Multi-source constraint reverse modeling method and system for chemical condition evolution of ancient lake water

Abstract

The invention provides a multi-source constraint reverse modeling method and system for ancient lake water chemical condition evolution, and relates to the technical field of mineral resource exploration, and the method comprises the steps: rock core element scanning, ore-bearing section XRD analysis, and obtaining of ancient lake water ion composition; judging the water chemical type of the ancient lake based on the water ion composition of the ancient lake, and determining the basic evolution path of the water chemical condition; measuring carbon, oxygen and sulfur isotopes, and analyzing isotope fractionation characteristics and control factors; a multiphase fluid mixing model is defined, a Monte Carlo algorithm is adopted to solve the mixing proportion, and reaction network initialization is carried out; dividing time steps of fresh water injection, evaporation concentration, salt precipitation critical periods and the like, carrying out dynamic simulation calculation, constructing a loss function between actual measurement and simulation results, carrying out back propagation correction by adopting an Adam optimizer, and verifying and optimizing the simulation results. The method has good applicability to the numerical simulation of the chemical condition evolution process of the ancient lake water.

Description

Technical Field

[0001] This application relates to the technical field of mineral resource exploration, and particularly to a multi-source constrained reverse modeling method and system for the evolution of paleo-lake hydrochemical conditions. Background Art

[0002] Paleo-lakes are important geological carriers of mineral resources such as salts, lithium, and potassium. These mineral resources are usually the products of the evolution of paleo-lake hydrochemical conditions to a specific stage under the background of evaporation and closed lake basins. Therefore, the simulation results of the evolution process of paleo-lake hydrochemical conditions can effectively guide mineral resource exploration work. The initial input parameters are the key to determining whether the simulation results can accurately reflect the actual geological process. The commonly used forward modeling method requires input of hydrochemical parameters such as temperature and pH, microbial kinetics such as sulfate reduction rate and methanogenic activity, mineralogical parameters such as dolomite solubility product and nucleation activation energy, and constraint conditions of reaction paths such as evaporation rate and CO2 partial pressure. For modern lakes, the above four types of parameters can be obtained through field investigations, but for paleo-lakes in the geological history period, it is almost impossible to accurately determine these four types of parameters.

[0003] The reverse modeling method is a mathematical method for inferring process parameters from known results. Its core idea is to, based on the measured results and combined with physical and chemical principles, invert the possible action paths that lead to these results. Compared with forward modeling, the reverse modeling method is more applicable to scenarios with missing or highly uncertain parameters. Its simulation input parameters include initial state characteristics, possible reaction paths, and necessary conditions for the final state to appear. For the above input parameters of the evolution process of paleo-lake hydrochemical conditions, they need to be jointly obtained by combining geological observation data, modern analog data, physical and chemical laws, and machine learning inversion results, etc. Summary of the Invention

[0004] The purpose of the present invention is to provide a multi-source constrained reverse modeling method and system for the evolution of paleo-lake hydrochemical conditions in order to solve the problem of establishing a digital model for the evolution process of paleo-lake hydrochemical conditions.

[0005] The above object of this application is achieved by the following technical solutions:

[0006] Step S1: Determine the ionic composition of the paleo-lake water body according to the core samples of the ore-bearing section of the paleo-lake;

[0007] Step S2: Based on the ionic composition of the paleo-lake water body, determine the basic evolution path of the hydrochemical conditions of the paleo-lake;

[0008] Step S3: Measure the three isotope ratios of carbon, oxygen, and sulfur in the core samples to determine the isotope fractionation characteristics and their controlling factors;

[0009] Step S4: Define a multiphase fluid mixing model, combine the isotope fractionation characteristics and their controlling factors, solve the mixing ratio, and initialize the reaction network;

[0010] Step S5: Set the time step of the reaction network and perform calculations to obtain the simulation results;

[0011] Step S6: Construct a loss function for the reaction network; verify and optimize the simulation results through the loss function and the Adam optimizer to complete the multi-source constrained inverse modeling of the evolution of the paleolake hydrochemical conditions.

[0012] Optionally, Step S1 includes:

[0013] Step S11: Use a handheld elemental analyzer to scan the core of the paleolake, obtain the vertical elemental variation information of the core, and determine the ore-bearing section of the paleolake;

[0014] Step S12: Collect core samples from the ore-bearing section; generate XRD diffraction patterns of the core samples through an X-ray diffractometer to determine the mineral composition of the core samples;

[0015] Step S13: Select representative samples from the core samples based on the mineral composition; determine the brine ion composition of the representative samples through a mass spectrometer, that is, obtain the ion composition of the paleolake water body.

