Markov chain Monte Carlo-based paleoclimate inversion method, system and equipment

Abstract

The invention belongs to the technical field of climate prediction, particularly relates to a Markov chain Monte Carlo-based paleoclimate inversion method, system and equipment, and aims to solve the problems of accuracy and high efficiency of tracking climate change details in global inversion. The method comprises the following steps: determining a to-be-inverted paleoclimate index; inputting the to-be-inverted paleoclimate index into a paleoclimate inversion model to obtain a paleoclimate inversion result output by the paleoclimate inversion model; wherein the paleoclimate inversion model is obtained by performing iterative matching based on a target curve of a paleoclimate index and an initial parameter set determined based on the target curve of the paleoclimate index so as to continuously update a to-be-inverted parameter set. According to the method, in global inversion, climate change details are tracked more accurately, calculation is more efficient, and the response capability and adaptability to the time sequence are improved.

Description

Background Art

[0002] At present, many methods in the field of paleoclimate simulation are mainly based on forward models, that is, by presetting the boundary conditions of the geological period (such as temperature, carbon dioxide concentration and solar radiation, etc.), the corresponding climate state is calculated using numerical climate models. Forward models have efficiency advantages in determining climate trends and overall changes, but their results are highly dependent on initial conditions and parameter settings. They have certain limitations in processing complex and multivariable paleoclimate data, and it is difficult to fully explore the potential uncertainties and detailed changes in the climate system.

[0003] In order to overcome the limitations of the forward modeling method, inversion technology has gradually been introduced into paleoclimate reconstruction. The inversion method integrates geological record data and statistical models, and combines observational and simulation data to infer the possible evolution path of climate change. Unlike forward modeling, inversion technology allows the gradual reconstruction of multiple possible states of climate change and their probability distribution based on geological evidence, making the simulation results more diverse and scientifically explanatory.

[0004] In the paleoclimate inversion technology, the combination of Markov chain Monte Carlo (MCMC) method and Bayesian inversion technology has made significant progress in recent years. The MCMC algorithm is an algorithm for sampling probability distribution by constructing a Markov chain, which is particularly suitable for sampling problems of high-dimensional and complex probability distributions. In paleoclimate reconstruction, MCMC can effectively handle the complexity of multi-parameter and high-dimensional data by approximating the posterior distribution of climate variables through random sampling in parameter space. Bayesian inversion is based on Bayesian statistical theory, combining the prior information of geological records and the likelihood function of observed data to estimate the posterior distribution of parameters. Through Bayesian inversion, researchers can gradually optimize the uncertainty range of parameters based on the matching degree of geological records, thereby obtaining a climate change path that conforms to the observed data. This multi-parameter estimation method that combines MCMC and Bayesian inversion is not only applicable to single variables or specific conditions, but also provides an effective way in multi-dimensional climate data reconstruction, providing a guarantee for the accuracy and reliability of paleoclimate reconstruction.

[0005] However, based on global inversion, the overall climate change trend is identified by processing the data of all time steps at one time. Although it performs well in identifying the overall change trend, it has the following defects in detail description and accuracy: Insufficient detail resolution: Due to the global inversion, this method tends to prioritize finding the overall trend in the data and ignores subtle changes in the time series. This results in the inversion results being less sensitive to small climate fluctuations and making it difficult to accurately capture the details of climate change.

[0006] Reason: The global inversion method processes all time step data simultaneously, which makes the method tend to look for trend characteristics rather than progressively characterize the changes in each time step. This is particularly obvious when processing long time series data.

[0007] Limited accuracy: In multi-parameter inversion, although the global model improves the calculation speed, it sacrifices a certain degree of parameter detail accuracy. In practical applications, this method is prone to over-smoothing the inversion results, resulting in smoothing of abnormal signals in complex climate events.

[0008] Reason: Global inversion introduces overall approximation, which may cause deviations from the actual parameter values ​​at individual time steps, making it difficult to effectively reflect the real changes in each stage.

[0009] Lack of time progressiveness: The technology lacks the ability to adjust and update in a time-series progressive manner, and cannot effectively track parameters that change dynamically over time. In the case of rapidly changing geological and climatic conditions, the global inversion method has poor adaptability to nonlinear changes in the time dimension.

[0010] Reason: The global processing mode cannot be dynamically adjusted according to the results of each time step, resulting in weak timing logic of the time series, which is not conducive to dynamic climate inversion needs. Summary of the invention

[0011] In order to solve the above-mentioned problems in the prior art, namely, the accuracy and efficiency of tracking climate change details in global inversion, the present invention provides a paleoclimate inversion method, system and device based on Markov Chain Monte Carlo, which not only makes the tracking of climate change details more accurate and the calculation more efficient in global inversion, but also improves the responsiveness and adaptability to time series.

