Method, device and medium for convective parameterization suitable for variable resolution

Abstract

The application provides a convection parameterization method, device, equipment and medium suitable for variable resolution, which comprises the following steps: forming each sub-region based on a mode grid of different resolutions, calculating an entrainment rate of each sub-region, establishing a statistical relationship, obtaining a scaling relationship of the entrainment rate along with the resolution, and obtaining a scale adaptive entrainment rate parameterization suitable for a variable resolution mode as a correction factor; calculating an outflow rate based on a parameterized convection cloud amount under a critical buoyancy condition, determining a cumulus model; generating sample data satisfying a Bessel function statistical distribution, for each time step, establishing a cumulus mass flux probability density distribution function based on a random seed number by using a Monte Carlo method; using the convection cloud amount to perform scale adaptive correction on the distribution function, obtaining a corrected cumulus mass flux, and coupling the corrected cumulus mass flux with the cumulus model to obtain a convection parameterization scheme, so as to solve the problem of designing a scale adaptive entrainment rate parameterization under a variable resolution condition.

Description

Technical Field

[0001] This invention relates to the field of numerical weather prediction technology, and in particular to a method, apparatus, equipment and medium for convective parameterization suitable for variable resolution. Background Technology

[0002] Cumulus convection is one of the most important non-adiabatic heating physical processes in numerical weather prediction models. It is a crucial regulator of global energy, moisture, and mass cycles, plays a vital role in the global climate system, and is also a major bottleneck affecting model performance. For example, the bi-equatorial convergence phenomenon reflected in climate models and the simulation bias of intra-seasonal atmospheric oscillations are related to imperfections in cumulus convection parameterization. Therefore, improving cumulus convection parameterization schemes is an important way to improve the overall performance of models. In recent years, with the continuous improvement of model resolution, developing scale-adaptive convection parameterization schemes to address the failure of theoretical assumptions in physical process parameterization developed under coarse-resolution grids has become one of the major challenges in the development of numerical weather prediction.

[0003] The existing adaptive scheme for scale adjustment of mass flux modulates the convection adjustment time of convection parameterization to a resolution-dependent manner, thereby achieving scale adaptation and improving the convection forecast of the ECMWF high-resolution model. However, it is based on the traditional quasi-equilibrium assumption and the original framework has not been broken. The research on the convection closure assumption and subgrid cloud model under non-quasi-equilibrium conditions is the direction for the development of scale-adaptive convection parameterization.

[0004] Besides the closure assumption, the mixing process between cumulus clouds and the ambient atmosphere (also known as entrainment) is also a crucial factor affecting the parameterization performance of cumulus convection. It influences temperature and humidity tendencies by altering the vertical distribution of cumulus mass flux, thereby impacting the overall atmospheric circulation. A low entrainment rate results in excessively tall simulated cumulus towers and overly vigorous convection; conversely, a high entrainment rate dilutes cloud buoyancy, preventing cumulus clouds from reaching their proper height. Previous entrainment rate parameterization methods were designed for coarse-grid models; therefore, designing a scale-adaptive entrainment rate parameterization scheme is a critical challenge in convection parameterization.

[0005] Furthermore, with the continuous improvement of model resolution, developing scale-adaptive physical process parameterization schemes to solve or avoid the problem of the failure of theoretical assumptions in traditional physical process parameterization is one of the major challenges in the development of numerical weather prediction. Traditional convection parameterization is based on statistical averaging under the premise of large sample size, which no longer satisfies the large sample size assumption in high-resolution models, and the randomness of convection becomes increasingly obvious.

[0006] In view of this, there is an urgent need to provide a scale-adaptive deep convection parameterization scheme that can solve the convection closure problem under non-quasi-equilibrium assumptions and is applicable to variable resolution grids, and a convection parameterization method applicable to variable resolution that can reasonably describe cumulus convection. Summary of the Invention

[0007] To overcome the problems existing in the related technologies, this disclosure provides a convection parameterization method, apparatus, device and medium applicable to variable resolution, so as to solve the technical problems in the related technologies.

[0008] This specification provides one or more embodiments of a convection parameterization method suitable for variable resolution, including the steps of:

[0009] Based on the coarse-grained method, the incorporation rate of each sub-region is calculated according to the pattern mesh with different resolutions. The statistical relationship between the incorporation rate and the corresponding resolution of each sub-region is established. The scaling relationship of the incorporation rate with the change of resolution is obtained and used as the correction factor of the incorporation rate parameterization formula. The scale-adaptive incorporation rate parameterization applicable to variable resolution patterns is obtained.

[0010] The convective cloud amount is parameterized based on the critical mixing ratio under critical buoyancy conditions, and the entrainment rate is calculated based on the entrainment rate to determine the cumulus cloud model.

[0011] Based on the closure assumption, sample data satisfying the statistical distribution of the Bessel function are generated. For each time step, the probability density distribution function of cumulus mass flux is established using the Monte Carlo method based on the number of random seeds.