[0016] Optionally, Step S2 includes:

[0017] Step S21: Determine the hydrochemical type of the paleolake according to the ion composition of the paleolake water body and the mineral composition of the core samples; the hydrochemical types include: carbonate type, sulfate type, chloride type, sodium type, and calcium-magnesium type;

[0018] Step S22: Determine the basic evolution path of the hydrochemical conditions of the paleolake according to the hydrochemical type of the paleolake, specifically as follows:

[0019] For a carbonate-type paleolake, the basic evolution path of its hydrochemical conditions is: fresh water period - slightly saline water period - saline water period - salt lake period, and the mineral composition includes: trona and lithium;

[0020] For a sulfate-type paleolake, the basic evolution path of its hydrochemical conditions is: fresh water period - gypsum precipitation period - mirabilite precipitation period - halite crystallization period, and the mineral composition includes: gypsum and potassium salt;

[0021] For a chloride-type paleolake, the basic evolution path of its hydrochemical conditions is: saline water period - halite precipitation period - potassium and magnesium salt precipitation period - dry salt lake period, and the mineral composition includes: halite, potassium and magnesium salts, and lithium and boron;

[0022] For the sodium-type ancient lake, the basic evolution path of its hydrochemical conditions is: alkaline fresh water period - soda precipitation period - diatomite deposition period - rare earth enrichment period, and the mineral composition includes: rare earth and diatomite;

[0023] For the calcium-magnesium type ancient lake, the basic evolution path of its hydrochemical conditions is: hard water period - dolomitization period - magnesite formation period - nitre deposition period, and the mineral composition includes: dolomite and magnesite.

[0024] Optionally, step S2 includes:

[0025] Step S21: Determine the hydrochemical type of the ancient lake according to the ion composition of the ancient lake water body and the mineral composition of the core samples; the hydrochemical types include: carbonate type, sulfate type, chloride type, sodium type and calcium-magnesium type;

[0026] Step S22: Determine the basic evolution path of the hydrochemical conditions of the ancient lake according to the hydrochemical type of the ancient lake, specifically as follows:

[0027] For the carbonate-type ancient lake, the basic evolution path of its hydrochemical conditions is: fresh water period - slightly saline water period - saline water period - salt lake period, and the mineral composition includes: trona and lithium;

[0028] For the sulfate-type ancient lake, the basic evolution path of its hydrochemical conditions is: fresh water period - gypsum precipitation period - mirabilite precipitation period - halite crystallization period, and the mineral composition includes: gypsum and potassium salt;

[0029] For the chloride-type ancient lake, the basic evolution path of its hydrochemical conditions is: saline water period - halite precipitation period - potassium and magnesium salt precipitation period - dry salt lake period, and the mineral composition includes: halite, potassium and magnesium salts, and lithium and boron;

[0030] For the sodium-type ancient lake, the basic evolution path of its hydrochemical conditions is: alkaline fresh water period - soda precipitation period - diatomite deposition period - rare earth enrichment period, and the mineral composition includes: rare earth and diatomite;

[0031] For the calcium-magnesium type ancient lake, the basic evolution path of its hydrochemical conditions is: hard water period - dolomitization period - magnesite formation period - nitre deposition period, and the mineral composition includes: dolomite and magnesite.

[0032] Optionally, step S4 includes:

[0033] Step S41: Define the atmospheric precipitation end-member fluid, deep heat flow end-member fluid and surface runoff end-member fluid;

[0034] Based on the three isotope ratios of carbon, oxygen and sulfur and the ion composition of the ancient lake water body, determine the identification markers of the atmospheric precipitation end-member fluid, deep heat flow end-member fluid and surface runoff end-member fluid;

[0035] Step S42: Use the Monte Carlo algorithm, combine the multi - end - member mixing model formula and the identification flag to solve the mixing ratio of the ancient fluid; initialize the reaction network according to the mixing ratio; the reaction network is a digital model used to simulate the evolution of the ancient lake water chemical conditions;

[0036] The multi - end - member mixing model formula is:

[0037] C obs = ∑(f i ·C i ) + ε

[0038] where C obs represents the measured chemical parameters, and the chemical parameters include concentration and isotope ratio; C i represents the chemical characteristics of the i - th end - member; f i represents the proportion of the i - th end - member in the mixing process; ε represents the error term.