[0012] One aspect of the present invention provides a paleoclimate inversion method based on Markov chain Monte Carlo, comprising: Determine the paleoclimate indicators to be inverted; Inputting the paleoclimate index to be inverted into the paleoclimate inversion model to obtain the paleoclimate inversion result output by the paleoclimate inversion model; The paleoclimate inversion model is obtained by iteratively matching a target curve of paleoclimate indicators and an initial parameter set determined based on the target curve of paleoclimate indicators to continuously update the parameter set to be inverted.

[0013] Preferably, the paleoclimate inversion model is obtained by iteratively matching a target curve of paleoclimate indicators and an initial parameter set determined based on the target curve of paleoclimate indicators to continuously update the parameter set to be inverted, and includes: Based on the initial parameter set, a plurality of parameter sets to be inverted are randomly perturbed in a time-series progressive perturbation manner to generate a plurality of perturbation parameter sets as perturbation parameter sets; Based on the initial parameter set and the disturbance parameter set, obtaining a simulated value of a model output; Obtaining a log likelihood or an error value based on the climate indicator, the simulation value, and a target value of the target curve; The parameter set to be inverted is iteratively updated based on the logarithmic likelihood or the error value to obtain a paleoclimate inversion model including a final inversion parameter set.

[0014] Preferably, the randomly perturbing the initial parameter set in a time-series progressive perturbation manner to generate a plurality of parameter sets to be inverted as perturbation parameter sets comprises: Preset initial time and time window for multiple time periods; Based on a preset initial time and gradually expanding according to a preset time window, the parameter values ​​of the initial parameter set are randomly perturbed within a preset range in each time period to obtain the multiple parameter sets to be inverted as the perturbed parameter sets.

[0015] Preferably, obtaining a simulated value of a model output based on the initial parameter set and the disturbance parameter set comprises: The initial parameter set and the disturbance parameter set are respectively input into the paleoclimate inversion model to obtain corresponding simulation results under the initial state and the disturbance state.

[0016] Preferably, the paleoclimate indicators include one or more of organic carbon isotopes, atmospheric carbon dioxide concentration, and global temperature; The logarithmic likelihood or error value is obtained based on the paleoclimate index, the simulation value and the target value of the target curve, including: When the paleoclimate indicator is organic carbon isotope, the logarithmic likelihood is calculated based on the simulation value and the target value of the target curve; Otherwise, the error value is calculated according to the simulation value and the target value of the target curve.

[0017] Preferably, the calculating the error value according to the simulation value and the target value of the target curve includes: If the simulated value is within the confidence interval of the target curve observation data, the error value is set to zero; If the simulated value exceeds the confidence interval of the target curve observation data, the error value is calculated based on the mean deviation.

[0018] Preferably, the iterative updating of the parameter set to be inverted based on the log-likelihood or the error value comprises: Based on the error value, the total error ll0 of the initial parameter set and the errors ll1 to llN of the perturbation parameter set are calculated and compared: if lli>ll0, φ0 is replaced by the new parameter set φi to enter the next iteration; otherwise, the new candidate value is accepted based on the log-likelihood; wherein 1<=i<=N.

[0019] Another aspect of the present invention provides a paleoclimate inversion system based on Markov chain Monte Carlo, comprising: An index determination module is used to determine the paleoclimate index to be inverted; A climate inversion module, used for inputting the paleoclimate index to be inverted into the paleoclimate inversion model to obtain the paleoclimate inversion result output by the paleoclimate inversion model; The paleoclimate inversion model is obtained by iteratively matching a target curve of paleoclimate indicators and an initial parameter set determined based on the target curve of paleoclimate indicators to continuously update the parameter set to be inverted.

[0020] According to a third aspect of the present invention, an electronic device is provided, comprising: at least one processor; and a memory communicatively coupled to at least one of the processors; The memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the above-mentioned paleoclimate inversion method based on Markov chain Monte Carlo.

[0021] In a fourth aspect of the present invention, a computer-readable storage medium is proposed, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to be executed by the computer to implement the above-mentioned Markov Chain Monte Carlo-based paleoclimate inversion method.