[0012] The probability density distribution function of cumulus mass flux is scale-adaptively corrected using convective cloud amount to obtain the corrected cumulus mass flux, which serves as the boundary condition for the cumulus model. The convective heating rate is then calculated to obtain the convective parameterization scheme.

[0013] This specification provides one or more embodiments of a convection parameterization device suitable for variable resolution, comprising:

[0014] The ingress rate parameterization module is used to calculate the ingress rate of each sub-region based on the pattern mesh of different resolutions formed by the coarse-grained method, establish the statistical relationship between the ingress rate and the corresponding resolution of each sub-region, obtain the scaling relationship of the ingress rate with the change of resolution, and use it as the correction factor of the ingress rate parameterization formula to obtain a scale-adaptive ingress rate parameterization suitable for variable resolution patterns.

[0015] The rollout rate calculation module is used to parameterize the convective cloud amount based on the critical mixing ratio under critical buoyancy conditions, and to calculate the rollout rate based on the calculated rollin rate to determine the cumulus cloud model.

[0016] The cumulus mass flux determination module is used to generate sample data that satisfies the statistical distribution of the Bessel function based on the closure assumption. For each time step, the probability density distribution function of cumulus mass flux is established using the Monte Carlo method based on the number of random seeds.

[0017] The cumulus mass flux correction module is used to perform scale-adaptive correction of the probability density distribution function of cumulus mass flux using convective cloud amount, to obtain the corrected cumulus mass flux and obtain the scale-adaptive closure scheme.

[0018] The convection parameterization scheme determination module is used to use the corrected cumulus mass flux as the boundary condition of the cumulus model, calculate the convective heating rate, and obtain the convection parameterization scheme.

[0019] This specification provides one or more embodiments of a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the convection parameterization method applicable to variable resolution as described above.

[0020] This specification provides one or more embodiments of a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described convection parameterization method applicable to variable resolution.

[0021] This disclosure provides a method, apparatus, device, and medium for convection parameterization applicable to variable resolution. Its advantages lie in its design for both high-resolution and variable-resolution models, suitability for simulating severe convective weather processes at different grid scales. It proposes a scale-adaptive incorporation rate parameterization scheme by establishing a statistical relationship between the convective incorporation rate and horizontal resolution. Secondly, it implicitly calculates the entrainment rate to determine the cumulus model by parameterizing convective cloud amount, then generates samples satisfying the statistical distribution of the Bessel function, and uses the Monte Carlo method to obtain the cumulus mass flux from this distribution. Using large eddy simulation results, it analyzes the probability density distribution of cumulus mass flux at different spatial scales, improves the probability density distribution function describing cumulus mass flux by correcting the theoretical model to solve the convection closure problem under non-quasi-equilibrium assumptions, and finally completes the entire convection parameterization scheme by performing scale-adaptive correction of the cumulus mass flux and coupling it with the cumulus model. This method introduces an improved stochastic model into traditional convection parameterization to solve the convection closure problem under non-quasi-equilibrium assumptions. It is applicable to any mass flux-type convection scheme, featuring advanced modeling and computational flexibility. It can be used in models of different resolutions and can improve the accuracy of model predictions. Especially for mid-to-high latitude land areas with frequent convection in summer, it meets the precipitation forecasting needs of severe weather processes such as typhoons and rainstorms, which is of great significance to the development of the meteorological industry. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in one or more embodiments of this specification or in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A flowchart illustrating a convection parameterization method suitable for variable resolution, provided for one or more embodiments of this specification;

[0024] Figure 2 A schematic diagram of the functional structure of the cumulus convection parameterization method module provided in one or more embodiments of this specification;

[0025] Figure 3 A block diagram of a convection parameterization device suitable for variable resolution, provided for one or more embodiments of this specification;

[0026] Figure 4 This is a schematic diagram of the structure of a computer device provided for one or more embodiments of this specification. Detailed Implementation

[0027] To enable those skilled in the art to better understand the technical solutions in one or more embodiments of this specification, the technical solutions in one or more embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, and not all of the embodiments. Based on one or more embodiments of this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of this invention.

[0028] The present invention will now be described in detail with reference to specific embodiments and accompanying drawings.

[0029] Method Implementation Examples

[0030] According to embodiments of the present invention, a convection parameterization method suitable for variable resolution is provided, such as... Figure 1 The diagram shown is a flowchart of a convection parameterization method suitable for variable resolution provided in this embodiment. The convection parameterization method suitable for variable resolution according to this embodiment of the invention includes the following steps:

[0031] Step S1: Based on the coarse-grained method, calculate the convolution rate of each sub-region formed by the pattern grid with different resolutions, establish the statistical relationship between the convolution rate and the corresponding resolution of each sub-region, obtain the scaling relationship of the convolution rate with the resolution, and use it as the correction factor of the convolution rate parameterization formula to obtain the scale-adaptive convolution rate parameterization applicable to the variable resolution pattern.

