineral phase. The principal components of the whole rock are calculated using an area-weighted average algorithm. The results are then volume-corrected by combining the density values ​​of each mineral phase, and a dataset of principal components of the whole rock at the micro-area scale is output.

6. The method for reconstructing the PTT trajectory of metamorphic rocks based on multi-source data fusion according to claim 1, characterized in that, S4 includes: S41. Identify the zircon and monazite dating mineral regions in the mineral classification map; S42. Micro-area in-situ ablation is performed using a laser ablation system with a spot diameter of 10–30 μm, and instrument drift is corrected based on a combination of external and internal standards. S43. Collect U, Th, and Pb isotope signals, calculate U-Pb age using the isotope dilution algorithm, and obtain chronological data.

7. The method for reconstructing the PTT trajectory of metamorphic rocks based on multi-source data fusion according to claim 1, characterized in that, In S5, constructing the PTt three-dimensional trajectory model of metamorphic rocks includes: Based on the chronological data, an absolute time scale is provided for the PT trajectory. A continuous functional relationship of temperature-pressure-time is established through an interpolation algorithm, and the entire path evolution process of rock metamorphism is reconstructed in a three-dimensional coordinate system.

8. A metamorphic rock PTt trajectory reconstruction system based on multi-source data fusion, characterized in that, include: Data acquisition and correction module: used to acquire electron probe microanalytical data and internal standard data of metamorphic rock samples, and to perform multi-level data correction; The identification and quantitative interpretation module is used to identify mineral phases and quantitatively interpret chemical components in the corrected micro-area spectral data, generating mineral classification maps, elemental concentration distribution maps, and whole-rock major component datasets; including: A multi-scale convolutional neural network based on an attention mechanism is used to identify mineral phases in the corrected micro-area map data, generating a mineral classification map reflecting the distribution of each mineral phase. Based on internal standard sample data, an empirical correction algorithm is used to quantitatively convert the X-ray intensity in the micro-area map into oxide concentration, generating an elemental concentration distribution map. The mineral classification map and the elemental concentration distribution map are fused, and after weighted averaging and density correction, a whole-rock principal component dataset at the micro-area scale is calculated and output. PT Trajectory Extraction Module: Used to perform thermodynamic phase equilibrium simulation based on the whole-rock major component dataset to generate PT profile; and to extract PT trajectories from the PT profile based on the elemental concentration distribution map and multiphase equilibrium constraints; including: Based on the whole-rock major component dataset, a thermodynamic database was selected and a solid solution model was defined. The pressure-temperature simulation range and step size were set. The Gibbs free energy minimization algorithm was used to initially solve the stable mineral assemblages at each PT point through linear programming. Then, the composition and proportion of each phase were refined through a nonlinear iterative optimization algorithm to generate a PT profile. The nonlinear iterative optimization algorithm adopted the simulated annealing algorithm. Based on the mineral composition zoning structure or composition gradient information provided by the elemental concentration distribution map, combined with the multiphase equilibrium reaction relationship, the change of the stable domain boundary of a specific mineral assemblage was tracked in the PT profile to extract the metamorphic PT trajectory. Chronology module: used to perform micro-area in-situ chronological analysis on selected minerals in the mineral classification map to obtain chronological data; PTt Coupled Modeling Module: Used to spatiotemporally couple the PT trajectory with geochronological data, construct a three-dimensional PTt trajectory model of metamorphic rocks, and output a visual reconstruction report.

Images