Radionuclide release inversion method and system fusing four-dimensional variation and eulerian modes

Abstract

The application relates to a radionuclide emission inversion method and system fusing four-dimensional variation and Euler mode, and belongs to the technical field of nuclear pollution prediction. The method comprises the following steps: collecting multiple types of radionuclide observation data, high-temporal and high-spatial resolution meteorological prediction data and nuclear accident / nuclear leakage reports, and dividing the multiple types of radionuclides into different particle size sections; constructing an assimilation system based on four-dimensional variation theory and a nuclear pollution prediction Euler mode; the assimilation system is designed to be differentially adapted to the turbulence diffusion and dry deposition law of different particle size sections; the prior information of the nuclear accident / nuclear leakage report is input into the nuclear pollution prediction Euler mode to carry out prediction, the high-temporal and high-spatial resolution meteorological prediction data are used to constrain the output prediction result and drive the assimilation system, and the assimilation inversion is carried out in combination with the constraint of the observation data, so that the emission position, intensity and emission amount of the radionuclide per hour in the research period are obtained. The method can improve the accuracy of nuclear pollution prediction.

Description

Technical Field

[0001] This application relates to the field of nuclear pollution forecasting technology, and in particular to a method and system for inverting radionuclide emissions by integrating four-dimensional variational and Eulerian models. Background Technology

[0002] Accurately obtaining the emission levels of radionuclides is a prerequisite for precise forecasting and early warning of nuclear pollution. However, in real-world scenarios, the intensity of radionuclide emissions in most nuclear accidents is difficult to obtain directly, and in nuclear leak scenarios, the location of radionuclide emissions is even more unknown. These problems not only severely restrict the accuracy of nuclear pollution forecasting but also pose a significant challenge to the efficient implementation of nuclear safety emergency response.

[0003] Currently, the mainstream technical approach for estimating radionuclide emissions in nuclear accidents is based on a traditional pollution diffusion model conforming to a Gaussian distribution, combined with observational data. This involves acquiring data on wind speed, wind direction, and radionuclide concentrations around the nuclear accident site, solving the inverse problem of emission source strength and concentration distribution, and then estimating the radionuclide emissions. However, the Gaussian model is poorly adapted to radionuclide diffusion under complex meteorological conditions, leading to high uncertainty in the obtained emissions. In addition, a few studies have used Lagrange models to estimate particle diffusion distribution combined with radionuclide observational data to infer emissions. However, this approach suffers from a significant decrease in accuracy in areas far from the pollution source due to a reduction in the number of tracked particles. Furthermore, it suffers from coupling and compatibility defects with the Eulerian model used in current mainstream weather forecasting, failing to fully utilize key data such as meteorological wind fields, thus limiting improvements in inversion performance. Summary of the Invention

[0004] Therefore, it is necessary to provide a radionuclide emission inversion method and system that integrates four-dimensional variational and Eulerian models to address the above-mentioned technical problems. This method and system can quickly and efficiently estimate the hourly emission location, emission intensity, and emission amount of radionuclides. Furthermore, it can be used for nuclear pollution forecasting and early warning model research, accurately improving the accuracy of nuclear pollution forecasting and overcoming the shortcomings of existing traditional methods.

[0005] A method for inverting radionuclide emissions by integrating four-dimensional variational and Eulerian models, the method comprising:

[0006] Collect observational data of multiple types of radionuclides after a nuclear accident, high spatiotemporal resolution meteorological forecast data, and nuclear accident / nuclear leak reports, and classify multiple types of radionuclides into fine-grained and coarse-grained radionuclides.

[0007] A four-dimensional variational objective functional for the radionuclide emission source term is constructed. Using the Eulerian model for nuclear pollution forecasting, a forward model operator is constructed in the four-dimensional variational objective functional to describe the evolution of radionuclides in the atmosphere. Based on the four-dimensional variational objective functional containing the forward model operator, an assimilation system for radionuclide emission intensity and emission location is built. In the forward model operator, the turbulent diffusion term corresponding to the two particle size ranges adopts a differentiated turbulent diffusion coefficient, and the corresponding dry deposition term adopts a differentiated dry deposition velocity.

[0008] The location, time, and amount of radionuclide emissions recorded and estimated in nuclear accident / nuclear leak reports are used as prior information. This prior information is then input into the Eulerian model for nuclear pollution forecasting to conduct simulation forecasts. Under the constraints of high spatiotemporal resolution meteorological forecast data, the simulation outputs continuous meteorological fields and radionuclide pollution concentration fields during the study period as driving data.

[0009] The driving data is input into the assimilation system, and assimilation inversion is performed under the constraints of observation data of multiple types of radionuclides to obtain the hourly radionucde emission location, emission intensity and emission amount during the study period.

[0010] In one embodiment, a four-dimensional variational objective functional of the radionuclide emission source term is constructed, expressed as:

[0011] ;

[0012] in, The objective function value; This is a state variable, representing the hourly radionuclide emissions at the grid point where the nuclear accident occurred or the grid point where a potential nuclear leak area is located. This refers to the prior radionuclide emissions derived from nuclear accident reports; when the location of the nuclear accident is identified, and Both are one-dimensional vectors; however, in the case of nuclear leakage, and All are three-dimensional vectors, containing grid regions representing potential nuclear leaks and the time of nuclear leaks; Indicates time; The duration of time during which radioactive nuclides are released into the atmosphere; The background error covariance of radionuclide emissions, its dimension is... Consistent, but the vector dimension may differ under different circumstances; It is the observation vector; The assimilation time window is defined as the time window within which observational data of multiple types of radionuclides are used as constraints in the assimilation and inversion process. The observation error covariance is in diagonal matrix form, and the observation error covariance differs for different types of radionuclide observation data. This is a forward model operator constructed using the Eulerian model for nuclear contamination forecasting; For observation operators, used to... The described radionuclide contamination concentration is converted into an observation operator identical to the radionuclide observation data type, while spatial matching between observation sites and model formats is performed; superscript This represents the transpose of a vector.