[0039] Optionally, step S5 includes:

[0040] Step S51: Divide the time step of the reaction network according to the modern lake data of the same water chemical type as the ancient lake; the time step includes: the step length of the fresh - water injection period, the step length of the evaporation and concentration period, and the step length of the critical period of mineral precipitation;

[0041] Step S52: Use the CUDA architecture, divide the computational domain, perform parallel computing acceleration, and obtain the simulation result.

[0042] Optionally, step S6 includes:

[0043] Obtain the measured results of the core samples;

[0044] Construct a loss function including a data fitting term and a physical constraint term between the measured results and the simulation results, as follows

[0045]

[0046] where L represents the loss function, which measures the overall error between the prediction result of the reaction network and the real observed data; N represents the total number of observed data points participating in the calculation, which is used for the averaging of the mean square error; represents the simulation result of the reaction network for the i - th data point; represents the measured result of the i - th data point; λ represents the weight coefficient that balances the data fitting term and the physical constraint term, which controls the model complexity and the over - fitting risk; represents the absolute value of the change gradient of the sulfur isotope in the spatial or temporal dimension, usually referring to the first - order derivative or difference.

[0047] A multi - source - constrained inverse modeling system for the evolution of ancient lake water chemical conditions, the system includes:

[0048] A data acquisition module, a data processing module, and a display module;

[0049] The data acquisition module, the data processing module, and the display module are sequentially connected;

[0050] The data acquisition module is used to collect the ionic composition of the ancient lake water body;

[0051] The data processing module is used to determine the basic evolution path of the hydrochemical conditions of the ancient lake based on the ionic composition of the ancient lake water body;

[0052] The data processing module is also used to measure the isotope ratios of carbon, oxygen, and sulfur in the core samples, and analyze the isotope fractionation characteristics and their controlling factors;

[0053] The data processing module is also used to define a multiphase fluid mixing model, combine the isotope fractionation characteristics and their controlling factors, solve the mixing ratio, and initialize the reaction network;

[0054] The data processing module is also used to set the time step of the reaction network and perform calculations to obtain simulation results;

[0055] The data processing module is also used to construct a loss function for the reaction network; through the loss function and the Adam optimizer, verify and optimize the simulation results, and complete the multi-source constrained inverse modeling of the evolution of the hydrochemical conditions of the ancient lake;

[0056] The display module is used to visually display the simulation results

[0057] The beneficial effects brought by the technical solution provided by this application are:

[0058] Through the multi-source constraints of geological observation data, modern analogy data, physical and chemical laws, and machine learning inversion results, the complementarity of heterogeneous data and the cross-validation of multidisciplinary theories are improved. Through backpropagation correction, combined with data-driven fitting and physical law constraints, the efficient optimization of geochemical model parameters is realized, and the inversion accuracy and inversion efficiency are improved. It provides a more accurate and efficient scientific method for the numerical simulation of the evolution process of the hydrochemical conditions of ancient lakes, and has good prospects in the field of mineral exploration. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The following will further illustrate this application in conjunction with the drawings. In the drawings:

[0060] Figure 1 is the step diagram in the embodiment of this application;

[0061] Figure 2 is the core element scanning result diagram in the embodiment of this application;

[0062] Figure 3 It is the system module diagram in the embodiment of the present application. Specific implementation manners

[0063] For a clearer understanding of the technical features, objectives, and effects of the present application, the specific implementation manners of the present application will now be described in detail with reference to the accompanying drawings.

[0064] The embodiment of the present application provides a multi-source constrained reverse modeling method for the evolution of paleolake hydrochemical conditions.

[0065] Please refer to Figure 1 , Figure 1 It is the step diagram of a multi-source constrained reverse modeling method for the evolution of paleolake hydrochemical conditions in the embodiment of the present application, including:

[0066] Step S1: Determine the ionic composition of the paleolake water body according to the core samples of the ore-bearing section of the paleolake;

[0067] Step S2: Based on the ionic composition of the paleolake water body, determine the basic evolution path of the hydrochemical conditions of the paleolake;

[0068] As an embodiment, the basic evolution path of the hydrochemical conditions provides a basic framework for the evolution process in Steps S4, S5, and S6. For example, the carbonate type generally follows the evolution process of "freshwater period - brackish water period - saline water period - salt lake period", and does not show evolution processes such as "gypsum precipitation period - mirabilite precipitation period".