[0022] Beneficial effects of the present invention: (1) A Bayesian inversion method based on a progressive time series model that gradually processes time series data to more accurately track climate change details; (2) By introducing a time series progressive model, the inversion process can be dynamically updated at each time step, thus improving the responsiveness and adaptability to time series; (3) By combining the Markov Chain Monte Carlo (MCMC) method with the time series progressive model, it is possible to retain high-precision detail inversion while maintaining computational efficiency, providing better performance in the inversion of long-term climate data. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings: Figure 1 It is a flow chart of the paleoclimate inversion method based on Markov Chain Monte Carlo; Figure 2 It is a flowchart of iterative matching of inversion parameters in the paleoclimate inversion model; Figure 3 It is a schematic diagram of the iterative process of Markov chain Monte Carlo temporal progressive Bayesian inversion for δ13C data; Figure 4 This is the high-precision fitting effect diagram of the target curve by Bayesian inversion; Figure 5 It is a structural block diagram of a paleoclimate inversion system based on Markov chain Monte Carlo; Figure 6 It is a structural diagram of a computer system of a server for implementing the method, system, and device embodiments of the present application. DETAILED DESCRIPTION

[0024] The present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the relevant invention are shown in the accompanying drawings.

[0025] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0026] The present invention provides a paleoclimate inversion method based on Markov Chain Monte Carlo. The method introduces a time-series progressive model so that the inversion process can be dynamically updated at each time step, thereby improving the responsiveness and adaptability to time series. At the same time, by combining the Markov Chain Monte Carlo (MCMC) method with the time-series progressive model, it is possible to retain high-precision detail inversion while maintaining computational efficiency.

[0027] A paleoclimate inversion method based on Markov chain Monte Carlo of the present invention comprises: Determine the paleoclimate indicators to be inverted; Inputting the paleoclimate index to be inverted into the paleoclimate inversion model to obtain the paleoclimate inversion result output by the paleoclimate inversion model; The paleoclimate inversion model is obtained by iteratively matching a target curve of paleoclimate indicators and an initial parameter set determined based on the target curve of paleoclimate indicators to continuously update the parameter set to be inverted.

[0028] In order to more clearly explain the paleoclimate inversion method based on Markov chain Monte Carlo of the present invention, Figure 1 Each step in the embodiment of the present invention is described in detail.

[0029] The paleoclimate inversion method based on Markov chain Monte Carlo in the first embodiment of the present invention is as follows: Figure 1 As shown, it includes step S101-step S102, and each step is described in detail as follows: Step S101, determining the paleoclimate index to be inverted; Specifically, first of all, in response to the needs of paleoclimate inversion, organic carbon isotopes (δ13C), atmospheric carbon dioxide concentration (pCO2) and temperature were selected as the main alternative indicators. These indicators can directly reflect the climate change characteristics of the geological period, and have high accessibility and reliability in the existing geological records. As an important record of the carbon cycle, organic carbon isotopes can indicate the changing process of carbon burial and atmospheric carbon dioxide, and provide information on the relationship between the carbon cycle and climate change. Carbon dioxide is one of the key driving factors of paleoclimate change, and changes in its concentration directly affect global temperature fluctuations. Therefore, as an important climate indicator, it can provide a reliable target curve for inversion. It should be noted that since temperature has a latitude effect, temperature changes in different geographical regions will be affected by latitude factors. Therefore, in the simulation process, the local temperature needs to be converted to the global average temperature to eliminate the latitude bias and obtain a more accurate global climate change trend.

[0030] Step S102, inputting the paleoclimate index to be inverted into the paleoclimate inversion model to obtain the paleoclimate inversion result output by the paleoclimate inversion model; The paleoclimate inversion model is obtained by iteratively matching a target curve of paleoclimate indicators and an initial parameter set determined based on the target curve of paleoclimate indicators to continuously update the parameter set to be inverted.

[0031] Specifically, in order to ensure the consistency of each alternative index in time, this method uses the interpolation method to unify the resolution of the target curve in view of the resolution difference, ensuring that δ13C, pCO2 and temperature data can be obtained at any time point. The pre-processed target curve can ensure the consistency of the difference calculation between the simulation results and the target value at any time in the subsequent calculation, thereby improving the accuracy and reliability of the inversion results.

[0032] After the target curve is selected and pre-processed, the parameter range required for inversion is determined and included in the parameter set to be inverted (denoted as φN). The setting of the initial parameter set φ0 takes into account the constraints of the target curve to ensure that the inversion results are consistent with the actual climate change trend.

[0033] The present invention introduces a time-series progressive model so that the inversion process can be dynamically updated at each time step, thereby improving the responsiveness and adaptability to time series. At the same time, by combining the Markov Chain Monte Carlo (MCMC) method with the time-series progressive model, it is possible to retain high-precision detail inversion while maintaining computational efficiency.

[0034] Although the various steps in the above embodiment are described in the above-mentioned order, those skilled in the art can understand that in order to achieve the effect of this embodiment, different steps do not have to be executed in such an order. They can be executed simultaneously (in parallel) or in a reverse order. These simple changes are within the scope of protection of the present invention.

