Atmospheric visibility prediction method based on physical diagnosis model

Abstract

The invention discloses an atmospheric visibility prediction method based on a physical diagnosis model, which performs visibility analysis based on the in-fog physical diagnosis model, and can obtain a visibility prediction result based on various mainstream numerical model products such as GFS, ECMWF and the like. The method has the advantages of being high in calculation speed, small in storage occupation, free of a large number of samples and high in expansibility, and theoretical and technical supports are provided for the development of the refined forecasting and early warning technology of fog.

Description

Technical Field

[0001] The present invention belongs to the technical field of meteorological disaster prevention, and in particular relates to an atmospheric visibility prediction method. Background Art

[0002] Fog is a typical low-visibility weather phenomenon that forms suddenly and develops rapidly, severely impacting transportation and causing disasters and economic losses. With the improvement of atmospheric science theories and the development of computer technology, numerical models have been widely used to simulate and forecast fog. Current operational numerical forecast models typically do not output fog and visibility, requiring statistical methods or artificial intelligence to calculate visibility. Establishing statistical or artificial intelligence models requires a large amount of historical observation and forecast product data as training samples to obtain accurate visibility calculations, which consumes significant computing and storage resources, and cannot guarantee satisfactory results from a single training session. When applying an existing statistical or artificial intelligence model to another numerical model product, retraining is also required, significantly increasing the R&D cycle of numerical product fog forecasting technology.

[0003] At present, statistical models and artificial intelligence models are generally used in existing technologies to output visibility forecasts, and it is not possible to directly establish and obtain ground visibility forecast output results based on numerical forecast products. Summary of the Invention

[0004] Purpose of the Invention: To address the aforementioned existing problems and shortcomings, the present invention provides a method for predicting atmospheric visibility based on a physical diagnostic model. Based on the fog physics diagnostic equation, the present invention analytically derives a visibility model. This model, applied to mainstream numerical model products such as the Global Forecasting System (GFS) and the European Central Meteorological Organization (ECMWF), can produce visibility forecasts. This method offers the advantages of fast computational speed, minimal storage requirements, the absence of a large sample size, and strong scalability, providing theoretical and technical support for the development of refined fog forecasting and early warning technologies.

[0005] Technical solution: To achieve the above-mentioned purpose, the present invention adopts the following technical solution: a method for predicting atmospheric visibility based on a physical diagnostic model, comprising the following steps:

[0006] S1, obtaining numerical model forecast product data, including surface meteorological elements and isobaric surface meteorological elements, wherein the surface meteorological elements include temperature, dew point, 10m wind speed, and air pressure, and the isobaric surface meteorological elements include temperature, absolute humidity, wind speed, and geopotential height;

[0007] S2: Construct a fog path variation model. Considering the influence of temperature changes and surface droplet deposition on the fog path, a differential equation for the time-varying fog path is established. Based on the linear relationship between the fog path and ground fog concentration in meteorological studies, algebraic equations for the time-varying rate of the fog path and ground fog concentration are established, respectively. These two equations are then combined to obtain an ordinary differential equation for the time-varying fog concentration q.

[0008] S3, assuming that the initial fog concentration q1 = 0, the fog concentration q at the n+1th moment is obtained based on the ordinary differential equation of the fog concentration in step S2 n+1 , and the visibility VISn+1 at the n+1th moment is obtained by the following formula: .

[0009] Furthermore, the ordinary differential equation of the time-varying mist concentration q in step S2 is as follows: , Where α is the fog settling rate constant, which is 0.062, H is the fog top height, and β represents the amount of water vapor condensation caused by temperature drop.

[0010] Furthermore, the mist concentration qn+1 in step S3 is obtained by the following formula: , And when season ; In the above formula, It represents the amount of water vapor condensation caused by the temperature drop at the nth moment; Indicates the relative size of water vapor condensation and fog water deposition at the nth moment, Indicates the height of the fog top at the nth moment.