[0069] Step S3: Measure the three isotope ratios of carbon, oxygen, and sulfur in the core samples to determine the isotope fractionation characteristics and their controlling factors;

[0070] Step S4: Define a multi-phase fluid mixing model, combine the isotope fractionation characteristics and their controlling factors, solve the mixing ratio, and initialize the reaction network;

[0071] Step S5: Set the time step of the reaction network and perform calculations to obtain the simulation results;

[0072] Step S6: Construct the loss function of the reaction network; through the loss function and the Adam optimizer, verify and optimize the simulation results to complete the multi-source constrained reverse modeling of the evolution of paleolake hydrochemical conditions.

[0073] Step S1 includes:

[0074] As an embodiment, conduct core element scanning: sample the ore-bearing section for X-ray diffraction analysis, and conduct laser ablation inductively coupled plasma mass spectrometry analysis on the primary inclusions to reveal the ionic composition of the paleolake water body.

[0075] Step S11: Use a handheld elemental analyzer to scan the core of the ancient lake, obtain the vertical variation information of the elements in the core, and determine the ore-bearing section of the ancient lake;

[0076] The present application provides an embodiment as follows. When using a handheld elemental analyzer, a micro X-ray tube, an Ag target, and an SDD detector need to be configured to scan the core, obtain the vertical variation of elements (or oxides) such as Si, Al, Fe, K, Ti, Ca, etc., and determine the ore-bearing section. For example, Figure 2 as shown. In the layer with high Ca and Mg contents, dolomite (CaMg(CO)) may be enriched; in the layer with only high Ca content, calcite (CaCO) may be enriched; in the layer with only high Mg content, magnesite (MgCO) may be enriched; in the layer with high Ca and S contents, gypsum (CaSO·2HO) may be enriched; in the layer with high Na and S contents, mirabilite (NaSO·10HO) may be enriched; in the layer with high K and Cl contents, potassium salt (KCl) may be enriched; in the layer with high Na and Cl contents, halite (NaCl) may be enriched.

[0077] Step S12: Collect the core samples of the ore-bearing section; use an X-ray diffractometer to generate the XRD diffraction pattern of the core samples and determine the mineral composition of the core samples;

[0078] The present application provides an embodiment as follows. In XRD analysis, different minerals will exhibit unique diffraction peaks (d values and relative intensities) due to differences in their crystal structures. The main peak d value of dolomite is and the second-strongest peak is The main peak d value of calcite is and the second-strongest peak is The main peak d value of magnesite is and the second-strongest peak is The main peak d value of gypsum is and the second-strongest peak is The main peak d value of mirabilite is and the second-strongest peak is The main peak d value of potassium salt is and the second-strongest peak is The main peak d value of halite is and the second-strongest peak is

[0079] Step S13: Select representative samples of the core samples based on the mineral composition; use a mass spectrometer to measure the brine ion composition of the representative samples, and thus obtain the ion composition of the ancient lake water body.

[0080] Step S2 includes:

[0081] Step S21: Determine the hydrochemical type of the ancient lake based on the ionic composition of the ancient lake water body and the mineral composition of the core samples; the hydrochemical types include: carbonate type, sulfate type, chloride type, sodium type, and calcium-magnesium type;

[0082] As an example, based on the ionic composition of the ancient lake water body, determine the hydrochemical type of the ancient lake and determine the basic evolution path of its hydrochemical conditions. The hydrochemical types include: carbonate type (CO3 2- +HCO3 - )>SO4 2- >Cl - , sulfate type SO4 2- >(CO3 2- +HCO3 - )>Cl - , chloride type Cl - >SO4 2- >(CO3 2- +HCO3 - ), sodium type Na + +K + >Mg 2+ >Ca 2+ (cation-dominated), calcium-magnesium type Ca 2+ +Mg 2+ >Na + +K + (cation-dominated).