[0032] Step S2: Parameterize the convective cloud amount based on the critical mixing ratio under critical buoyancy conditions, and calculate the entrainment rate based on the calculated entrainment rate to determine the cumulus model.

[0033] Step S3: Based on the closure assumption, generate sample data that satisfies the statistical distribution of the Bessel function. For each time step, establish the probability density distribution function of cumulus mass flux using the Monte Carlo method based on the number of random seeds.

[0034] Step S4: Use convective cloud amount to perform scale-adaptive correction on the probability density distribution function of cumulus mass flux to obtain the corrected cumulus mass flux, obtain the scale-adaptive closure scheme, and couple it with the cumulus model to obtain the convection parameterization scheme.

[0035] The method provided in this embodiment is designed for high-resolution and variable-resolution models and is applicable to the simulation of severe convective weather processes at different grid scales. It proposes a scale-adaptive entrainment rate parameterization scheme by establishing a statistical relationship between convective entrainment rate and horizontal resolution. Secondly, it determines the cumulus model by implicitly calculating the entrainment rate through parameterized convective cloud amount. Then, it generates samples that satisfy the statistical distribution of the Bessel function and uses the Monte Carlo method to obtain the cumulus mass flux from this distribution. Using large eddy simulation results, it analyzes the probability density distribution of cumulus mass flux at different spatial scales and improves the probability density describing cumulus mass flux by correcting the theoretical model. This invention employs a distribution function to address the convection closure problem under non-quasi-equilibrium assumptions. Finally, it completes the entire convection parameterization scheme by performing scale-adaptive correction on cumulus mass flux and coupling it with a cumulus model. This method introduces an improved stochastic model into traditional convection parameterization to solve the convection closure problem under non-quasi-equilibrium assumptions. It is applicable to any mass flux-based convection scheme, such as precipitation forecasting for severe weather events like typhoons and heavy rainstorms. For example, this method can be applied to the IFS convection scheme and coupled with the variable resolution model GRIST and a typhoon track and intensity post-processing system to establish a variable resolution typhoon forecasting system. This invention features advanced modeling and computational flexibility, can be used with models of different resolutions, and can improve model prediction accuracy, especially for mid-to-high latitude landmasses with frequent convection in summer. It meets the precipitation forecasting needs for severe weather events such as typhoons and heavy rainstorms, and is of great significance to the development of the meteorological industry.

[0036] In this embodiment, reference Figure 2The diagram shown illustrates the functional structure of the cumulus convection parameterization method module provided in this embodiment. The convection parameterization scheme is divided into two parts: a cumulus model and a closure assumption. The first part is the cumulus model. In this embodiment, the cumulus model consists of two parts: scale-adaptive ingress rate parameterization and implicit ingress rate solution. The closure assumption scheme includes two parts: first, establishing the statistical distribution function of cumulus mass flux; and second, performing scale-aware correction on the cumulus mass flux.

[0037] In this embodiment, the scale-adaptive convolution rate parameterization in step S1 specifically includes the following steps.

[0038] Step S11: Using a coarse-grained method, the model simulation area is divided into several sub-regions of equal size in stages, with each sub-region being of a different size. This is used to analogize model grids of different resolutions in climate models. Based on the large eddy simulation dataset, the model simulation area in this invention is 6.4 km. Using the coarse-grained method, the entire area is successively divided into sub-regions of sizes 3.2 km × 3.2 km, 1.6 km × 1.6 km, 0.8 km × 0.8 km, 0.4 km × 0.4 km, and 0.2 km × 0.2 km, respectively, corresponding to climate model grids with resolutions of 6.4 km, 3.2 km, 1.6 km, 0.8 km, 0.4 km, and 0.2 km.

[0039] Step S12: Calculate the enumeration rate of the simulation region and each sub-region of the model, establish the statistical relationship between the enumeration rate of each sub-region and the size of the sub-region (equivalent to the model grid resolution), and obtain the scaling relationship between the enumeration rate and the resolution (from traditional coarse resolution to high resolution) by referring to Equation 2.

[0040] The entrainment rate was calculated using the formula published by Wang Xiaocong and Zhang Minghua in the journal *Progress in Earth System Modeling* on May 14, 2014, for "Shallow Convection Vertical Velocity of Different Stream Types," as follows:

[0041]

[0042] The convection recognition operator is defined using the updraft definition (i.e., simultaneously satisfying both upward motion and condensation conditions), ε φ φ represents the incorporation rate calculated from tracer diagnostics, where φ is the tracer concentration (e.g., total water volume qt, liquid water temperature θ). ι The subscript 'c' represents the average attribute within the cloud, and 'e' represents the average attribute outside the cloud environment.

[0043] In this embodiment, step S12 specifically includes the following steps:

[0044] Step S121: Average the convolution rate of the model simulation region and the convolution rate in different resolution grids obtained by the coarsening method in the vertical direction to obtain the vertical average convolution rate of the model simulation region and the vertical average convolution rate in different resolution grids, respectively.