[0011] Furthermore, the amount of water vapor condensed due to the temperature drop at the nth moment is and the relative size of water vapor condensation and fog deposition λ n , are calculated by the following formulas: , Where Lv represents the latent heat constant of water vapor condensation, which is 2.5×10 6 J / kg; Rv represents the water vapor ratio gas constant, which is 461.5 J / kg; α is the fog water deposition rate constant, which is 0.062; basic meteorological quantities include air pressure p, temperature T, 1000 hPa isobaric surface height z1000 hPa, and time interval Δt, all of which can be directly read from numerical forecast products; represents the condensation rate function with respect to air pressure p and T; β n represents the amount of water vapor condensation caused by the temperature drop at the nth moment, λ n Indicates the relative size of water vapor condensation and fog water deposition at the nth moment, and are the condensation rate functions at 1000 hPa and the ground, respectively. and Respectively represent the temperature change values at 1000hPa and the ground. and Respectively represent the temperature at 1000hPa at the n+1th moment and the nth moment, and Represent the ground temperature at the n+1th moment and the nth moment respectively.

[0012] Furthermore, the numerical model forecast product data in step S1 is collected from the ECMWF model, and the spatial resolution of the product data is 0.25 ∘ ×0.25 ∘ , select surface, 1000hPa, 925hPa, 850hPa, 500hPa for vertical levels; the forecast time interval is 3h, select forecast product data at 12, 15, 18, 21, 24, 27, 30 o'clock.

[0013] Beneficial Effects: Compared with existing technologies, this method uses a fog physics diagnostic model to derive a visibility model. This model, applied to mainstream numerical model products such as the Global Forecasting System (GFS) and the European Central Meteorological Organization (ECMWF), can produce visibility forecasts. This method offers the advantages of fast computational speed, minimal storage requirements, a lack of sample size, and strong scalability, providing theoretical and technical support for the development of refined fog forecasting and warning technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a flow chart of the atmospheric visibility prediction method based on the physical diagnostic model described in the present invention. DETAILED DESCRIPTION

[0015] The present invention is further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, modifications of various equivalent forms of the present invention made by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0016] Based on various numerical models such as GFS and ECMWF, the present invention establishes a physical diagnostic model in fog, simulates and predicts the visibility forecast results. During the implementation process, a large number of historical samples are not required, and the requirements for computing and storage are very low. The technical method of the present invention can not only be applied to mainstream numerical model products such as GFS and ECMWF, but also be extended to any other numerical model products, saving a large amount of computing and storage resources.

[0017] As Figure 1 shown, the present invention innovatively proposes an atmospheric visibility prediction method based on a physical diagnostic model, and the specific process steps are as follows:

[0018] S1. Obtain numerical model forecast product data

[0019] (1) Taking the ECMWF model as an example, according to the current starting forecast time, obtain the corresponding data. The product data is in grib format, covering the whole world, with a spatial resolution of 0.25°×0.25°, and the vertical levels are selected as the surface, 1000 hPa, 925 hPa, 850 hPa, and 500 hPa. The interval of the forecast time is 3 h, and the forecasts at 12, 15, 18, 21, 24, 27, and 30 hours are selected. The variables are selected as surface meteorological elements (temperature, dew point, 10 m wind speed, pressure) and isobaric surface meteorological elements (temperature, absolute humidity, wind speed, geopotential height).

[0020] S2. Construct a change model of the fog water path (the total amount of fog water concentration in the vertical direction), considering the influence of the fog water path by temperature change and surface fog droplet sedimentation, establish a differential equation of the fog water path changing with time; according to the linear relationship between the fog water path and the fog water concentration at the ground in meteorological research, establish an algebraic equation of the time change rates of the fog water path and the ground fog water concentration;联立 these two equations, eliminate the common terms, and obtain an ordinary differential equation of the ground fog water concentration q changing with time. The specific process is as follows: First, establish a physical model of fog water concentration, as shown in the following formula, , In the formula, LWC(z) is the fog water concentration at height z; K is the turbulent diffusion coefficient; α is the fog water sedimentation rate constant 0.062; p and T respectively represent the air pressure and temperature; γ is the condensation rate function about p and T; Lv represents the latent heat of vapor condensation constant 2.5×10 6 J / kg; Rv represents the specific gas constant of water vapor 461.5 J / kg; C represents the temperature change rate. The boundary conditions of the equation are 0 < z < H, where H is the fog top height, which can be defined as the maximum height where the relative humidity continuously exceeds 95% above the ground, and LWC(H) is fixed at 0.