[0035] Based on the above embodiment, the paleoclimate inversion model is obtained by iteratively matching the target curve of the paleoclimate index and the initial parameter set determined based on the target curve of the paleoclimate index to continuously update the parameter set to be inverted, such as Figure 2 As shown, the following steps S201-S204 are included: Step S201, based on the initial parameter set, a plurality of parameter sets to be inverted are randomly perturbed in a time-series progressive perturbation manner to generate a plurality of perturbation parameter sets as perturbation parameter sets; Step S202, obtaining a simulation value of a model output based on the initial parameter set and the disturbance parameter set; Step S203, obtaining a logarithmic likelihood or an error value based on the climate index, the simulation value and the target value of the target curve; Step S204, iteratively updating the parameter set to be inverted based on the log-likelihood or the error value to obtain a paleoclimate inversion model including a final inversion parameter set.

[0036] Specifically, when the error in the iterative process converges to the preset tolerance range, or reaches the specified maximum number of iterations, the simulation results of the model are considered to have effectively approached the target curve, and the final inversion results are output. The output parameter set contains the best estimate of the climate indicator and its uncertainty range, providing an accurate numerical description of the various characteristics of the target curve.

[0037] It should be noted that when the parameter output is completed, the inversion method of the present invention will record the optimal value and credible interval of each parameter, and generate the final fitted climate curve to intuitively display the fitting between the simulation results and the target curve. The output includes the capture results of global trends and local high-frequency fluctuation characteristics, providing a panoramic restoration of the climate change process. At the same time, the output results include complete parameter paths and disturbance records, which provide support for subsequent model verification and parameter sensitivity analysis, so that the inversion model has good reproducibility and interpretability.

[0038] Based on the above embodiment, the method of randomly perturbing the initial parameter set in a time-series progressive perturbation manner to generate a plurality of parameter sets to be inverted as perturbation parameter sets includes: Preset initial time and time window for multiple time periods; Based on a preset initial time and gradually expanding according to a preset time window, the parameter values ​​of the initial parameter set are randomly perturbed within a preset range in each time period to obtain the multiple parameter sets to be inverted as the perturbed parameter sets.

[0039] Specifically, starting from the initial time, this method advances step by step according to the set time window, and perturbs the parameters in a small range in each time period. This time-series progressive perturbation method is implemented through the Monte Carlo random generation method, which ensures that the parameter values ​​are randomly generated in each time window, enhancing the global exploration ability of the algorithm. In each time window, the generated parameter perturbations are independently randomly generated, avoiding the mutual influence between the parameters, thereby ensuring the full coverage of the parameter space. Ultimately, this process will generate N parameter sets to be inverted (denoted as φN), each of which corresponds to a different climate state and can effectively capture the uncertainty in the climate change process. At the same time, the perturbation process of this method can dynamically adjust the perturbation amplitude, gradually optimize the parameter range according to the degree of fit of the target curve, so that the generated parameters are more in line with the characteristics of the target curve, and improve the accuracy and stability of the inversion results.

[0040] It should be noted that there are different options for the time series progression mode. The most direct way is to progressively progress according to the minimum resolution of the target curve. Although this method can achieve the highest resolution fitting effect, it is computationally intensive and runs slowly. In order to improve computational efficiency, another way is to set up multiple combination schemes, allowing the time window to be gradually shortened. For example, in the initial stage, a larger time window is used for a rough global trend fitting, and then the window is gradually narrowed to capture finer high-frequency changes. This progressive progressive mode can achieve high-resolution fitting effects while maintaining computational efficiency, and can meet the accuracy requirements of different studies.

[0041] Based on the above embodiment, obtaining a simulation value of a model output based on the initial parameter set and the disturbance parameter set includes: The initial parameter set and the disturbance parameter set are respectively input into the paleoclimate inversion model to obtain corresponding simulation results under the initial state and the disturbance state.

[0042] Specifically, after completing the setting of the initial parameter set φ0 and the disturbance parameter set φN, the two sets of parameters are respectively put into the model to run and obtain the simulation results under the initial state and the disturbance state.

[0043] It should be noted that in the model execution part, different global biogeochemical models can be selected according to specific research needs to achieve higher applicability. For example, models related to the carbon cycle or climate simulation models can be selected to adapt to different geological events and target curve requirements. This flexible model selection method can ensure that the scope of application is expanded without changing the core algorithm framework, making this method widely used in paleoclimate inversion research under different climate events and chemical environments.