[0083] Step S22: Determine the basic evolution path of the hydrochemical conditions of the ancient lake according to the hydrochemical type of the ancient lake, specifically as follows:

[0084] For an ancient lake of carbonate type, the basic evolution path of its hydrochemical conditions is: fresh water period - slightly saline water period - saline water period - salt lake period, and the mineral composition includes: trona and lithium;

[0085] For an ancient lake of sulfate type, the basic evolution path of its hydrochemical conditions is: fresh water period - gypsum precipitation period - mirabilite precipitation period - halite crystallization period, and the mineral composition includes: gypsum and potassium salt;

[0086] For an ancient lake of chloride type, the basic evolution path of its hydrochemical conditions is: saline water period - halite precipitation period - potassium and magnesium salt precipitation period - dry salt lake period, and the mineral composition includes: halite, potassium and magnesium salts, and lithium and boron;

[0087] For an ancient lake of sodium type, the basic evolution path of its hydrochemical conditions is: alkaline fresh water period - soda precipitation period - diatomite deposition period - rare earth enrichment period, and the mineral composition includes: rare earth and diatomite;

[0088] For an ancient lake of the calcium-magnesium type, the basic evolution path of its hydrochemical conditions is: hard water period - dolomitization period - magnesite formation period - nitrate deposition period, and the mineral composition includes dolomite and magnesite.

[0089] Step S3 includes:

[0090] The isotope fractionation characteristics and their controlling factors include: carbon isotope fractionation characteristics and their controlling factors, oxygen isotope fractionation characteristics and their controlling factors, and sulfur isotope fractionation characteristics and their controlling factors;

[0091] Step S31: Use a stable isotope mass spectrometer to measure the isotopes of carbon, oxygen, and sulfur in the core sample to obtain the carbon isotope ratio, oxygen isotope ratio, and sulfur isotope ratio;

[0092] Step S32: According to the carbon isotope ratio, analyze the control of biological processes, temperature, chemical reaction types, and diffusion on carbon isotope fractionation, and determine the carbon isotope fractionation characteristics and their controlling factors;

[0093] As an example, biological processes include photosynthesis (such as C3 / C4 plant fractionation), microbial methane production / oxidation (fractionation can reach -50‰ to +20‰). Temperature means that high temperature reduces the fractionation effect, such as the d13C of thermally cracked methane approaching the parent material. Chemical reaction types refer to the thermal evolution of organic matter (such as kerogen cracking), carbonate precipitation (fractionation is controlled by pH and mineral phase). Diffusion means that light isotopes (12C) migrate preferentially, and usually have little effect on the carbon isotope fractionation of rocks. If d13C is significantly negatively biased (such as < -30‰), it may be related to microbial methane production / oxidation; if it is related to the organic matter maturity index (such as vitrinite reflectance), it indicates that isotope fractionation is mainly affected by thermal evolution.

[0094] Step S33: According to the oxygen isotope ratio, analyze the control of temperature, water-rock interaction, and evaporation on oxygen isotope fractionation, and determine the oxygen isotope fractionation characteristics and their controlling factors;

[0095] As an example, temperature means that the oxygen isotope fractionation between minerals and water is smaller at high temperature. Water-rock interaction includes hydrothermal alteration and diagenesis. Evaporation means that the evaporation of water causes the preferential loss of oxygen isotopes. Use the calcite-water fractionation equation to calculate the mineral formation temperature. If it is consistent with the temperature of the coexisting minerals, it is equilibrium fractionation; if it deviates, it may be affected by later fluids.

[0096] Step S34: According to the sulfur isotope ratio, analyze the control of bacterial sulfate reduction, thermochemical sulfate reduction, volcanic activity, and sedimentary environment on sulfur isotope fractionation, and determine the sulfur isotope fractionation characteristics and their controlling factors.

[0097] As an example, the fractionation of bacterial sulfate reduction (BSR) can reach -70‰. The fractionation of thermochemical sulfate reduction (TSR) is relatively small, usually <20‰, and the kinetic effect dominates at high temperatures. Volcanism affects sulfur isotope fractionation through thermal effects. In anoxic environments, the δ 34 S of pyrite can reflect the evolution of the ancient water body sulfate reservoir better than that of the whole rock sample. The δ 34 S fractionation range >40‰ strongly indicates BSR; if it is positively correlated with the organic matter content, it may be dominated by biology.