[0045] Step S122: Using the vertical average incursion rate result of the model simulation area as a reference, obtain the ratio of the vertical average incursion rate in different resolution grids to the vertical average incursion rate in the model simulation area, that is, the scaling factor Dε of the vertical average incursion rate in different resolution grids.

[0046] Step S123: Simultaneously calculate the ratio D of the size of the coarsened mesh at different resolutions to the size of the model simulation region. Take the logarithm of D to the base 2 to obtain Dx, which is the logarithmic rate of change of the mesh scale at different resolutions. By fitting, it is determined that the vertical average convolution scaling factor Dε and the logarithmic rate of change of the mesh scale at different resolutions Dx have an exponential relationship approximately equal to Equation (2), as shown in Equation 2 below:

[0047] Dε=exp(0.027687*Dx^1.9223) (2);

[0048] Using the above exponential relationship, the convolution rate applicable to unknown resolution modes (e.g., high resolution) can be determined based on the convolution rate parameterization scheme applicable to known resolution grid modes (e.g., coarse resolution).

[0049] Step S13: Multiply the traditional convolution rate parameterization formula applicable to coarse resolution grids with the scaling relationship between convolution rate and resolution to obtain the corrected convolution rate applicable to high resolution mode.

[0050] In this embodiment, the rollout rate is implicitly calculated in step S2 by parameterizing the convective cloud amount (i.e., the percentage of convective cloud coverage obtained from the large eddy simulation nested in the climate system model scale grid). Since the rollout process mainly occurs at the top of the convective cloud, i.e., near the neutral buoyancy layer, this embodiment adopts the critical buoyancy (neutral buoyancy) concept, parameterizes the convective cloud amount, and the specific steps for calculating the rollout rate are as follows.

[0051] Step 21: Calculate the critical mixing ratio χ c The critical mixing ratio is the mixing ratio of ambient air contained in a convective cloud when the convective cloud is in a neutral buoyancy state (i.e., the virtual potential temperature of the convective cloud is equal to the virtual potential temperature of the environment). Therefore, for a certain atmospheric layer, the critical mixing ratio χ is... c The smaller the value, the less positive buoyancy mixture and the more negative buoyancy mixture in that layer, which is conducive to increasing the roll-out rate and thus reducing the amount of convective clouds.

[0052] Critical mixing ratio χ cThe calculation formula is as shown in equation (3):

[0053]

[0054] In the above formula, Δθ v , Δθ ι , Δq t θ represents the difference between the convective cloud wet potential temperature, liquid water potential temperature, and total water vapor specific humidity relative to the environment, respectively; where θ ι q t The cloud interior value can be obtained from θ at the cloud base. ι q t The involvement rate diagnosed in the previous step is calculated by back-calculating θ using equation (1). v Can be combined with θ ι q t Solving the system of equations simultaneously, we obtain the environmental θ. v θ ι q t This can be directly regarded as the grid average value of the corresponding variable, thereby obtaining the temperature and humidity difference between the convective cloud and the environment; other coefficients involved, α, β, γ, are defined as in equations (4), (5), and (6):

[0055]

[0056]

[0057]

[0058] In the above formula, the subscript u represents the convective ascending branch, which is equivalent to a convective cloud; C p Where L is the isobaric specific heat, p is the latent heat of vaporization, and θ is the gas pressure. l q represents the liquid water level temperature. s Where saturation specific humidity is given, Y is temperature, and T is... u Temperature within the convective cloud.

[0059] Step 22: Based on the critical mixing ratio χ c The difference Δq between the total water vapor specific humidity of convective clouds and the environment was calculated using a large eddy simulation dataset. t The standard deviation of the total water vapor specific humidity of the grid σ qt The ratio of the total water content surplus Y under neutral buoyancy conditions is used to calculate the total water content surplus Y. c As shown in equation (7):

[0060]

[0061] Step 23: The total water vapor specific humidity gradient under normalized neutral buoyancy conditions can be determined through calculation in step S22. The normalized gradient of convective cloud amount σ is roughly linearly correlated with the data, and a fitting relationship is obtained; the fitting equation is as follows (8):

[0062]

[0063] Step 24: Based on the fitting relationship and given the boundary conditions of the convective cloud base, iteratively calculate and reconstruct the convective cloud volume of each layer from the cloud base to the cloud top by layer upwards.

[0064] Because of Y c Since the vertical profile is known, the convective cloud volume profile of each layer from the cloud base to the cloud top can be reconstructed by iteratively calculating layer by layer from the convective cloud base upwards based on the linear fitting relationship in Equation 8.

[0065] That is, given the boundary conditions of the convective cloud base, the cloud amount reconstruction formula for the i-th layer between the cloud base k and the cloud top k+N is shown in equation (9):

[0066]

[0067] Step 25: Parameterize the convective cloud amount using the reconstructed convective cloud amount profile, and determine the entrainment rate based on the calculated entrainment rate.