[0013] In one embodiment, the forward model operator constructed using the Eulerian model for nuclear contamination prediction is expressed as:

[0014] ;

[0015] in, For the first The concentration of radioactive nuclide contamination at time t, with subscript s and l These represent fine-particle-size nuclides and coarse-particle-size nuclides, respectively, and the total concentration of the two types of nuclides is taken as the radionuclide contamination concentration. This refers to the advection diffusion term for nuclides with a small particle size. This refers to the advection diffusion term for nuclides with coarse particle sizes. Wind speed in the horizontal direction. Wind speed in the vertical direction; For the turbulent diffusion term of fine-particle-size nuclides, The turbulent diffusion terms are for coarse-grained nuclides, and the turbulent diffusion terms for both types of nuclides are much smaller than the corresponding advection diffusion terms. The turbulent diffusion coefficients represent the horizontal and vertical directions for two types of nuclides with different particle sizes. To address the differences in the motion characteristics of the two types of nuclides, a differentiated adaptation is used for the turbulent diffusion coefficients. Specifically, for the finer-particle-size nuclides, a dynamic turbulent diffusion coefficient in the horizontal direction, corrected in real-time based on the intensity of turbulent fluctuations, is employed. and The coefficient is dynamically adjusted to follow changes in atmospheric turbulence to improve the accuracy of fine particle diffusion simulation; for coarse-particle nuclides, the static turbulent diffusion coefficient in the horizontal direction coupled with gravity settling effect is used. and It is adapted to the motion characteristics dominated by the gravity sedimentation of coarse particles and meets the requirements. That is, the proportion of turbulent diffusion terms is higher for fine-grained nuclides than for coarse-grained nuclides; These represent the dry precipitation terms for fine-grained and coarse-grained nuclides, respectively. Representing the dry sedimentation velocities of fine-grained and coarse-grained nuclides respectively, and satisfying... Among them, the dry deposition velocity model of fine-particle-size nuclides is constructed based on the gas-particle exchange coefficient and surface adsorption efficiency, which is suitable for their characteristics of being easily adsorbed by atmospheric particulate matter and having frequent gas-particle exchange; the dry deposition velocity model of coarse-particle-size nuclides is constructed based on the coupling of gravity deposition velocity and ground roughness, which characterizes their characteristics of being dominated by gravity deposition and significantly affected by ground resistance. For decay terms, The decay rate of a radioactive nuclide; For emissions, this represents the increase in the concentration of radionuclides in the atmosphere caused by the emission of radionuclides. It is the actual air density. The model layer height corresponding to the height of the emission source. and The model is a dynamic allocation model for total emissions among fine-grained and coarse-grained nuclides. Based on the historical statistical distribution characteristics of measured particle size distribution of emission sources, the total emissions are dynamically allocated according to the proportion of the two types of nuclides. The allocation coefficient is updated according to real-time measured data to adapt to the differences in particle size distribution of different emission sources. and In the horizontal direction, It is in the vertical direction.

[0016] In one embodiment, the method further includes, when constructing the four-dimensional variational objective functional of the radionuclide emission source term:

[0017] Based on prior radionuclide emissions and considering the uncertainty of state variables, the position and height of state variables are dynamically adjusted to calculate the background error covariance of prior radionuclide emissions to meet the requirements of four-dimensional variational assimilation. Furthermore, considering the heterogeneity of observation data for various types of radionuclides, the observation error covariance of each type of radionuclide observation data is calculated, and specific observation operators for each type of radionuclide observation data are designed to adapt to the four-dimensional variational assimilation model. Specifically, for cases where the location of emission sources in nuclear accident scenarios is determined, the location of state variables is considered fixed, and only the emission intensity is inverted and assimilated. For cases where the location of radionuclides in nuclear leak scenarios is uncertain, the potential emission location area is dynamically adjusted, and all radionuclides within the grid of that area are treated as state variables.

[0018] In one embodiment, a dedicated observation operator is designed specifically for adapting four-dimensional variational assimilation modes to various types of radionuclide observation data, including:

[0019] The observation data for various types of radionuclides include specific activity observation data, sedimentation rate observation data, and dose rate observation data. Among them, the specific activity observation operator corresponding to the specific activity observation data is composed of specific activity observation data unit conversion and grid point interpolation, which is a linear calculation problem. For the sedimentation rate observation operator corresponding to the sedimentation rate observation data and the dose rate observation operator corresponding to the dose rate observation data, the correlation between sedimentation rate, dose rate and radionuclide emission intensity is statistically analyzed based on the forecast results output by the Eulerian model for nuclear pollution forecasting, and a lookup table is constructed. Then, the corresponding observation operators and their accompanying programs are provided through the lookup table for direct invocation.

[0020] In one embodiment, when inputting prior information into the Eulerian model for nuclear contamination prediction to conduct simulation prediction, the method further includes:

[0021] In the simulation and forecasting process of the Eulerian model for nuclear pollution forecasting, the advection diffusion term in the forward model operator is calculated using the upwind difference scheme. The turbulent diffusion coefficient and dry deposition velocity of the two types of nuclides with different sizes are updated in the forward model operator. In the integration process of the Eulerian model for nuclear pollution forecasting, the concentration gradient diffusion equation is solved spatiotemporally using an implicit difference scheme.

[0022] In one embodiment, updating the turbulent diffusion coefficient and dry sedimentation velocity for differentiated adaptation of two types of nuclides in the forward mode operator includes:

[0023] The update method for the turbulent diffusion coefficient of fine-grained nuclides is as follows:

[0024] ;

[0025] ;

[0026] in, These are the turbulent diffusion coefficients of fine-particle-size nuclides in the horizontal and vertical directions, respectively. The turbulence constant is It is atmospheric turbulent kinetic energy; This is the turbulent energy dissipation rate, used to reflect the turbulence decay characteristics; This is a correction function for the turbulent diffusion coefficient, and relates to the adsorption efficiency of the fine-particle-size nuclide surface. It is also related to the exchange coefficient of nuclides in the fine-particle size range. related, Used to improve the efficiency of mass exchange between fine-particle-size nuclides and the atmosphere;

[0027] The update method for the turbulent diffusion coefficient of coarse-grained nuclides is as follows:

[0028] ;

[0029] ;

[0030] in, These represent the turbulent diffusion coefficients of coarse-grained nuclides in the horizontal and vertical directions, respectively. It is the gravitational settling velocity of coarse-grained nuclides. Relevant correction functions; The mixing length is a characteristic length that characterizes turbulent motion and is related to atmospheric stratification and the roughness of the underlying surface.

[0031] The update method for the dry sedimentation rate of the two types of nuclides is as follows:

[0032] ;

[0033] in, For dry settlement velocity, subscript Representing fine-grained and coarse-grained nuclides, It is the gravitational settling velocity. It is aerodynamic drag, which is positively correlated with the roughness of the underlying surface; It is the sublayer drag of laminar flow, which reflects the magnitude of the resistance to particles passing through the atmospheric laminar sublayer. It is the surface resistance, set as a lookup table according to the underlying surface type, used to correct for dry settling velocity.

