Turbulence parameterization method oriented to improved stable boundary layer simulation

Abstract

The invention provides an improved stable boundary layer simulation-oriented turbulence parameterization method, and relates to the field of silicon semiconductor component packaging. Comprising the following steps: acquiring various boundary layer observation data; determining a vertical profile and a time evolution characteristic of a turbulence key parameter under a stable boundary layer condition and determining a plurality of boundary layer schemes based on the observation data of the plurality of boundary layers according to a multi-source observation means; the mixed length is improved according to wind shear, and the improved mixed length is obtained; improving the minimum turbulence diffusion coefficient and the turbulence Prandtt number to obtain an improved minimum turbulence diffusion coefficient and an improved turbulence Prandtt number; obtaining an improved boundary layer scheme based on the improved mixing length, the improved minimum turbulence diffusion coefficient and the improved turbulence Prandtt number; and evaluating the prediction effect of the radiation fog process according to the improved boundary layer scheme to obtain an evaluation result. According to the invention, the problem of low numerical simulation capability of stable stratification atmospheric turbulence in the prior art is solved.

Description

Technical Field

[0001] The invention relates to the technical field of weather forecasting, and in particular to a turbulence parameterization method for improving stable boundary layer simulation. Background Art

[0002] The stable boundary layer is an important form of boundary layer. The inversion structure in the stable boundary layer is a key factor in the occurrence and development of fog. The thicker and stronger the inversion layer, the denser the fog. The weaker the turbulent mixing effect of the stable stratification is, the less conducive it is to the diffusion of atmospheric pollution, leading to an increase in the concentration of pollutants. The low-altitude jet phenomenon that often appears in the stable boundary layer can lead to drastic changes in the intensity and energy of the near-surface turbulence, thereby affecting the generation and evolution of fog. In addition, the intermittent nature of the turbulent motion of the stable stratification will also have a significant impact on the flux of matter and energy between the ground and the air. Therefore, accurately understanding and describing the structure and turbulence characteristics of the stable boundary layer is of great scientific significance for understanding the generation and evolution mechanism of haze weather and improving the forecast effect of numerical weather and air quality models.

[0003] However, due to the lack of sufficient field and laboratory observation data, as well as the complexity of the dynamic mechanism of the stable boundary layer itself, the parameterization of the stable stratified turbulent mixing process is still very difficult at this stage, and it is the weakest and most uncertain area in the study of the atmospheric boundary layer. The inadequacy of atmospheric models in simulating the stable boundary layer also puts forward an urgent requirement for in-depth research on the stable boundary layer.

[0004] In the stable boundary layer, the vertical shear of the horizontal wind (wind shear for short) becomes the main source of turbulence. The vortex scale is small and has strong locality. The parameterization calculation of local turbulence is closely related to the mixing length. However, there is still great uncertainty in the description of the mixing length in the current model, which is an important reason affecting the simulation effect of the stable boundary layer. A series of studies have shown that existing boundary layer schemes are usually difficult to fully describe the mechanical turbulence caused by wind shear (such as low-level jets) in the stable boundary layer. One of the important reasons is that the influence of wind shear is not considered in the parameterization calculation of the mixing length. In addition, the turbulent motion of the stable stratified atmosphere often shows obvious intermittentity. However, the current boundary layer parameterization scheme does not consider the influence of turbulence intermittentity, which is also a key reason for the simulation error of the stable boundary layer.

[0005] Fog is a weather phenomenon that occurs in a stable boundary layer. The low atmospheric visibility it causes seriously affects transportation, production and life. At present, the ability of numerical models to predict fog is still at a low level. Summary of the invention

[0006] In order to overcome the deficiencies of the prior art, the purpose of the present invention is to provide a turbulence parameterization method for improving stable boundary layer simulation, which solves the problem of low numerical simulation capability of stable stratified atmospheric turbulence in the prior art.