[0068] The core is the total water content surplus gradient under normalized neutral buoyancy conditions. Since the convective cloud amount can be parameterized by the above methods, it is no longer necessary to parameterize the entrainment rate. Given the entrainment rate and the convective cloud amount, the entrainment rate δ can be directly calculated by formula (10).

[0069]

[0070] In the formula, ε is the convolution rate obtained by parameterization in step S12.

[0071] In this embodiment, the closure hypothesis scheme comprises two parts: first, establishing the statistical distribution function of cumulus cloud mass flux; and second, performing scale-aware correction on the cumulus cloud mass flux. The details of these two parts are as follows.

[0072] In the most commonly used mass flux-based parameterization schemes, mass flux is a key variable for parameterizing the impact of convective activity on climate system models. It is defined as M = ρσω, where air density ρ, convective cloud amount σ, and cumulus vertical velocity ω are the parameters. If the convective cloud amount σ is known, air density ρ is usually considered constant. Therefore, parameterizing the cumulus vertical velocity is sufficient to obtain the vertical distribution of cumulus mass flux M, essentially completing the cloud model. The boundary conditions of the cumulus mass flux equation will be introduced in the following section on closure assumptions.

[0073] In this embodiment, the cumulus mass flux distribution based on the radiation-convection balance model differs from the Gaussian distribution in traditional coarse-resolution climate models. The high-resolution model's cumulus mass flux distribution more closely approximates the Bessel function statistical distribution. Therefore, samples satisfying the Bessel function statistical distribution are first generated, and then the Monte Carlo method is used to obtain the cumulus mass flux probability density distribution function from this distribution. Specifically, the following steps are included:

[0074] Step S31: Based on the Bessel function distribution, generate sample data of the Bessel function distribution that satisfy the first and second order statistical intervals.

[0075] For example, the number of samples is set to 200; the sample generation process can be as follows: random parameterization is based on the Bessel function distribution, as shown in Equation (11). Since the convection parameterization scheme also needs to characterize convective clouds in different development states through mass flux after the model grid resolution is increased, in this embodiment, convective activity is described by clouds of different quantities and mass fluxes existing within a given time. The sample data generation process in this embodiment is as follows.

[0076] Assuming that the size of a single cloud is much smaller than the size of the pattern grid and is randomly distributed in space, there is no correlation between the mass fluxes appearing in adjacent grid frames. The aforementioned Bessel function distribution involves two parameters: ensemble average total cumulus mass flux. <M b > and cumulus mass flux of a single cloud m b Similar to other parameterization methods based on quality flux, the ensemble average total cumulus quality flux < M. b >Based on the closure assumption, the average cumulus mass flux m of a single cloud b Then assume it to be a constant, with a value of 2 × 10. 7 kg s -1 .

[0077]

[0078] In the formula, I1 is the modified first-order Bessel function, obtained using the quasi-equilibrium assumption. <M b After that, sample data that follows the distribution of equation (12) can be generated.

[0079] Step S32: Based on the sample data generated in step S31, and according to the random seed number, combined with the calculated ensemble average total cumulus mass flux, the Monte Carlo method is used to obtain the convective cloud base mass flux.

[0080] In this embodiment, the scale-adaptive correction of cumulus mass flux in step S4 is the second part of the closure assumption of this scheme. As shown in equation (12), the convection scheme applicable to variable resolution forecasting adjusts the convective cloud base mass flux calculated by the stochastic process based on the convective cloud amount σ (i.e., the percentage of convective cloud coverage area obtained from the large eddy simulation diagnosis nested in the climate system model scale grid). The corrected ensemble average total cumulus mass flux M was obtained. b This, in turn, affects the intensity of convection. The greater the convective cloud cover, the higher the analytical capability of the model grid for convection, and the parameterization part can be weakened and not considered.

[0081] For example, under extreme conditions, when convective clouds cover the entire grid, i.e., the convective cloud amount σ is 1, M b If σ is 0, the parameterization scheme is automatically turned off. If σ < 1, then the average total cumulus mass flux M obtained from the corrected set is used. b Update the thermodynamic properties of the cumulus cloud model and complete calculations such as convective-scale eddy transport and latent heat release.

[0082]

[0083] Finally, the cloud base mass flux obtained by the closure assumption is used as the boundary condition of the cumulus model to calculate the convective heating rate, thus completing the improvement of the traditional convection parameterization scheme.