[0034] In one embodiment, the forward model operator and the driving data are both constructed and calculated by the Eulerian model for nuclear contamination prediction in the Eulerian coordinate system. The two are fully coupled to achieve matching based on the same coordinate system, and the assimilation and inversion of radionuclides are driven by the fully coupled structure.

[0035] In one embodiment, after collecting observational data on multiple types of radionuclides following a nuclear accident, the method further includes:

[0036] Preprocessing is performed on observation data of various types of radionuclides. The preprocessing methods include background data analysis, threshold check analysis, spatiotemporal consistency analysis, and cross-validation analysis.

[0037] A radionuclide emission inversion system integrating four-dimensional variational and Eulerian models, the system comprising:

[0038] The data acquisition module is used to collect observation data of multiple types of radionuclides, high spatiotemporal resolution meteorological forecast data and nuclear accident / nuclear leakage reports after a nuclear accident, and to classify multiple types of radionuclides into fine-grained and coarse-grained nuclides.

[0039] The assimilation system construction module is used to construct a four-dimensional variational objective functional for the radionuclide emission source term. Using the Eulerian model for nuclear pollution forecasting, it constructs a forward model operator in the four-dimensional variational objective functional to describe the evolution of radionuclides in the atmosphere. Based on the four-dimensional variational objective functional containing the forward model operator, an assimilation system for the emission intensity and location of radionuclides is built. In the forward model operator, the turbulent diffusion term corresponding to the two particle size ranges adopts a differentiated turbulent diffusion coefficient, and the corresponding dry deposition term adopts a differentiated dry deposition velocity.

[0040] The prior forecast module is used to take the location, time and amount of radionuclide emission recorded and estimated in the nuclear accident / nuclear leak report as prior information, input the prior information into the Eulerian model of nuclear pollution forecast to carry out simulation forecast, and under the constraint of high spatiotemporal resolution meteorological forecast data, simulate and output the continuous meteorological field and radionuclide pollution concentration field during the study period as driving data.

[0041] The emission inversion module is used to input the driving data into the assimilation system and perform assimilation inversion under the constraints of observation data of multiple types of radionuclides to obtain the hourly emission location, emission intensity and emission amount of radionuclides during the study period.

[0042] The aforementioned method and system for inverting radionuclide emissions by integrating four-dimensional variational and Eulerian models offer the following advantages over existing technologies: First, relying on the Eulerian model for nuclear pollution forecasting, it can more fully and precisely utilize the high spatiotemporal resolution meteorological forecast data output by the Eulerian model, without being limited by complex meteorological conditions, topographical differences, or transmission distances. Second, by integrating a four-dimensional variational objective functional constructed based on four-dimensional variational theory and a forward model operator constructed based on the Eulerian model for nuclear pollution forecasting, it can significantly improve the simulation accuracy and forecast timeliness of the nuclear pollution diffusion and evolution process, thereby enabling accurate inversion of the hourly radionuclide emission location, intensity, and emission rate within the research period. First, it has the advantages of high throughput, fast computing speed, and small storage space requirement, making it suitable for the high-efficiency needs of emergency response. Second, by dividing different particle size segments of various types of radionuclides and designing differentiated turbulent diffusion coefficients and dry deposition velocities for coarse and fine particle size segments in the forward model operator, it can more accurately reflect the turbulent diffusion and dry deposition laws of different particle size nuclides in the atmosphere, improving the accuracy and applicability of radionuclide diffusion simulation. Third, the radionuclide emission data obtained by inversion can be directly applied to nuclear pollution forecasting and early warning work, which has important practical application value for improving the accuracy of nuclear pollution forecasting, strengthening nuclear accident emergency support capabilities, and ensuring nuclear safety. Attached Figure Description

[0043] Figure 1This is a flowchart illustrating a radionuclide emission inversion method that integrates four-dimensional variational and Eulerian models in one embodiment.

[0044] Figure 2 This is a schematic diagram illustrating the actual and spurious locations of radionuclide emissions in one embodiment.

[0045] Figure 3 This is a schematic diagram of the intensity of cesium emission sources in a nuclear accident, both prior and derived from the inversion in this application, in one embodiment.

[0046] Figure 4 This is a schematic diagram illustrating the efficiency reduction of the cost function of a four-dimensional variational objective functional as the number of iterations converges in one embodiment. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0048] In one embodiment, such as Figure 1 As shown, a method for inverting radionuclide emissions by integrating four-dimensional variational and Eulerian models is provided, including the following steps:

[0049] Step 1: Collect observation data of multiple types of radionuclides, high spatiotemporal resolution meteorological forecast data, and nuclear accident / nuclear leak reports after the nuclear accident, and classify the multiple types of radionuclides into fine-grained and coarse-grained radionuclides.

[0050] Step 2: Construct a four-dimensional variational objective functional for the radionuclide emission source term, and use the Eulerian model for nuclear pollution forecasting to construct a forward model operator in the four-dimensional variational objective functional to describe the evolution of radionuclides in the atmosphere. Based on the four-dimensional variational objective functional containing the forward model operator, construct an assimilation system for radionuclide emission intensity and emission location. In the forward model operator, the turbulent diffusion term corresponding to the two particle size ranges adopts a differentiated turbulent diffusion coefficient, and the corresponding dry deposition term adopts a differentiated dry deposition velocity.

[0051] Step 3: The location, time, and amount of radionuclide emissions recorded and estimated in the nuclear accident / nuclear leak report are used as prior information. The prior information is input into the Eulerian model for nuclear pollution forecasting to carry out simulation forecasting. Under the constraints of high spatiotemporal resolution meteorological forecast data, the continuous meteorological field and radionuclide pollution concentration field during the study period are simulated and output as driving data.

[0052] Step 4: Input the driving data into the assimilation system and perform assimilation inversion under the constraints of observation data of multiple types of radionuclides to obtain the hourly radionucde emission location, emission intensity and emission amount during the study period.