[0044] Based on the above embodiment, the paleoclimate indicator includes one or more of organic carbon isotopes, atmospheric carbon dioxide concentration and global temperature; It should be noted that the paleoclimate substitute indicators used can be adjusted and expanded according to research needs. In addition to the organic carbon isotopes, atmospheric carbon dioxide concentration and temperature selected by the present invention, multiple isotope indicators (such as carbon, mercury and zinc isotopes) can also be used in combination to achieve the inversion of more parameters. In principle, based on the principle of "several equations to solve several unknowns", adding substitute indicators can more comprehensively characterize the changes in the carbon cycle and climate system. For different geological events, such as climate change caused by volcanic eruptions, large-scale carbon emission events, etc., the use of different combinations of substitute indicators can flexibly adapt to different parameter solution requirements, thereby improving the inversion accuracy and applicability.

[0045] The logarithmic likelihood or error value is obtained based on the paleoclimate index, the simulation value and the target value of the target curve, including: When the paleoclimate indicator is organic carbon isotope, the logarithmic likelihood is calculated based on the simulation value and the target value of the target curve; Otherwise, the error value is calculated according to the simulation value and the target value of the target curve.

[0046] Specifically, in order to evaluate the match between the model results and the target curve, this method uses a combination of log-likelihood and error calculation. Specifically, log-likelihood calculation is used for δ13C data to capture its subtle changes; errors are directly calculated for pCO2 concentration and temperature data. The calculation formula is as follows: (1) (2) (3) (4) in, is the total error, for The calculation error of the data, for The calculation error of concentration is is the calculation error of temperature data, represents the log-likelihood, for Observed values, for The standard deviation represents the uncertainty or variability of the statistical distribution of observed data. Obtained for the model Simulation results, is the observed carbon dioxide concentration The carbon dioxide concentration is obtained for the model, is the observed sea surface temperature is the sea surface temperature obtained by the model, represents a normal distribution, Indicates selecting the maximum value. Indicates selecting the minimum value.

[0047] In this process, the error calculation methods for different target indicators vary due to different data characteristics and reliability. For δ13C data, as a key indicator of the carbon cycle, it has obtained high accuracy and reliability through geological records, and is suitable for log-likelihood calculation based on normal distribution. The use of normal distribution can make full use of the high-quality records of δ13C, thereby reflecting the true fluctuation characteristics of climate change. The log-likelihood provides an accurate fitting evaluation of the simulation results by calculating the probability density function value of the model output and the δ13C observation value.

[0048] It should be noted that in the error calculation, the weights of different target curves can be flexibly allocated according to the effect requirements of the inversion target curve. Different climate indicators may have different importance in the inversion, so the accurate fitting of specific indicators can be achieved by adjusting the weights of each indicator in the log-likelihood calculation. For example, in some events, temperature may be the main indicator of concern, and its weight can be increased accordingly, so as to more accurately control the temperature fitting effect. This weight distribution adjustment scheme enhances the flexibility and pertinence of this method, so that the fitting effect can be optimized according to specific research needs.

[0049] Based on the above embodiment, the error value is calculated according to the simulation value and the target value of the target curve, including: If the simulated value is within the confidence interval of the target curve observation data, the error value is set to zero; If the simulated value exceeds the confidence interval of the target curve observation data, the error value is calculated based on the mean deviation.

[0050] In contrast, the paleoclimate reconstruction of pCO2 concentration and temperature data is limited by the uncertainty of paleodata, and the accuracy of the observed values ​​is constrained by a variety of natural and technical factors. Therefore, for pCO2 concentration and temperature data, an error calculation method is used to evaluate their matching. Specifically, if the model output value is within the confidence interval of the observed data, the error is set to zero, and the model simulation value is considered to be consistent with the observation; if it exceeds the confidence interval, the error value is calculated based on the mean deviation to reflect the degree of deviation between the simulation value and the observed data. This method can reasonably evaluate the matching degree of pCO2 and temperature data within a rough range and avoid comparison bias caused by data uncertainty.

[0051] It should be noted that in this process, only the parameter set within the current time window is disturbed each time to ensure that the parameter adjustment is gradually optimized in the local range. However, when running the model and calculating the error, the evaluation process is global, that is, the entire simulation result will be calculated for global matching in each time period. As the time window slides gradually, the new parameter disturbance gradually covers the entire time period after the model runs, and finally achieves global optimization and evaluation. Through the combination of local disturbance and global matching calculation in this time window, the present invention ensures the accuracy of the global fitting of the model while realizing the sequential progressive disturbance.