[0021] The key to the equation is: the fog water concentration LWC(0) at the ground, which is recorded as q. The fog water concentration q is the quantity to be solved in the present invention, and it is a variable that changes with time.

[0022] Integrate the equation vertically from 0 to H. The left-hand side term LWC(z) is integrated to obtain the fog path LWP (the definition of LWP in meteorology, which indicates the total amount of fog concentration in the vertical direction); the first term on the right-hand side is is turbulent diffusion. According to meteorological theory, it is assumed that turbulent diffusion only changes the vertical distribution of fog water and does not change the total amount of fog water. Therefore, it is considered that this item is 0 after integration. The second item on the right is After integration, according to the boundary conditions LWC(0)=q, LWC(H)=0, the result is -αq 2 ; The third item on the right The integral represents the amount of water vapor condensation caused by the temperature drop of the entire layer, which is recorded as β. A value greater than 0 indicates condensation, and a value less than 0 indicates evaporation. .

[0023] Since the fog layer is usually thin, the occurrence range is basically from the ground to 1000hPa, so the integration range is replaced by the ground to 1000hPa, γ and C are expressed by the average of these two layers, C represents the temperature change rate, and is calculated using the temperature change values at two adjacent moments. Therefore, β is calculated using the following formula: , Among them, γ 1000hpa and γ sfc They represent the condensation rate values of 1000hPa and the ground respectively, ΔT represents the temperature change value at two adjacent moments, ΔT 1000hpa and ΔT sfc They represent ΔT at 1000hPa and the ground respectively, Δt is the time interval, z 1000hpa It is 1000hpa altitude.

[0024] After the above processing, equation (1) is obtained, and combined with the model of fog water path (LWP) and ground fog water concentration (q) (as shown in equation 2), (1) (2) Among them, the adiabatic parameter a eq and function Γ ad These two quantities will not be used later. H is also the height of the fog top. Assuming that only the time variation of q is considered and the time variation of H and other variables is ignored, we can combine (1) and (2) and take the time derivative of equation (2). , The ordinary differential equation of the time-varying fog water concentration q is obtained, as shown in formula (3): (3) Equation (3) can be solved analytically for each time interval of the numerical forecast product. Let the forecast time of the numerical model be t1, t2, ..., t n , the interval is Δt, and the corresponding q is q1,q2,…,q n , and let q1 corresponding to the first moment t1 be 0. Taking t1 to t2 as an example, take q1 as the known quantity and q2 as the quantity to be solved, and calculate q2 according to the following detailed derivation steps. Then for t2 to t3, t3 to t4, ..., t n-1 to t n , using the same solution method, we can get q at all moments. The specific process is shown in step S3.

[0025] S3, for the ordinary differential equation obtained in S2, let the fog water concentration q1=0 (initial condition) at the first moment, and find the analytical solution to obtain the fog water concentration q at the next moment n+1 and the current q n The recursive relationship, and the visibility VIS at the corresponding moment n+1 The calculation is as follows: , The intermediate variables are as follows: , In the above formula, the constants include α, L v 、R v , L v The latent heat constant of water vapor condensation is 2.5×10 6 J / kg, R v represents the water vapor ratio gas constant of 461.5 J / kg, α is the fog deposition rate constant of 0.062; basic meteorological quantities include air pressure p, temperature T, 1000 hPa isobaric surface height z1000 hPa, and time interval Δt, all of which are variables that can be directly read in numerical forecast products; γ is a condensation rate function related to p and T in meteorology; the intermediate calculation quantity is 、 、 、 、 、 、 、 、 、 , , their subscript n represents the current moment, and n+1 represents the next moment, The meaning of is the amount of water vapor condensation caused by temperature drop. The meaning of is the relative size of water vapor condensation and fog water deposition. and Indicates the γ value at 1000hPa and the ground, and Indicates the temperature change value between 1000hPa and the ground. and Indicates the temperature at 1000hPa at the current moment n and the next moment n+1, and represents the ground temperature at the current time n and the next time n+1, H n is the fog top height, which can be defined as the maximum height from the ground where the relative humidity is continuously greater than 95%. The final result is q n+1 and VIS n+1 , indicating the fog concentration and visibility at the next moment.