[0053] The aforementioned radionuclide emission inversion method, which integrates four-dimensional variational theory and the Eulerian model, constructs a framework for solving the inverse problem of emission source-concentration inversion by combining four-dimensional variational theory with the Eulerian model for nuclear pollution forecasting. This allows for precise source tracing of radionuclide emissions by solving the inverse problem of radionuclide emission source and concentration inversion. It can rapidly and efficiently estimate the hourly emission location, intensity, and amount of radionuclides, and can be used in nuclear pollution forecasting and early warning model research, significantly improving the accuracy of nuclear pollution forecasts. Furthermore, by segmenting multiple types of radionuclides by particle size and setting appropriate turbulent diffusion coefficients and dry deposition velocities for coarse and fine particle size segments, the method can more accurately simulate the atmospheric turbulent diffusion and dry deposition processes of radionuclides of different particle sizes, improving the accuracy and reliability of radionuclide diffusion simulation.

[0054] In one embodiment, a four-dimensional variational objective functional of the radionuclide emission source term is constructed, expressed as:

[0055] ;

[0056] in, The objective function value; This is a state variable, representing the hourly radionuclide emissions at the grid point where the nuclear accident occurred or the grid point where a potential nuclear leak area is located. For prior radionuclide emissions derived from nuclear accident reports, nuclear accident / leakage reports specifically include reports on the source term intensity and leakage process corresponding to the nuclear accident / leakage; when the location of the nuclear accident is identified, and Both are one-dimensional vectors; however, in the case of nuclear leakage, and All are three-dimensional vectors, containing grid regions representing potential nuclear leaks and the time of nuclear leaks; Indicates time; The duration of time during which radioactive nuclides are released into the atmosphere; The background error covariance of radionuclide emissions, its dimension is... Consistent, but the vector dimension may differ under different circumstances; It is the observation vector; The assimilation time window is defined as the time window within which observational data of multiple types of radionuclides are used as constraints in the assimilation and inversion process. The observation error covariance is in diagonal matrix form. The observation error covariance varies for different types of radionuclide observation data. Setting the observation error covariance in diagonal matrix form is beneficial for balancing calculation speed and calculation accuracy. This is a forward model operator constructed using the Eulerian model for nuclear pollution forecasting, used to describe processes such as horizontal diffusion, turbulent diffusion, dry / wet deposition, decay, and emission of radionuclides in the atmosphere. For observation operators, used to... The described radionuclide contamination concentration is converted into an observation operator identical to the radionuclide observation data type, while spatial matching between observation sites and model formats is performed; superscript This represents the transpose of a vector.

[0057] It should be understood that this four-dimensional variational objective functional can integrate all observational data within a specific time window to achieve synergistic constraint and inversion of radionuclide emission data and the location of nuclear accident pollution sources.

[0058] In one embodiment, the forward model operator constructed using the Eulerian model for nuclear contamination prediction is expressed as:

[0059] ;

[0060] in, For the first The concentration of radioactive nuclide contamination at time t, with subscript s and l These represent fine-particle-size nuclides and coarse-particle-size nuclides, respectively. The total concentration of the two types of nuclides is taken as the radionuclide contamination concentration. Specifically, fine-particle-size nuclides represent particles with a diameter of 0-2.5 µm, and coarse-particle-size nuclides represent particles with a diameter of 2.5-10.0 µm. This refers to the advection diffusion term for nuclides with a small particle size. This is the advection diffusion term for nuclides with a coarse particle size. Wind speed in the horizontal direction. Wind speed in the vertical direction; For the turbulent diffusion term of fine-particle-size nuclides, The turbulent diffusion terms are for coarse-grained nuclides, and the turbulent diffusion terms for both types of nuclides are much smaller than the corresponding advection diffusion terms. The turbulent diffusion coefficients represent the horizontal and vertical directions for two types of nuclides with different particle sizes. To address the differences in the motion characteristics of the two types of nuclides, a differentiated adaptation is used for the turbulent diffusion coefficients. Specifically, for the finer-particle-size nuclides, a dynamic turbulent diffusion coefficient in the horizontal direction, corrected in real-time based on the intensity of turbulent fluctuations, is employed. and The coefficient is dynamically adjusted to follow changes in atmospheric turbulence to improve the accuracy of fine particle diffusion simulation; for coarse-particle nuclides, the static turbulent diffusion coefficient in the horizontal direction coupled with gravity settling effect is used. and It is adapted to the motion characteristics dominated by the gravity sedimentation of coarse particles and meets the requirements. That is, the proportion of turbulent diffusion term is higher for fine-grained nuclides than for coarse-grained nuclides. It should be noted that the turbulent diffusion term of coarse-grained nuclides is negligible due to gravity sedimentation, while the turbulent diffusion term of fine-grained nuclides is much smaller than that of their advection diffusion term, but still retains a dynamic correction mechanism. These represent the dry precipitation terms for fine-grained and coarse-grained nuclides, respectively. Representing the dry sedimentation velocities of fine-grained and coarse-grained nuclides respectively, and satisfying... Among them, a dry deposition velocity model was constructed for fine-particle-size nuclides based on the gas-particle exchange coefficient and surface adsorption efficiency, which is adapted to their characteristics of being easily adsorbed by atmospheric particulate matter and having frequent gas-particle exchange. The value range is 0.01-0.1 m / s; for coarse-grained nuclides, a dry settlement velocity model is constructed based on the coupling of gravity settlement velocity and ground roughness to characterize its characteristics of being dominated by gravity settlement and significantly affected by ground resistance. The value range is 0.1-1.0 m / s; For decay terms, The decay rate of radioactive nuclides is relatively small compared to other physical processes, so the decay rate of particles of different sizes is no longer distinguished. For emissions, this represents the increase in the concentration of radionuclides in the atmosphere caused by the emission of radionuclides. It is the actual air density. The model layer height corresponding to the height of the emission source. and The model is a dynamic allocation model for total emissions among fine-grained and coarse-grained nuclides. Based on the historical statistical distribution characteristics of measured particle size distribution of emission sources, the total emissions are dynamically allocated according to the proportion of the two types of nuclides. The allocation coefficient is updated according to real-time measured data to adapt to the differences in particle size distribution of different emission sources. and In the horizontal direction, It is in the vertical direction.

[0061] In one embodiment, the method further includes, when constructing the four-dimensional variational objective functional of the radionuclide emission source term:

[0062] Based on prior radionuclide emissions and considering the uncertainty of state variables, the position and height of state variables are dynamically adjusted to calculate the background error covariance of prior radionuclide emissions to meet the requirements of four-dimensional variational assimilation. Furthermore, considering the heterogeneity of observation data for various types of radionuclides, the observation error covariance of each type of radionuclide observation data is calculated, and specific observation operators for each type of radionuclide observation data are designed to adapt to the four-dimensional variational assimilation model. Specifically, for cases where the location of emission sources in nuclear accident scenarios is determined, the location of state variables is considered fixed, and only the emission intensity is inverted and assimilated. For cases where the location of radionuclides in nuclear leak scenarios is uncertain, the potential emission location area is dynamically adjusted, and all radionuclides within the grid of that area are treated as state variables.