[0052] Based on the above embodiment, the iterative updating of the parameter set to be inverted based on the logarithmic likelihood or the error value includes: Based on the error value, the total error ll0 of the initial parameter set and the errors ll1 to llN of the perturbation parameter set are calculated and compared: if lli>ll0, φ0 is replaced by the new parameter set φi to enter the next iteration; otherwise, the new candidate value is accepted based on the log-likelihood; wherein 1<=i<=N.

[0053] That is, in each iteration, the total error ll0 of the initial parameter set and the errors ll1 to llN of the perturbed parameter set are calculated and compared. If lli>ll0, the new parameter set φi replaces φ0 and enters the next iteration; otherwise, according to the Metropolis algorithm, the new candidate value is accepted with the probability of the likelihood ratio. This update mechanism ensures that the parameters gradually converge to the posterior distribution and obtain a better solution.

[0054] To improve adaptability, the present invention establishes a dynamic adjustment strategy: if a new parameter combination is rejected five times in a row, the average width and amplitude of the disturbance will be reduced by 10% and readjusted in the next time window to increase the probability of accepting the new proposed parameters. The initial time window is usually set to 20% of the time length of the target curve. The number of iterations is increased by gradually shortening the window, thereby improving the model's ability to capture and fit high-frequency information. This time-series progressive adjustment method ensures that the model gradually approaches the target curve based on the global trend, improving the accuracy of the inversion results.

[0055] In the early stage of inversion, the model adopts a low-frequency search strategy to capture the global trend of overall climate change with a longer time window. This process quickly identifies the macro characteristics of the target curve by establishing a preliminary fitting curve over a large time period, providing a reliable reference point for subsequent segment-by-segment fitting. As the inversion progresses, the time window is gradually shortened to capture subtle changes in the climate curve with more sophisticated high-frequency perturbation simulations. This progressive window shortening strategy allows the model to gradually transition from macro trends to detailed fitting, effectively capturing high-frequency fluctuation characteristics in climate records. Through segment-by-segment optimization of high-frequency perturbations, the model gradually reduces the degrees of freedom of the inversion parameters, enhances sensitivity to short-term climate fluctuations, and improves the stability of the simulation output, ultimately achieving convergence of the steady-state distribution.

[0056] The method of the present invention has significant application effects in the analysis of extreme carbon emission events in geological history, and is particularly suitable for simulation and parameter inversion of rapid responses of carbon cycle and climate system. Through accurate fitting of key climate indicators and high-frequency capture of time-by-time progression, this method can effectively solve different carbon emission scenarios, deduce possible emission mechanisms and environmental responses, and provide in-depth analysis tools for paleoclimate and paleoecological research.

[0057] For example, the dramatic carbon emission event in the Toarcian period of the Early Jurassic recorded a large negative shift of carbon isotopes (δ13C) of about 7‰, accompanied by a sharp rise in atmospheric carbon dioxide concentration and global temperature, which became one of the possible driving forces of biological extinction at that time. The carbon emission event in the Toarcian period had a large amplitude and rapid change, so a high-precision time series fitting method was needed to identify the carbon emission and climate change process during the event. This method selected δ13C, pCO2 and temperature as the target curves for inversion, and captured high-frequency changes by gradually shortening the time window, ensuring accurate simulation of carbon emission events in long time series.

[0058] like Figure 3As shown in the figure, the iterative process of δ13C data through Markov Chain Monte Carlo (MCMC) time-series progressive Bayesian inversion is given. In each iteration, the red curve gradually fits the black observed data points, indicating that the model parameters are constantly adjusted to gradually improve the fit to the data. The initial model is deliberately set as a straight line, demonstrating the algorithm's strong search and convergence capabilities, which are not affected by the initial value. Early iteration process: Although the initial model is quite different from the target, the algorithm quickly adjusts and searches for the general trend of the data. In the first few iterations, the red curve gradually begins to approach the changes in the black data points, indicating that the algorithm can quickly identify the main trend direction. Mid-term iteration process: As the number of iterations increases, the red curve gradually captures the general trend of δ13C, especially the V-shaped change characteristics, showing a deep fit to the data structure. At this point, the model has shown a high fitting accuracy. Late iteration process: After reaching 250 iterations, the model is further refined and gradually fits the subtle fluctuations of the data. By 350 iterations, the red curve almost perfectly fits the observed data, indicating the strong convergence performance of the algorithm.

[0059] In general, the algorithm converges quickly from the initial straight line, demonstrating its advantage of not relying on initial values, as well as its excellent performance in parameter search and fitting accuracy. This step-by-step optimization process demonstrates the effectiveness of Bayesian inversion in complex time series data.