[0098] Step S4 includes:

[0099] Step S41: Define the atmospheric precipitation end-member fluid, the deep heat flow end-member fluid, and the surface runoff end-member fluid;

[0100] Based on the three isotope ratios of carbon, oxygen, and sulfur and the ionic composition of the ancient lake water body, determine the identification markers of the atmospheric precipitation end-member fluid, the deep heat flow end-member fluid, and the surface runoff end-member fluid;

[0101] Step S42: Use the Monte Carlo algorithm, combine the multi-end-member mixing model formula and the identification markers to solve the mixing ratio of the ancient fluid; initialize the reaction network according to the mixing ratio; the reaction network is a digital model used to simulate the evolution of the ancient lake water chemical conditions;

[0102] The multi-end-member mixing model formula is:

[0103] C obs =∑(f i ·C i )+ε

[0104] where C obs represents the measured chemical parameter, and the chemical parameter includes: concentration and isotope ratio; C i represents the chemical characteristics of the i-th end-member; f i represents the proportion of the i-th end-member in the mixing process; ε represents the error term.

[0105] Step S5 includes:

[0106] Step S51: According to the modern lake data of the same water chemical type as the ancient lake, divide the time step of the reaction network; the time step includes: the time step of the fresh water injection period, the time step of the evaporation and concentration period, and the time step of the critical period of mineral precipitation;

[0107] As an example, according to the modern lake data of the same water chemical type, divide the time step; the time step includes: the time step of the fresh water injection period (used to simulate the ion dilution process), the time step of the evaporation and concentration period (used for the diffusion calculation driven by the concentration gradient), and the time step of the critical period of mineral precipitation (used for the dynamic monitoring of mineral saturation).

[0108] Step S52: Adopt the CUDA architecture, divide the computational domain, perform parallel computing acceleration, and obtain the simulation results.

[0109] As an embodiment, adopt the CUDA architecture, divide the computational domain into 512×512 grid blocks, and each GPU thread processes a chemical reaction equation. The computational domain refers to the simulation scope of each step.

[0110] Step S6 includes:

[0111] Obtain the measured results of the core samples;

[0112] Construct a loss function containing a data fitting term and a physical constraint term between the measured results and the simulation results, as follows

[0113]

[0114] where L represents the loss function, which measures the overall error between the prediction results of the reaction network and the real observed data; N represents the total number of observed data points participating in the calculation, which is used for the averaging of the mean square error; represents the simulation result of the reaction network for the i-th data point; represents the measured result of the i-th data point; λ represents the weight coefficient for balancing the data fitting term and the physical constraint term, which controls the model complexity and the overfitting risk; represents the absolute value of the change gradient of sulfur isotopes in the spatial or temporal dimension, usually referring to the first derivative or difference.

[0115] As an embodiment, the loss function containing the data fitting term and the physical constraint term is used for the simulation of the evolution of the paleolake water chemical conditions, which modifies the meaning of the parameters and redefines the constraint of the simulation results using sulfur isotopes (i.e., ), which is the innovation point of the present invention.

[0116] Please refer to Figure 3 , Figure 3 which is a multi-source constraint inverse modeling system for the evolution of paleolake water chemical conditions in an embodiment of the present application. The system includes:

[0117] A data acquisition module, a data processing module, and a display module;

[0118] The data acquisition module, the data processing module, and the display module are sequentially connected;

[0119] The data acquisition module is used to collect the ionic composition of the paleolake water body;

[0120] The data processing module is used to determine the basic evolution path of the water chemical conditions of the paleolake based on the ionic composition of the paleolake water body;

[0121] The data processing module is also used to measure the isotope ratios of carbon, oxygen and sulfur in core samples, and analyze the isotope fractionation characteristics and their controlling factors;

[0122] The data processing module is also used to define a multiphase fluid mixing model, combine the isotope fractionation characteristics and their controlling factors, solve the mixing ratio, and initialize the reaction network;

[0123] The data processing module is also used to set the time step of the reaction network and perform calculations to obtain simulation results;

[0124] The data processing module is also used to construct a loss function for the reaction network; through the loss function and the Adam optimizer, verify and optimize the simulation results, and complete the multi-source constrained inverse modeling of the evolution of paleolake hydrochemical conditions;

[0125] The display module is used to visually display the simulation results.

[0126] Table 1 is a table of the chemical compositions of three types of end-member fluids in the embodiments of the present invention. Table 2 is a table of the time step division in the embodiments of the present invention.

[0127] Endmember type Characteristic index Atmospheric precipitation δ18O < -8‰ Deep hydrothermal fluid Li / B > 0.15 Terrestrial runoff Sr / Ca > 0.02

[0128] Table 1

[0129] Stage Time resolution Calculation content Freshwater injection period 10 years / step Simulation of ion dilution process Evaporation and concentration period 1 year / step Diffusion calculation driven by concentration gradient Critical period of salt precipitation 0.1 year / step Dynamic mineral saturation

[0130] Table 2

[0131] This application also discloses a computer-readable storage medium, which stores multiple instructions suitable for being loaded by a processor to execute the above-mentioned multi-source constrained inverse modeling method for the evolution of paleolake hydrochemical conditions.