[0063] In one embodiment, a dedicated observation operator is designed specifically for adapting four-dimensional variational assimilation modes to various types of radionuclide observation data, including:

[0064] The observation data for various types of radionuclides include specific activity observation data, sedimentation rate observation data, and dose rate observation data. Among them, the specific activity observation operator corresponding to the specific activity observation data is composed of specific activity observation data unit conversion and grid point interpolation, which is a linear calculation problem. For the sedimentation rate observation operator corresponding to the sedimentation rate observation data and the dose rate observation operator corresponding to the dose rate observation data, the correlation between sedimentation rate, dose rate and radionuclide emission intensity is statistically analyzed based on the forecast results output by the Eulerian model for nuclear pollution forecasting, and a lookup table is constructed. Then, the corresponding observation operators and their accompanying programs are provided through the lookup table for direct invocation.

[0065] In one embodiment, when inputting prior information into the Eulerian model for nuclear contamination prediction to conduct simulation prediction, the method further includes:

[0066] In the simulation and forecasting process of the Eulerian model for nuclear pollution forecasting, the advection diffusion term in the forward model operator is calculated using an upwind difference scheme. This updates the turbulent diffusion coefficients and dry deposition velocities of the two particle size ranges in the forward model operator. Furthermore, during the integration process of the Eulerian model for nuclear pollution forecasting, an implicit difference scheme is used to solve the concentration gradient diffusion equation spatiotemporally. The prior information input into the Eulerian model for nuclear pollution forecasting incorporates the aforementioned differential emission allocation logic for fine-particle and coarse-particle nuclides. Simultaneously, high spatiotemporal resolution meteorological forecast data is used as a constraint, and the meteorological boundary conditions of the study area are dynamically constrained every 6 hours to ensure the accuracy of the Eulerian model for nuclear pollution forecasting.

[0067] In one embodiment, updating the turbulent diffusion coefficient and dry sedimentation velocity for differentiated adaptation of two types of nuclides in the forward mode operator includes:

[0068] The update method for the turbulent diffusion coefficient of fine-grained nuclides is as follows:

[0069] ;

[0070] ;

[0071] in, These are the turbulent diffusion coefficients of fine-particle-size nuclides in the horizontal and vertical directions, respectively. The turbulence constant is It is atmospheric turbulent kinetic energy; This is the turbulent energy dissipation rate, used to reflect the turbulence decay characteristics; This is a correction function for the turbulent diffusion coefficient, and relates to the adsorption efficiency of the fine-particle-size nuclide surface. It is also related to the exchange coefficient of nuclides in the fine-particle size range. related, Used to improve the efficiency of mass exchange between fine-particle-size nuclides and the atmosphere;

[0072] The update method for the turbulent diffusion coefficient of coarse-grained nuclides is as follows:

[0073] ;

[0074] ;

[0075] in, These represent the turbulent diffusion coefficients of coarse-grained nuclides in the horizontal and vertical directions, respectively. It is the gravitational settling velocity of coarse-grained nuclides. Relevant correction functions; The mixing length is a characteristic length that characterizes turbulent motion and is related to atmospheric stratification and the roughness of the underlying surface.

[0076] The update method for the dry sedimentation rate of the two types of nuclides is as follows:

[0077] ;

[0078] in, For dry settlement velocity, subscript Representing fine-grained and coarse-grained nuclides, It is the gravitational settling velocity. It is aerodynamic drag, which is positively correlated with the roughness of the underlying surface; It is the sublayer drag of laminar flow, which reflects the magnitude of the resistance to particles passing through the atmospheric laminar sublayer. It is the surface resistance, set as a lookup table according to the underlying surface type, used to correct for dry settling velocity.

[0079] In one embodiment, the forward model operator and the driving data are both constructed and calculated by the Eulerian model for nuclear contamination prediction in the Eulerian coordinate system. The two are fully coupled to achieve matching based on the same coordinate system, and the assimilation and inversion of radionuclides are driven by the fully coupled structure.

[0080] In one embodiment, after collecting observational data on multiple types of radionuclides following a nuclear accident, the method further includes:

[0081] Preprocessing is performed on observation data of various types of radionuclides. The preprocessing methods include background data analysis, threshold check analysis, spatiotemporal consistency analysis, and cross-validation analysis.

[0082] Furthermore, to verify the performance of the method proposed in this application, the following detailed explanation uses a specific application of this method in a particular scenario as an example. Taking the nuclear contamination diffusion process in a certain area from 06:00 on March 15, 2011 to 00:00 on March 16, 2011 as an example, the emissions announced after the nuclear leak accident are taken as the actual emissions, and the spatial resolution of the study area is 3 km. The specific steps include the following:

[0083] 1. Collect observational data of various types of radionuclides following a nuclear accident in a certain area (including radionuclide air specific activity data, ground subsidence data, air dose rate data, and observational data from international radiation stations, and conduct background data analysis, threshold check analysis, spatiotemporal consistency analysis, cross-validation analysis, etc.), and high spatiotemporal resolution meteorological forecast data (every three hours, resolution of [missing information]). The data includes meteorological forecasts and reports of nuclear accidents / leaks, and classifies various types of radionuclides into fine-particle-size nuclides (0-2.5 µm) and coarse-particle-size nuclides (2.5-10.0 µm).

[0084] 2. Construct a four-dimensional variational objective functional for the radionuclide emission source term, and use the Eulerian model for nuclear pollution forecasting to construct a forward model operator in the four-dimensional variational objective functional to describe the evolution of radionuclides in the atmosphere. Based on the four-dimensional variational objective functional containing the forward model operator, construct an assimilation system for radionuclide emission intensity and emission location. In the forward model operator, the turbulent diffusion term corresponding to the two particle size ranges adopts a differentiated turbulent diffusion coefficient, and the corresponding dry deposition term adopts a differentiated dry deposition velocity.

[0085] 3. The location, time, and amount of radionuclide emissions recorded and estimated in nuclear accident / nuclear leak reports are used as prior information. This prior information is then input into the Eulerian model for nuclear pollution forecasting to conduct simulation forecasts. Under the constraints of high spatiotemporal resolution meteorological forecast data, the continuous meteorological field and radionuclide pollution concentration field during the study period are simulated and output as driving data.