[0060] like Figure 4 As shown in the figure, the MCMC Bayesian inversion has a high-precision fitting effect on the target curve in the converged state. The results show that the fitting quality of the target curve is good, and the key variables such as CO2 and methane release flux, isotopic composition, etc. solved by the inversion also show high resolution. These inversion results fully demonstrate the accuracy of the algorithm and its good ability to capture complex gas release processes.

[0061] In the simulation of the Toarcian carbon emission event, the parameter perturbations adjusted step by step allow the model to closely track the rapid negative excursion of carbon isotopes and accurately capture the dynamic changes of pCO2 and temperature associated with it, ensuring high consistency between the model output and the actual observed characteristics. The present invention can not only reproduce the significant negative excursion process of δ13C in the Toarcian period, but also reveal the high-frequency oscillation characteristics of rapid warming during the event through detailed simulation of temperature and pCO2. Through this method, the impact of carbon release rate, emission sources and different emission scenarios on the climate system can be inferred, providing a reliable basis for understanding the climate driving mechanism of the Toarcian period.

[0062] This method is also applicable to extreme climate events and carbon emission events in other geological historical periods, such as the mass extinction event at the end of the Permian period, the greenhouse climate of the Cretaceous period, the glacial-interglacial cycle of the Quaternary period, etc. The dramatic fluctuations in the carbon cycle and the rapid response of the climate in these events are important research topics in geology and climatology. The present invention can refine the complex climate change process in these events through flexible parameter adjustment and high-resolution inversion with piece-by-piece fitting, thereby achieving a comprehensive exploration of the climate-carbon cycle system under different driving mechanisms, providing valuable scientific reference for future climate change prediction and carbon management.

[0063] The paleoclimate inversion system based on Markov chain Monte Carlo in the second embodiment of the present invention is as follows: Figure 5 As shown, including: An index determination module 501 is used to determine the paleoclimate index to be inverted; The climate inversion module 502 is used to input the paleoclimate index to be inverted into the paleoclimate inversion model to obtain the paleoclimate inversion result output by the paleoclimate inversion model; The paleoclimate inversion model is obtained by iteratively matching a target curve of paleoclimate indicators and an initial parameter set determined based on the target curve of paleoclimate indicators to continuously update the parameter set to be inverted.

[0064] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related instructions of the system described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0065] It should be noted that the paleoclimate inversion system based on Markov chain Monte Carlo provided in the above embodiment is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the modules or steps in the embodiments of the present invention can be decomposed or combined. For example, the modules in the above embodiment can be combined into one module, or further divided into multiple sub-modules to complete all or part of the functions described above. The names of the modules and steps involved in the embodiments of the present invention are only for distinguishing the modules or steps, and are not regarded as improper limitations of the present invention.

[0066] An electronic device according to a third embodiment of the present invention includes: at least one processor; and a memory communicatively coupled to at least one of the processors; The memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the above-mentioned paleoclimate inversion method based on Markov chain Monte Carlo.

[0067] A fourth embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to be executed by the computer to implement the above-mentioned Markov Chain Monte Carlo-based paleoclimate inversion method.

[0068] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process and related instructions of the storage device and processing device described above can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0069] Those skilled in the art should be able to appreciate that the modules and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, computer software, or a combination of the two, and the programs corresponding to the software modules and method steps can be placed in random access memory (RAM), memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disks, removable disks, CD-ROMs, or any other form of storage medium known in the technical field. In order to clearly illustrate the interchangeability of electronic hardware and software, the composition and steps of each example have been generally described in the above description according to the function. Whether these functions are performed in electronic hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0070] Reference below Figure 6 , which shows a schematic diagram of the structure of a computer system of a server for implementing the method, system, and device embodiments of the present application. Figure 6 The server shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0071] like Figure 6 As shown, the computer system includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 602 or the program loaded from the storage part 608 to the random access memory (RAM) 603. In the RAM 603, various programs and data required for system operation are also stored. The CPU 601, ROM 602 and RAM 603 are connected to each other through a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0072] The following components are connected to the I / O interface 605: an input section 606 including a keyboard, a mouse, etc.; an output section 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I / O interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed so that a computer program read therefrom is installed into the storage section 608 as needed.

[0073] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 609, and / or installed from the removable medium 611. When the computer program is executed by the central processing unit (CPU) 601, the above-mentioned functions defined in the method of the present application are executed. It should be noted that the above-mentioned computer-readable medium of the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, - but not limited to - an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection with one or more conductors, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, an apparatus or a device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in combination with an instruction execution system, an apparatus or a device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0074] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0075] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0076] The terms "first", "second", etc. are used to distinguish similar objects rather than to describe or indicate a particular order or sequence.