[0132] The above are only exemplary embodiments of the present disclosure, and the scope of the present disclosure cannot be limited thereby. That is, any equivalent changes and modifications made in accordance with the teachings of the present disclosure still fall within the scope covered by the present disclosure.

[0133] This application aims to cover any variations, uses or adaptive changes of the present disclosure, which follow the general principles of the present disclosure and include common general knowledge or conventional technical means in the technical field not recorded in the present disclosure. The description and embodiments are only regarded as exemplary, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A multi-source constrained inverse modeling method for the evolution of water chemical conditions in ancient lakes, characterized in that: The method comprises the following steps: Step S1: Determine the ion composition of the water body of the ancient lake based on the core sample of the ore-bearing section of the ancient lake; Step S2: Determine the basic evolution path of the water chemical conditions of the ancient lake based on the ion composition of the ancient lake water body; Step S3: Determine the isotope ratios of carbon, oxygen and sulfur in the core sample to determine the isotope fractionation characteristics and their controlling factors; Step S4: define a multiphase fluid mixing model, combine the isotope fractionation characteristics and its control factors, solve the mixing ratio, and initialize the reaction network; Step S5: setting the time step of the reaction network and performing calculations to obtain simulation results; Step S6: Construct the loss function of the reaction network; verify and optimize the simulation results through the loss function and Adam optimizer to complete the multi-source constrained inverse modeling of the evolution of water chemical conditions in ancient lakes.

2. A multi-source constrained inverse modeling method for the evolution of water chemical conditions in ancient lakes as claimed in claim 1, characterized in that: Step S1 includes: Step S11: using a handheld element analyzer to scan the core of the ancient lake, obtain the vertical element change information of the core, and determine the ore-bearing section of the ancient lake; Step S12: collecting core samples from the ore-bearing section; generating an XRD diffraction pattern of the core samples by an X-ray diffractometer to determine the mineral composition of the core samples; Step S13: Select representative samples of the core samples according to the mineral composition; measure the brine ion composition of the representative samples by mass spectrometer to obtain the ion composition of the ancient lake water.

3. A multi-source constrained inverse modeling method for the evolution of water chemical conditions in ancient lakes as claimed in claim 2, characterized in that: Step S2 includes: Step S21: Determine the hydrochemical type of the ancient lake according to the ion composition of the ancient lake water and the mineral composition of the core sample; the hydrochemical types include: carbonate type, sulfate type, chloride type, sodium type and calcium-magnesium type; Step S22: Determine the basic evolution path of the hydrochemical conditions of the ancient lake according to the hydrochemical type of the ancient lake, as follows: The basic evolution path of water chemical conditions of carbonate paleo-lakes is: freshwater period - brackish water period - saltwater period - salt lake period, and the mineral composition includes: natural alkali and lithium; The basic evolution path of water chemical conditions of sulfate-type ancient lakes is: freshwater period - gypsum precipitation period - mirabilite precipitation period - rock salt crystallization period, and the mineral composition includes: gypsum and potassium salt; The basic evolution path of hydrochemical conditions of chloride-type ancient lakes is: salt water period - rock salt precipitation period - potassium and magnesium salt precipitation period - dry salt lake period. The mineral composition includes: rock salt, potassium and magnesium salts, and lithium boron. The basic evolution path of water chemical conditions of sodium-type ancient lakes is: alkaline freshwater period - soda precipitation period - diatomaceous earth deposition period - rare earth enrichment period. The mineral composition includes rare earth and diatomaceous earth. The basic evolution path of the water chemical conditions of the calcium-magnesium type ancient lake is: hard water period - dolomite period - magnesite formation period - saltpeter deposition period, and the mineral composition includes: dolomite and magnesite.