[0086] 4. Input the driving data into the assimilation system and perform assimilation inversion under the constraints of observation data of multiple types of radionuclides to obtain the hourly radionucde emission location, emission intensity and emission amount during the study period.

[0087] In this embodiment, cesium in a nuclear accident is used as an example. Figure 2 The actual and spurious locations of nuclide emissions are shown, represented by hollow triangles and solid dots, respectively. It can be seen that the distance between the actual and spurious emission locations in this embodiment exceeds 12 km, thereby testing the applicability of this dynamic inversion nuclide emission assimilation system.

[0088] Figure 3 The paper demonstrates the intensity of cesium emission sources in a nuclear accident, both a priori and derived from the inversion in this application. Figure 3 It can be seen that the assimilation system for radionuclide emission intensity and emission location constructed in this application can basically correct the location of the radionuclide emission source, with an inversion accuracy of kilometers. Even if the error between false emission and true emission is more than ten kilometers, it can be corrected well, and the radionuclide emission source is inverted at the correct location. The radionuclide emission rate and emission amount have good consistency with the true results, while reducing the false emission amount at the wrong location.

[0089] Figure 4 This demonstrates the efficiency of the cost function of the four-dimensional variational objective functional constructed in this application as it converges with the number of iterations. Figure 4 It can be seen that as the number of iterations increases, the cost function value of the four-dimensional variational objective functional shows a decreasing trend. When the number of iterations is 6, the function value approaches 0. This shows that the four-dimensional variational objective functional constructed in this application has a fast convergence speed and reliable calculation results in the nuclear pollution prediction scenario.

[0090] In summary, this method can effectively address the issue of tracing the source of radioactive nuclides emitted into the atmosphere due to nuclear accidents, and accurate estimation of nuclide emissions is of significant scientific importance. This method is simple and easy to implement, requires little computation and funding, and the obtained nuclide emission data can also be used for nuclear pollution forecasting and early warning research, improving the accuracy of nuclear pollution forecasts and playing an important role in formulating nuclear safety emergency response measures.

[0091] In one embodiment, a radionuclide emission inversion system integrating four-dimensional variational and Eulerian models is provided, comprising:

[0092] The data acquisition module is used to collect observation data of multiple types of radionuclides, high spatiotemporal resolution meteorological forecast data and nuclear accident / nuclear leakage reports after a nuclear accident, and to classify multiple types of radionuclides into fine-grained and coarse-grained nuclides.

[0093] The assimilation system construction module is used to construct a four-dimensional variational objective functional for the radionuclide emission source term. Using the Eulerian model for nuclear pollution forecasting, it constructs a forward model operator in the four-dimensional variational objective functional to describe the evolution of radionuclides in the atmosphere. Based on the four-dimensional variational objective functional containing the forward model operator, an assimilation system for the emission intensity and location of radionuclides is built. In the forward model operator, the turbulent diffusion term corresponding to the two particle size ranges adopts a differentiated turbulent diffusion coefficient, and the corresponding dry deposition term adopts a differentiated dry deposition velocity.

[0094] The prior forecast module is used to take the location, time and amount of radionuclide emission recorded and estimated in the nuclear accident / nuclear leak report as prior information, input the prior information into the Eulerian model of nuclear pollution forecast to carry out simulation forecast, and under the constraint of high spatiotemporal resolution meteorological forecast data, simulate and output the continuous meteorological field and radionuclide pollution concentration field during the study period as driving data.

[0095] The emission inversion module is used to input the driving data into the assimilation system and perform assimilation inversion under the constraints of observation data of multiple types of radionuclides to obtain the hourly emission location, emission intensity and emission amount of radionuclides during the study period.

[0096] Specific limitations regarding the radionuclide emission inversion system integrating four-dimensional variational and Eulerian models can be found in the limitations of the radionuclide emission inversion method integrating four-dimensional variational and Eulerian models mentioned above, and will not be repeated here. Each module in the aforementioned radionuclide emission inversion system integrating four-dimensional variational and Eulerian models can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0097] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0098] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application.

Claims

1. A method for inverting radionuclide emissions by integrating four-dimensional variational and Eulerian models, characterized in that, The method includes: Collect observational data of multiple types of radionuclides after a nuclear accident, high spatiotemporal resolution meteorological forecast data, and nuclear accident / nuclear leak reports, and classify multiple types of radionuclides into fine-grained and coarse-grained radionuclides. A four-dimensional variational objective functional for the radionuclide emission source term is constructed. Using the Eulerian model for nuclear pollution forecasting, a forward model operator is constructed in the four-dimensional variational objective functional to describe the evolution of radionuclides in the atmosphere. Based on the four-dimensional variational objective functional containing the forward model operator, an assimilation system for radionuclide emission intensity and emission location is built. In the forward model operator, the turbulent diffusion term corresponding to the two particle size ranges adopts a differentiated turbulent diffusion coefficient, and the corresponding dry deposition term adopts a differentiated dry deposition velocity. The location, time, and amount of radionuclide emissions recorded and estimated in nuclear accident / nuclear leak reports are used as prior information. This prior information is then input into the Eulerian model for nuclear pollution forecasting to conduct simulation forecasts. Under the constraints of high spatiotemporal resolution meteorological forecast data, the simulation outputs continuous meteorological fields and radionuclide pollution concentration fields during the study period as driving data. The driving data is input into the assimilation system, and assimilation inversion is performed under the constraints of observation data of multiple types of radionuclides to obtain the hourly radionucde emission location, emission intensity and emission amount during the study period. The forward model operator constructed using the Eulerian model for nuclear contamination forecasting is expressed as: ； in, For the first The concentration of radioactive nuclide contamination at time t, with subscript s and l These represent fine-particle-size nuclides and coarse-particle-size nuclides, respectively, and the total concentration of the two types of nuclides is taken as the radionuclide contamination concentration. This refers to the advection diffusion term for nuclides with a small particle size. This refers to the advection diffusion term for nuclides with coarse particle sizes. Wind speed in the horizontal direction. Wind speed in the vertical direction; For the turbulent diffusion term of fine-particle-size nuclides, The turbulent diffusion term represents the coarse-grained nuclide, and the turbulent diffusion terms for both types of nuclide are smaller than the corresponding advection diffusion terms. The turbulent diffusion coefficients represent the horizontal and vertical directions for two types of nuclides with different particle sizes. To address the differences in the motion characteristics of the two types of nuclides, a differentiated adaptation is used for the turbulent diffusion coefficients. Specifically, for the finer-particle-size nuclides, a dynamic turbulent diffusion coefficient in the horizontal direction, corrected in real-time based on the intensity of turbulent fluctuations, is employed. and The coefficient is dynamically adjusted to follow changes in atmospheric turbulence to improve the accuracy of fine particle diffusion simulation; for coarse-particle nuclides, the static turbulent diffusion coefficient in the horizontal direction coupled with gravity settling effect is used. and It is adapted to the motion characteristics dominated by the gravity sedimentation of coarse particles and meets the requirements. That is, the proportion of turbulent diffusion terms is higher for fine-grained nuclides than for coarse-grained nuclides; These represent the dry precipitation terms for fine-grained and coarse-grained nuclides, respectively. Representing the dry sedimentation velocities of fine-grained and coarse-grained nuclides respectively, and satisfying... Among them, the dry deposition velocity model of fine-particle-size nuclides is constructed based on the gas-particle exchange coefficient and surface adsorption efficiency, which is suitable for their characteristics of being easily adsorbed by atmospheric particulate matter and having frequent gas-particle exchange; the dry deposition velocity model of coarse-particle-size nuclides is constructed based on the coupling of gravity deposition velocity and ground roughness, which characterizes their characteristics of being dominated by gravity deposition and significantly affected by ground resistance. For decay terms, The decay rate of a radioactive nuclide; For emissions, this represents the increase in the concentration of radionuclides in the atmosphere caused by the emission of radionuclides. It is the actual air density. The model layer height corresponding to the height of the emission source. and The model is a dynamic allocation model for total emissions among fine-grained and coarse-grained nuclides. Based on the historical statistical distribution characteristics of measured particle size distribution of emission sources, the total emissions are dynamically allocated according to the proportion of the two types of nuclides. The allocation coefficient is updated according to real-time measured data to adapt to the differences in particle size distribution of different emission sources. and In the horizontal direction, It is in the vertical direction.