[0026] Example: Based on the atmospheric visibility prediction method of the present invention, a heavy fog forecast for Hebei Province on December 29-30, 2023, was tested. The TS score of the fog forecast was quantitatively calculated. The results showed that the official visibility forecast from ECMWF had a TS score of only 0.17, while using the method of the present invention, the TS score increased to 0.38, indicating that the forecast and the measured visibility were significantly better.

[0027] This study developed a visibility prediction model based on numerical modeling and a physical diagnostic model for fog. This model was applied to mainstream numerical model products, such as the Global Forest Service (GFS) and the European Central Meteorological Organization (ECMWF), to generate visibility forecasts. This method offers the advantages of fast computational speed, minimal storage requirements, a lack of sample size, and strong scalability, providing theoretical and technical support for the development of refined fog forecasting and warning technologies.

[0028] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solutions and technical concepts of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for predicting atmospheric visibility based on a physical diagnostic model, characterized in that The following steps are involved: S1, obtaining numerical model forecast product data, including surface meteorological elements and isobaric surface meteorological elements, wherein the surface meteorological elements include temperature, dew point, 10m wind speed, and air pressure, and the isobaric surface meteorological elements include temperature, absolute humidity, wind speed, and geopotential height; S2: Construct a fog path variation model. Considering the influence of temperature changes and surface droplet deposition on the fog path, a differential equation for the time-varying fog path is established. Based on the linear relationship between the fog path and ground fog concentration in meteorological studies, algebraic equations for the time-varying rate of the fog path and ground fog concentration are established, respectively. These two equations are then combined to obtain an ordinary differential equation for the time-varying fog concentration q. S3, assuming that the initial fog concentration q1 = 0, the fog concentration q at the n+1th moment is obtained based on the ordinary differential equation of the fog concentration in step S2 n+1 , and the visibility VISn+1 at the n+1th moment is obtained by the following formula: 。 2. The atmospheric visibility prediction method based on the physical diagnostic model according to claim 1 is characterized in that: The ordinary differential equation of the time-varying mist concentration q in step S2 is as follows: ， Where α is the fog settling rate constant, which is 0.062, H is the fog top height, and β represents the amount of water vapor condensation caused by temperature drop.

3. The atmospheric visibility prediction method based on the physical diagnostic model according to claim 2 is characterized in that: The mist concentration qn+1 in step S3 is obtained by the following formula: ， And when Time, order ; In the above formula, It represents the amount of water vapor condensation caused by the temperature drop at the nth moment; Indicates the relative size of water vapor condensation and fog water deposition at the nth moment, Indicates the height of the fog top at the nth moment.

4. The atmospheric visibility prediction method based on the physical diagnostic model according to claim 3 is characterized by: The amount of water vapor condensation caused by the temperature drop at the nth moment and the relative size of water vapor condensation and fog deposition λ n , are calculated by the following formulas: ， Where Lv represents the latent heat constant of water vapor condensation, which is 2.5×10 6 J / kg; Rv represents the water vapor ratio gas constant, which is 461.5 J / kg; α is the fog water deposition rate constant, which is 0.062; Basic meteorological quantities include air pressure p, temperature T, 1000hPa isobaric surface height z1000hPa, and time interval Δt, all of which can be directly read from numerical forecast products; represents the condensation rate function with respect to air pressure p and T; β n represents the amount of water vapor condensation caused by the temperature drop at the nth moment, λ n Indicates the relative size of water vapor condensation and fog water deposition at the nth moment, and are the condensation rate functions at 1000 hPa and the ground, respectively. and Respectively represent the temperature change values at 1000hPa and the ground. and Represent the temperature at 1000hPa at the n+1th moment and the nth moment, and Represent the ground temperature at the n+1th moment and the nth moment respectively.

5. The atmospheric visibility prediction method based on the physical diagnostic model according to claim 3 is characterized by: The numerical model forecast product data in step S1 is collected from the ECMWF model, and the spatial resolution of the product data is 0.25 ∘ ×0.25 ∘ , select surface, 1000hPa, 925hPa, 850hPa, 500hPa for vertical levels; the forecast time interval is 3h, select forecast product data at 12, 15, 18, 21, 24, 27, 30 o'clock.

Images