[0084] In this embodiment, the method for convection parameterization based on the mass flux method controls mass and other conserved quantities by the convective mass flux at the cloud base. Solving for the cumulus mass flux is also called convection closure. One method for handling convection closure under non-quasi-equilibrium assumptions is to treat the grid in the high-resolution mode as a sub-grid in the coarse-resolution mode. The coarse-resolution mode containing several sub-grids satisfies the quasi-equilibrium assumption, thus allowing the introduction of a stochastic model (i.e., establishing a model based on Bessel functions and obtaining the convective cloud base mass flux using the Monte Carlo method) to obtain the convection intensity of the sub-grid (corresponding to the high-resolution grid). Specifically, using large eddy simulation results, the probability density distribution of cumulus mass flux at different spatial scales is analyzed and compared with the theoretical model derived from statistical mechanics. By correcting the theoretical model, the probability density distribution function describing the cumulus mass flux is improved. Then, the cumulus mass flux obtained under the quasi-equilibrium assumption at different spatial scales is calculated and compared with the corresponding large eddy simulation results, revealing the applicable spatial scale of the quasi-equilibrium assumption. When the model grid spacing is smaller than this scale, based on the improved probability density distribution function of cumulus mass flux, a random method is used to obtain cumulus mass flux to update the thermodynamic properties of the cumulus model, completing the feedback calculation of convection on large-scale temperature, humidity, and wind. Considering that the occurrence and intensity of convection are closely related to the underlying surface conditions of the subgrid and the environmental field, the following improvements are made based on random sampling: the non-uniformity of the subgrid environmental field is fully considered, so that a larger cumulus mass flux is obtained in the active convection region. The scale-adaptive closure scheme is coupled with the cumulus model to complete the entire deep convection parameterization scheme.

[0085] Device Examples

[0086] According to embodiments of the present invention, a convection parameterization device suitable for variable resolution is provided, such as... Figure 3 The diagram shown is a block diagram of a convection parameterization device suitable for variable resolution provided in this embodiment. According to an embodiment of the present invention, the convection parameterization device suitable for variable resolution includes:

[0087] The incorporation rate parameterization module 10 is used to calculate the incorporation rate of each sub-region based on the pattern grid of different resolutions formed by the coarse-grained method, establish the statistical relationship between the incorporation rate and the corresponding resolution of each sub-region, obtain the scaling relationship of the incorporation rate with the change of resolution, and use it as the correction factor of the incorporation rate parameterization formula to obtain the scale-adaptive incorporation rate parameterization applicable to the variable resolution pattern.

[0088] The rollout rate calculation module 20 is used to parameterize the convective cloud amount based on the critical mixing ratio under critical buoyancy conditions, and to calculate the rollout rate based on the calculated rollout rate to determine the cumulus cloud model.

[0089] The cumulus mass flux determination module 30 is used to generate sample data that satisfies the statistical distribution of the Bessel function based on the closure assumption. For each time step, the probability density distribution function of the cumulus mass flux is established using the Monte Carlo method based on the number of random seeds.

[0090] The cumulus mass flux correction module 40 is used to perform scale-adaptive correction of the probability density distribution function of cumulus mass flux using convective cloud amount, to obtain the corrected cumulus mass flux, and to obtain the scale-adaptive closure scheme.

[0091] The convection parameterization scheme determination module 50 is used to use the corrected cumulus mass flux as the boundary condition of the cumulus model, calculate the convective heating rate, and obtain the convection parameterization scheme.

[0092] The apparatus provided in this embodiment is designed for high-resolution and variable-resolution models and is suitable for simulating convective weather processes at different grid scales. The entrainment rate parameterization module 10 establishes a statistical relationship between the convective entrainment rate and horizontal resolution, proposing a scale-adaptive entrainment rate parameterization scheme. The re-entrainment rate calculation module 20 implicitly calculates the entrainment rate to determine the cumulus model by parameterizing convective cloud amount. The re-cumulus mass flux determination module 30 generates samples that satisfy the statistical distribution of the Bessel function and uses the Monte Carlo method to obtain the cumulus mass flux from this distribution. The cumulus mass flux correction module 40 uses the large eddy simulation results to analyze different... This method improves the probability density distribution of cumulus mass flux at spatial scales by correcting the theoretical model and refining the probability density distribution function describing cumulus mass flux to address the convection closure problem under non-quasi-equilibrium assumptions. Finally, by performing scale-adaptive correction on cumulus mass flux and coupling it with the cumulus model, the entire convection parameterization scheme is completed. This method introduces an improved stochastic model into traditional convection parameterization to solve the convection closure problem under non-quasi-equilibrium assumptions. It is applicable to any mass flux-type convection scheme, has advanced modeling capabilities and computational flexibility, can be used for different resolution models, and can improve model prediction accuracy.

[0093] In this embodiment, the rollout rate calculation module 20 is used to parameterize the convective cloud cover based on the critical buoyancy condition and calculate the rollout rate. Specifically, it is configured to calculate the rollout rate by executing the following steps:

[0094] Step A1: Calculate the critical mixing ratio, which is the mixing ratio of ambient air contained in the convective cloud when the convective cloud is in a neutral buoyancy state.

[0095] Step A2: Based on the critical mixing ratio, combined with the difference between the total water vapor specific humidity of the convective cloud and the environment, and the standard deviation of the total water vapor specific humidity of the convective cloud, calculate the total water content surplus under neutral buoyancy conditions.