2. The radionuclide emission inversion method integrating four-dimensional variational and Eulerian models according to claim 1, characterized in that, The four-dimensional variational objective functional of the radionuclide emission source term is constructed as follows: ； in, The objective function value; This is a state variable, representing the hourly radionuclide emissions at the grid point where the nuclear accident occurred or the grid point where a potential nuclear leak area is located. This refers to the prior radionuclide emissions derived from nuclear accident reports; when the location of the nuclear accident is identified, and Both are one-dimensional vectors; however, in the case of nuclear leakage, and All are three-dimensional vectors, containing grid regions representing potential nuclear leaks and the time of nuclear leaks; Indicates time; The duration of time during which radioactive nuclides are released into the atmosphere; The background error covariance of radionuclide emissions, its dimension is... Consistent, but the vector dimension may differ under different circumstances; It is the observation vector; The assimilation time window is defined as the time window within which observational data of multiple types of radionuclides are used as constraints in the assimilation and inversion process. The observation error covariance is in diagonal matrix form, and the observation error covariance differs for different types of radionuclide observation data. This is a forward model operator constructed using the Eulerian model for nuclear contamination forecasting; For observation operators, used to... The described radionuclide contamination concentration is converted into an observation operator identical to the radionuclide observation data type, while spatial matching between observation sites and model formats is performed; superscript This represents the transpose of a vector.

3. The radionuclide emission inversion method integrating four-dimensional variational and Eulerian models according to claim 2, characterized in that, When constructing the four-dimensional variational objective functional of the radionuclide emission source term, the method further includes: Based on the prior radionuclide emissions, and considering the uncertainty of the state variables, the position and height of the state variables are dynamically adjusted to calculate the background error covariance of the prior radionuclide emissions to meet the requirements of four-dimensional variational assimilation. Furthermore, considering the heterogeneity of observation data for various types of radionuclides, the observation error covariance of each type of radionuclide observation data is calculated, and specific observation operators for each type of radionuclide observation data are designed to adapt to the four-dimensional variational assimilation model. Specifically, for cases where the location of a nuclear accident-type emission source is determined, the location of the state variables is considered fixed, and only the emission intensity is inverted and assimilated. For cases where the location of a nuclear leak-type radionuclide is uncertain, the potential emission location area is dynamically adjusted, and all radionuclides within the grid of that area are treated as state variables.

4. The radionuclide emission inversion method integrating four-dimensional variational and Eulerian models according to claim 3, characterized in that, Dedicated observation operators were designed specifically for adapting to four-dimensional variational assimilation models of various radionuclide observation data, including: The observation data for various types of radionuclides include specific activity observation data, sedimentation rate observation data, and dose rate observation data. Among them, the specific activity observation operator corresponding to the specific activity observation data is composed of specific activity observation data unit conversion and grid point interpolation, which is a linear calculation problem. For the sedimentation rate observation operator corresponding to the sedimentation rate observation data and the dose rate observation operator corresponding to the dose rate observation data, the correlation between sedimentation rate, dose rate and radionuclide emission intensity is statistically analyzed based on the forecast results output by the Eulerian model for nuclear pollution forecasting, and a lookup table is constructed. Then, the corresponding observation operators and their accompanying programs are provided through the lookup table for direct invocation.

5. The radionuclide emission inversion method integrating four-dimensional variational and Eulerian models according to claim 1, characterized in that, When inputting prior information into the Eulerian model for nuclear contamination prediction to conduct simulation forecasts, the method further includes: In the simulation and forecasting process of the Eulerian model for nuclear pollution forecasting, the advection diffusion term in the forward model operator is calculated using the upwind difference scheme. The turbulent diffusion coefficient and dry deposition velocity of the two types of nuclides with different sizes are updated in the forward model operator. In the integration process of the Eulerian model for nuclear pollution forecasting, the concentration gradient diffusion equation is solved spatiotemporally using an implicit difference scheme.