4. The multi-source constrained inverse modeling method for the evolution of water chemical conditions in ancient lakes according to claim 1, characterized in that: Step S3 includes: The isotope fractionation characteristics and their controlling factors include: carbon isotope fractionation characteristics and their controlling factors, oxygen isotope fractionation characteristics and their controlling factors, and sulfur isotope fractionation characteristics and their controlling factors; Step S31: using a stable isotope mass spectrometer to measure the carbon, oxygen and sulfur isotopes of the core sample to obtain the carbon isotope ratio, oxygen isotope ratio and sulfur isotope ratio; Step S32: Analyze the control of carbon isotope fractionation by biological process, temperature, chemical reaction type, and diffusion according to the carbon isotope ratio, and determine the carbon isotope fractionation characteristics and its control factors; Step S33: Analyze the control of temperature, water-rock interaction, and evaporation on oxygen isotope fractionation according to the oxygen isotope ratio, and determine the oxygen isotope fractionation characteristics and its controlling factors; Step S34: Analyze the control of sulfur isotope fractionation by bacterial sulfate reduction, thermochemical sulfate reduction, volcanic activity, and sedimentary environment based on the sulfur isotope ratio, and determine the sulfur isotope fractionation characteristics and its controlling factors.

5. The multi-source constrained inverse modeling method for the evolution of water chemical conditions in ancient lakes according to claim 1, characterized in that: Step S4 includes: Step S41: defining atmospheric precipitation end member fluid, deep heat flow end member fluid and surface runoff end member fluid; Based on the isotopic ratios of carbon, oxygen and sulfur and the ion composition of ancient lake water, the identification marks of atmospheric precipitation end member fluid, deep thermal flow end member fluid and surface runoff end member fluid are determined; Step S42: using a Monte Carlo algorithm, combined with a multi-end member mixing model formula and identification marks, to solve the mixing ratio of the paleofluid; initializing the reaction network according to the mixing ratio; the reaction network is a digital model for simulating the evolution of the hydrochemical conditions of the paleolake; The multi-endmember mixing model formula is: C obs =∑(f i ·C i )+ε Among them, C obs Represents the measured chemical parameters, including concentration and isotope ratio; C i represents the chemical characteristics of the i-th end member; f i represents the proportion of the i-th end member in the mixing process; ε represents the error term.

6. The multi-source constrained inverse modeling method for the evolution of water chemical conditions in ancient lakes according to claim 1, characterized in that: Step S5 includes: Step S51: dividing the time steps of the reaction network according to the modern lake data of the same hydrochemical type as the ancient lake; the time steps include: freshwater injection period step, evaporation concentration period step and mineral precipitation critical period step; Step S52: using the CUDA architecture, dividing the computational domain, performing parallel computation acceleration, and obtaining simulation results.

7. The multi-source constrained inverse modeling method for the evolution of water chemical conditions in ancient lakes according to claim 1, characterized in that: Step S6 includes: Obtain measured results of core samples; Between the measured results and the simulation results, a loss function including data fitting terms and physical constraint terms is constructed as follows Where L represents the loss function, which measures the overall error between the prediction results of the reaction network and the actual observation data; N represents the total number of observation data points involved in the calculation, which is used to average the mean square error; represents the simulation result of the reaction network for the i-th data point; represents the measured result of the ith data point; λ represents the weight coefficient of the balanced data fitting term and the physical constraint term, controlling the model complexity and overfitting risk; Represents the absolute value of the gradient of sulfur isotope change in space or time dimension, usually refers to the first-order derivative or difference.

8. A multi-source constrained inverse modeling system for the evolution of water chemical conditions in ancient lakes, used to implement a multi-source constrained inverse modeling method for the evolution of water chemical conditions in ancient lakes as described in any one of claims 1 to 7, characterized in that: The system comprises: Data acquisition module, data processing module and display module; The data acquisition module, the data processing module and the display module are connected in sequence; The data acquisition module is used to collect the ion composition of ancient lake water; The data processing module is used to determine the basic evolution path of the water chemical conditions of the ancient lake based on the ion composition of the ancient lake water body; The data processing module is also used to determine the isotope ratios of carbon, oxygen and sulfur in the core sample and analyze the isotope fractionation characteristics and their controlling factors; The data processing module is also used to define a multiphase fluid mixing model, combine isotope fractionation characteristics and control factors thereof, solve the mixing ratio, and initialize the reaction network; The data processing module is also used to set the time step of the reaction network and perform calculations to obtain simulation results; The data processing module is also used to construct a loss function of the reaction network; through the loss function and the Adam optimizer, the simulation results are verified and optimized to complete the multi-source constrained reverse modeling of the evolution of the water chemical conditions of the ancient lake; The display module is used to visualize the simulation results.

Images