[0096] Step A3: Through calculation, it can be determined that the normalized total water vapor specific humidity gradient under normalized neutral buoyancy conditions is roughly linearly correlated with the normalized gradient of convective cloud cover, thus obtaining a fitting relationship.

[0097] Step A4: Based on the fitting relationship and given the boundary conditions of the convective cloud base, the convective cloud volume of each layer from the cloud base to the cloud top is reconstructed by iteratively calculating layer by layer upwards from the convective cloud base.

[0098] Step A5: Parameterize the convective cloud amount using the reconstructed convective cloud amount profile, and determine the entrainment rate based on the calculated entrainment rate.

[0099] The embodiments of the present invention are device embodiments corresponding to the above method embodiments. The specific operations of each module processing step can be understood with reference to the description of the method embodiments, and will not be repeated here.

[0100] like Figure 4 As shown, the present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, it implements the convection parameterization method applicable to variable resolution in the above embodiments, or when the computer program is executed by a processor, it implements the convection parameterization method applicable to variable resolution in the above embodiments. When the computer program is executed by the processor, it implements the following method steps:

[0101] Step S1: Based on the coarse-grained method, calculate the convolution rate of each sub-region formed by the pattern grid with different resolutions, establish the statistical relationship between the convolution rate and the corresponding resolution of each sub-region, obtain the scaling relationship of the convolution rate with the change of resolution, and use it as the correction factor of the convolution rate parameterization formula to obtain the scale-adaptive convolution rate parameterization applicable to the variable resolution pattern.

[0102] Step S2: Parameterize the convective cloud amount based on the critical mixing ratio under critical buoyancy conditions, and calculate the entrainment rate based on the calculated entrainment rate to determine the cumulus model.

[0103] Step S3: Based on the closure assumption, generate sample data that satisfies the statistical distribution of the Bessel function. For each time step, establish the probability density distribution function of cumulus mass flux using the Monte Carlo method based on the number of random seeds.

[0104] Step S4: Use convective cloud amount to perform scale adaptive correction on the probability density distribution function of cumulus mass flux to obtain the corrected cumulus mass flux, which is used as the boundary condition of the cumulus model to calculate the convective heating rate and obtain the convective parameterization scheme.

[0105] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0106] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for apparatus or system embodiments, since they are basically similar to method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The apparatus and system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0107] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and the contents not described in detail in the specification of the present invention are well known to those skilled in the art.

Claims

1. A convection parameterization method suitable for variable resolution, characterized in that, Including the following steps: Based on the coarse-grained method, the incorporation rate of each sub-region is calculated according to the pattern mesh with different resolutions. The statistical relationship between the incorporation rate and the corresponding resolution of each sub-region is established. The scaling relationship of the incorporation rate with the change of resolution is obtained and used as the correction factor of the incorporation rate parameterization formula. The scale-adaptive incorporation rate parameterization applicable to variable resolution patterns is obtained. The convective cloud amount is parameterized based on the critical mixing ratio under critical buoyancy conditions, and the entrainment rate is calculated based on the entrainment rate to determine the cumulus cloud model. Based on the closure assumption, sample data satisfying the statistical distribution of the Bessel function are generated. For each time step, the probability density distribution function of cumulus mass flux is established using the Monte Carlo method based on the number of random seeds. The probability density distribution function of cumulus mass flux is scale-adaptively corrected using convective cloud amount. The corrected cumulus mass flux is then used as the boundary condition for the cumulus model. The convective heating rate is calculated, and the convective parameterization scheme is obtained.

2. The convection parameterization method applicable to variable resolution as described in claim 1, characterized in that, The specific statistical relationship between the entrainment rate and the corresponding resolution of each sub-region is as follows: The incorporation rate of the model simulation region and the incorporation rate of different resolution grids obtained by the coarsening method are averaged in the vertical direction to obtain the vertical average incorporation rate of the model simulation region and the vertical average incorporation rate of different resolution grids, respectively. Using the vertical average entrainment rate of the model simulation region as a reference, the ratio of the vertical average entrainment rate in different resolution grids to the vertical average entrainment rate of the model simulation region is obtained, which is the scaling factor of the vertical average entrainment rate in different resolution grids. D ; Calculate the ratio of the size of the coarsened mesh at different resolutions to the size of the model simulation region. D ,right D Taking the logarithm to base 2 to obtain the logarithmic rate of change for different grid resolutions Dx The scaling factor for the vertical average incorporation rate is determined by fitting. D Logarithmic rate of change with different resolution grid scales Dx The exponential relationship is as follows: 。 3. The convection parameterization method applicable to variable resolution as described in claim 1, characterized in that, The parameterization of convective cloud amount based on the critical mixing ratio under critical buoyancy conditions and the calculation of the entrainment rate include the following steps: Calculate the critical mixing ratio, which is the ratio of ambient air contained in the mixture when the convective cloud is in a neutral buoyancy state; Based on the critical mixing ratio, the ratio of the difference between the total water vapor specific humidity of the convective cloud and the environment to the standard deviation of the total water vapor specific humidity of the grid cloud is calculated using the large eddy simulation dataset, and the total water content surplus rate under neutral buoyancy conditions is calculated. The total water vapor specific humidity gradient under normalized neutral buoyancy conditions was determined by calculation, and it was found to be roughly linearly correlated with the normalized gradient of convective cloud cover, thus obtaining a fitting relationship. Based on the fitting relationship and given convective cloud base boundary conditions, the convective cloud volume of each layer from the cloud base to the cloud top is reconstructed by iteratively calculating and reconstructing the convective cloud volume of each layer from the cloud base to the cloud top. The convective cloud amount is parameterized by reconstructing the convective cloud amount profile, and the entrainment rate is determined based on the calculated entrainment rate.