6. The radionuclide emission inversion method integrating four-dimensional variational and Eulerian models according to claim 5, characterized in that, Update the turbulent diffusion coefficient and dry sedimentation velocity for the two particle size range nuclides in the forward mode operator for differentiated adaptation, including: The update method for the turbulent diffusion coefficient of fine-grained nuclides is as follows: ； ； in, These are the turbulent diffusion coefficients of fine-particle-size nuclides in the horizontal and vertical directions, respectively. The turbulence constant is It is atmospheric turbulent kinetic energy; This is the turbulent energy dissipation rate, used to reflect the turbulence decay characteristics; This is a correction function for the turbulent diffusion coefficient, and relates to the adsorption efficiency of the fine-particle-size nuclide surface. It is also related to the exchange coefficient of nuclides in the fine-particle size range. related, Used to improve the efficiency of mass exchange between fine-particle-size nuclides and the atmosphere; The update method for the turbulent diffusion coefficient of coarse-grained nuclides is as follows: ； ； in, These represent the turbulent diffusion coefficients of coarse-grained nuclides in the horizontal and vertical directions, respectively. It is the gravitational settling velocity of coarse-grained nuclides. Relevant correction functions; The mixing length is a characteristic length that characterizes turbulent motion and is related to atmospheric stratification and the roughness of the underlying surface. The update method for the dry sedimentation rate of the two types of nuclides is as follows: ； in, For dry settlement velocity, subscript Representing fine-grained and coarse-grained nuclides, It is the gravitational settling velocity. It is aerodynamic drag, which is positively correlated with the roughness of the underlying surface; It is the sublayer drag of laminar flow, which reflects the magnitude of the resistance to particles passing through the atmospheric laminar sublayer. It is the surface resistance, set as a lookup table according to the underlying surface type, used to correct for dry settling velocity.

7. The radionuclide emission inversion method integrating four-dimensional variational and Eulerian models according to claim 1, characterized in that, Both the forward model operator and the driving data are constructed and calculated from the Eulerian model for nuclear contamination prediction in the Eulerian coordinate system. The two are fully coupled to achieve matching based on the same coordinate system, and the assimilation and inversion of radionuclides are driven through the fully coupled structure.

8. The radionuclide emission inversion method integrating four-dimensional variational and Eulerian models according to claim 1, characterized in that, After collecting observational data on multiple types of radionuclides following a nuclear accident, the method further includes: Preprocessing is performed on observation data of various types of radionuclides. The preprocessing methods include background data analysis, threshold check analysis, spatiotemporal consistency analysis, and cross-validation analysis.

9. A radionuclide emission inversion system integrating four-dimensional variational and Eulerian models, characterized in that, The system includes: The data acquisition module is used to collect observation data of multiple types of radionuclides, high spatiotemporal resolution meteorological forecast data and nuclear accident / nuclear leakage reports after a nuclear accident, and to classify multiple types of radionuclides into fine-grained and coarse-grained nuclides. The assimilation system construction module is used to construct a four-dimensional variational objective functional for the radionuclide emission source term. Using the Eulerian model for nuclear pollution forecasting, it constructs a forward model operator in the four-dimensional variational objective functional to describe the evolution of radionuclides in the atmosphere. Based on the four-dimensional variational objective functional containing the forward model operator, an assimilation system for the emission intensity and location of radionuclides is built. In the forward model operator, the turbulent diffusion term corresponding to the two particle size ranges adopts a differentiated turbulent diffusion coefficient, and the corresponding dry deposition term adopts a differentiated dry deposition velocity. The prior forecast module is used to take the location, time and amount of radionuclide emission recorded and estimated in the nuclear accident / nuclear leak report as prior information, input the prior information into the Eulerian model of nuclear pollution forecast to carry out simulation forecast, and under the constraint of high spatiotemporal resolution meteorological forecast data, simulate and output the continuous meteorological field and radionuclide pollution concentration field during the study period as driving data. The emission inversion module is used to input the driving data into the assimilation system and perform assimilation inversion under the constraints of observation data of multiple types of radionuclides to obtain the hourly emission location, emission intensity and emission amount of radionuclides during the study period; The forward model operator constructed using the Eulerian model for nuclear contamination forecasting is expressed as: ； in, For the first The concentration of radioactive nuclide contamination at time t, with subscript s and l These represent fine-particle-size nuclides and coarse-particle-size nuclides, respectively, and the total concentration of the two types of nuclides is taken as the radionuclide contamination concentration. This refers to the advection diffusion term for nuclides with a small particle size. This refers to the advection diffusion term for nuclides with coarse particle sizes. Wind speed in the horizontal direction. Wind speed in the vertical direction; For the turbulent diffusion term of fine-particle-size nuclides, The turbulent diffusion term represents the coarse-grained nuclide, and the turbulent diffusion terms for both types of nuclide are smaller than the corresponding advection diffusion terms. The turbulent diffusion coefficients represent the horizontal and vertical directions for two types of nuclides with different particle sizes. To address the differences in the motion characteristics of the two types of nuclides, a differentiated adaptation is used for the turbulent diffusion coefficients. Specifically, for the finer-particle-size nuclides, a dynamic turbulent diffusion coefficient in the horizontal direction, corrected in real-time based on the intensity of turbulent fluctuations, is employed. and The coefficient is dynamically adjusted to follow changes in atmospheric turbulence to improve the accuracy of fine particle diffusion simulation; for coarse-particle nuclides, the static turbulent diffusion coefficient in the horizontal direction coupled with gravity settling effect is used. and It is adapted to the motion characteristics dominated by the gravity sedimentation of coarse particles and meets the requirements. That is, the proportion of turbulent diffusion terms is higher for fine-grained nuclides than for coarse-grained nuclides; These represent the dry precipitation terms for fine-grained and coarse-grained nuclides, respectively. Representing the dry sedimentation velocities of fine-grained and coarse-grained nuclides respectively, and satisfying Among them, the dry deposition velocity model of fine-particle-size nuclides is constructed based on the gas-particle exchange coefficient and surface adsorption efficiency, which is suitable for their characteristics of being easily adsorbed by atmospheric particulate matter and having frequent gas-particle exchange; the dry deposition velocity model of coarse-particle-size nuclides is constructed based on the coupling of gravity deposition velocity and ground roughness, which characterizes their characteristics of being dominated by gravity deposition and significantly affected by ground resistance. For decay terms, The decay rate of a radioactive nuclide; For emissions, this represents the increase in the concentration of radionuclides in the atmosphere caused by the emission of radionuclides. It is the actual air density. The model layer height corresponding to the height of the emission source. and The model is a dynamic allocation model for total emissions among fine-grained and coarse-grained nuclides. Based on the historical statistical distribution characteristics of measured particle size distribution of emission sources, the total emissions are dynamically allocated according to the proportion of the two types of nuclides. The allocation coefficient is updated according to real-time measured data to adapt to the differences in particle size distribution of different emission sources. and In the horizontal direction, It is in the vertical direction.

Images