[0077] The term "comprise" or any other similar term is intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus / device that includes a list of elements includes not only those elements but also other elements not expressly listed, or also includes elements inherent to such process, method, article, or apparatus / device.

[0078] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A paleoclimate inversion method based on Markov chain Monte Carlo, characterized in that: include: Determine the paleoclimate indicators to be inverted; Inputting the paleoclimate index to be inverted into the paleoclimate inversion model to obtain the paleoclimate inversion result output by the paleoclimate inversion model; The paleoclimate inversion model is obtained by iteratively matching a target curve of paleoclimate indicators and an initial parameter set determined based on the target curve of paleoclimate indicators to continuously update the parameter set to be inverted.

2. The paleoclimate inversion method based on Markov Chain Monte Carlo according to claim 1, characterized in that: The paleoclimate inversion model is obtained by iteratively matching the target curve of the paleoclimate index and the initial parameter set determined based on the target curve of the paleoclimate index to continuously update the parameter set to be inverted, and includes: Based on the initial parameter set, a plurality of parameter sets to be inverted are randomly perturbed in a time-series progressive perturbation manner to generate a plurality of perturbation parameter sets as perturbation parameter sets; Based on the initial parameter set and the disturbance parameter set, obtaining a simulated value of a model output; Obtaining a log likelihood or an error value based on the climate indicator, the simulation value, and a target value of the target curve; The parameter set to be inverted is iteratively updated based on the logarithmic likelihood or the error value to obtain a paleoclimate inversion model including a final inversion parameter set.

3. The paleoclimate inversion method based on Markov Chain Monte Carlo according to claim 2, characterized in that: The method of randomly perturbing the initial parameter set in a time-series progressive perturbation manner to generate a plurality of parameter sets to be inverted as perturbation parameter sets includes: Preset initial time and time window for multiple time periods; Based on a preset initial time and gradually expanding according to a preset time window, the parameter values ​​of the initial parameter set are randomly perturbed within a preset range in each time period to obtain the multiple parameter sets to be inverted as the perturbed parameter sets.

4. The paleoclimate inversion method based on Markov Chain Monte Carlo according to claim 2, characterized in that: The obtaining a simulated value of a model output based on the initial parameter set and the disturbance parameter set comprises: The initial parameter set and the disturbance parameter set are respectively input into the paleoclimate inversion model to obtain corresponding simulation results under the initial state and the disturbance state.

5. The paleoclimate inversion method based on Markov Chain Monte Carlo according to claim 2, characterized in that: The paleoclimate indicators include one or more of organic carbon isotopes, atmospheric carbon dioxide concentration, and global temperature; The logarithmic likelihood or error value is obtained based on the paleoclimate index, the simulation value and the target value of the target curve, including: When the paleoclimate indicator is organic carbon isotope, the logarithmic likelihood is calculated based on the simulation value and the target value of the target curve; Otherwise, the error value is calculated according to the simulation value and the target value of the target curve.

6. The paleoclimate inversion method based on Markov chain Monte Carlo according to claim 5, characterized in that: The calculating the error value according to the simulation value and the target value of the target curve includes: If the simulated value is within the confidence interval of the target curve observation data, the error value is set to zero; If the simulated value exceeds the confidence interval of the target curve observation data, the error value is calculated based on the mean deviation.

7. The paleoclimate inversion method based on Markov Chain Monte Carlo according to claim 2, characterized in that: The iterative updating of the parameter set to be inverted based on the log-likelihood or the error value comprises: Based on the error value, the total error ll0 of the initial parameter set and the errors ll1 to llN of the perturbation parameter set are calculated and compared: if lli > ll0, φ0 is replaced by the new parameter set φi to enter the next iteration; otherwise, the new candidate value is accepted with the log-likelihood; wherein 1<=i<=N.

8. A paleoclimate inversion system based on Markov chain Monte Carlo, characterized in that: include: An index determination module is used to determine the paleoclimate index to be inverted; A climate inversion module, used for inputting the paleoclimate index to be inverted into the paleoclimate inversion model to obtain the paleoclimate inversion result output by the paleoclimate inversion model; The paleoclimate inversion model is obtained by iteratively matching a target curve of paleoclimate indicators and an initial parameter set determined based on the target curve of paleoclimate indicators to continuously update the parameter set to be inverted.

9. An electronic device, characterized in that: include: at least one processor; and a memory communicatively coupled to at least one of the processors; Wherein, the memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the Markov Chain Monte Carlo based paleoclimate inversion method described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to be executed by the computer to implement the Markov Chain Monte Carlo-based paleoclimate inversion method described in any one of claims 1 to 7.

Images