4. The convection parameterization method applicable to variable resolution as described in any one of claims 1-3, characterized in that, The critical mixing ratio Calculate as follows: ； ； ； ； In the above formula, These represent the differences between the wet potential temperature of the mixture, the liquid water level temperature, and the total water vapor specific humidity relative to the environment, respectively. Subscript u Representing convective clouds, Where L is the isobaric specific heat and L is the latent heat of vaporization. For liquid water level temperature, saturated specific humidity, For air pressure, Temperature within the convective cloud.

5. The convection parameterization method applicable to variable resolution as described in claim 4, characterized in that, The total water vapor specific humidity gradient under normalized neutral buoyancy conditions is determined by calculation, along with convective cloud cover. The normalized gradients are roughly linearly correlated, and a fitting relationship is obtained. The specific fitting relationship is as follows: ； ； In the formula, Y c The total moisture content surplus rate, The total water vapor specific humidity gradient under neutral buoyancy conditions. For convective cloud cover, The standard deviation of the total water vapor specific humidity of the grid.

6. The convection parameterization method applicable to variable resolution as described in claim 4, characterized in that, The convective cloud volume of each layer from the cloud base to the cloud top is reconstructed by iterative calculation from the cloud base upwards. The specific calculation formula is as follows: ； ； In the formula, For the bottom of the clouds, For the cloud base k to cloud top k+N, the first... layer, Y c The total moisture content surplus rate, The standard deviation of the total water vapor specific humidity of the grid.

7. A convection parameterization device suitable for variable resolution, characterized in that, include: The ingress rate parameterization module calculates the ingress rate of each sub-region based on the pattern mesh of different resolutions using a coarse-grained method. It establishes a statistical relationship between the ingress rate and the corresponding resolution of each sub-region, obtains the scaling relationship of the ingress rate with the change of resolution, and uses it as a correction factor for the ingress rate parameterization formula to obtain a scale-adaptive ingress rate parameterization suitable for variable resolution patterns. The rollout rate calculation module is used to parameterize the convective cloud amount based on the critical mixing ratio under critical buoyancy conditions, and to calculate the rollout rate based on the calculated rollin rate to determine the cumulus cloud model. The cumulus mass flux determination module is used to generate sample data that satisfies the statistical distribution of the Bessel function based on the closure assumption. For each time step, the probability density distribution function of cumulus mass flux is established using the Monte Carlo method based on the number of random seeds. The cumulus mass flux correction module is used to perform scale-adaptive correction of the probability density distribution function of cumulus mass flux using convective cloud amount, to obtain the corrected cumulus mass flux and obtain the scale-adaptive closure scheme. The convection parameterization scheme determination module is used to use the corrected cumulus mass flux as the boundary condition of the cumulus model, calculate the convective heating rate, and obtain the convection parameterization scheme.

8. The convection parameterization device suitable for variable resolution as described in claim 7, characterized in that, The roll-out rate calculation module is configured to perform the following steps to calculate the roll-out rate: Calculate the critical mixing ratio, which is the ratio of ambient air contained in the mixture when the mixture is in a neutral buoyancy state; Based on the critical mixing ratio, the ratio of the difference between the total water vapor specific humidity of the convective cloud and the environment to the standard deviation of the total water vapor specific humidity of the grid cloud is calculated using the large eddy simulation dataset, and the total water content surplus rate under neutral buoyancy conditions is calculated. Calculations show that the normalized total water vapor specific humidity gradient under normalized neutral buoyancy conditions is roughly linearly correlated with the normalized gradient of convective cloud cover, thus obtaining a fitting relationship. Based on the fitting relationship and given the boundary conditions of the convective cloud base, the convective cloud volume of each layer from the cloud base to the cloud top is reconstructed by iteratively calculating and reconstructing the convective cloud volume of each layer from the cloud base to the cloud top. The convective cloud amount is parameterized by reconstructing the convective cloud amount profile, and the entrainment rate is determined based on the calculated entrainment rate.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the convection parameterization method for variable resolution as described in any one of claims 1 to 6.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the convection parameterization method applicable to variable resolution as described in any one of claims 1 to 6.

Images