[0007] To achieve the above object, the present invention provides the following solutions:

[0008] A turbulence parameterization method for improved stable boundary layer simulations, including:

[0009] Obtain various boundary layer observation data;

[0010] Based on the various boundary layer observation data, the vertical profile and time evolution characteristics of key turbulent parameters (turbulent diffusion coefficient, mixing length, turbulent intermittent intensity index, etc.) under stable boundary layer conditions are determined according to multi-source observation methods;

[0011] determining a plurality of boundary layer scenarios based on the vertical profile and the time evolution characteristics;

[0012] improving the mixing length according to wind shear to obtain an improved mixing length;

[0013] The minimum turbulent diffusion coefficient and the turbulent Prandtl number are improved to obtain the improved minimum turbulent diffusion coefficient and the turbulent Prandtl number;

[0014] An improved boundary layer scheme is obtained based on the improved mixing length, the improved minimum turbulent diffusion coefficient and the improved turbulent Prandtl number;

[0015] The prediction effect of the radiation fog process is evaluated according to the improved boundary layer scheme to obtain an evaluation result.

[0016] Preferably, the plurality of boundary layer observation data include:

[0017] Meteorological tower data, ground weather station data, microwave radiometer data, wind profiler radar data, and sounding data.

[0018] Preferably, the key turbulence parameters include:

[0019] Turbulent diffusion coefficient, mixing length, and turbulent intermittency intensity index.

[0020] The present invention discloses the following technical effects:

[0021] The present invention provides a turbulence parameterization method for improving stable boundary layer simulation, including: obtaining multiple boundary layer observation data; determining the vertical profile and time evolution characteristics of key turbulence parameters under stable boundary layer conditions based on the multiple boundary layer observation data according to multi-source observation means; determining multiple boundary layer schemes based on the vertical profile and time evolution characteristics; improving the mixing length according to wind shear to obtain an improved mixing length; improving the minimum turbulent diffusion coefficient and the turbulent Prandtl number to obtain an improved minimum turbulent diffusion coefficient and turbulent Prandtl number; obtaining an improved boundary layer scheme based on the improved mixing length, the improved minimum turbulent diffusion coefficient and the improved turbulent Prandtl number; evaluating the prediction effect of the radiation fog process according to the improved boundary layer scheme to obtain an evaluation result. The present invention improves the simulation ability of the boundary layer scheme for the stable stratified turbulent exchange mechanism by improving the parameterized expression of key turbulent parameters in the boundary layer scheme. The improved parameterization scheme is implanted into the WRF model, and the stable boundary layer processes such as haze are used as the test objects to evaluate the simulation ability and improvement effect of the new scheme on the stable boundary layer process. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0023] Figure 1 A schematic diagram of a turbulence parameterization method for improving stable boundary layer simulation provided by an embodiment of the present invention;

[0024] Figure 2 The turbulent diffusion coefficient variation and statistical diagram provided by the embodiment of the present invention, wherein: Figure 2 (a) The turbulent diffusion coefficient K observed at different heights (40m, 80m, 120m, 160m, 200m) of the meteorological tower during the fog process in December 2016. (The shadow in the figure represents the period when the fog occurred) Figure 2 (b) The statistical results of the turbulent diffusion coefficient K obtained by observing at different heights of the meteorological tower during four fog processes between 2016 and 2019 (the orange curve in the figure represents the average value profile of K at different heights);

[0025] Figure 3 A schematic diagram of non-optimal frequency selection for different boundary layer solutions provided in an embodiment of the present invention;

[0026] Figure 4The different physical and chemical mechanisms for PM2.5 in Tianjin from December 27 to 31, 2017 provided in the embodiments of the present invention 2.5 Schematic diagram of process analysis results for concentration contribution. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0029] like Figure 1 As shown, the present invention provides a turbulence parameterization method for improving stable boundary layer simulation, comprising:

[0030] Obtain various boundary layer observation data;

[0031] Based on the various boundary layer observation data, the vertical profile and time evolution characteristics of key turbulent parameters (turbulent diffusion coefficient, mixing length, turbulent intermittent intensity index, etc.) under stable boundary layer conditions are determined according to multi-source observation methods;

[0032] determining a plurality of boundary layer scenarios based on the vertical profile and the time evolution characteristics;

[0033] improving the mixing length according to wind shear to obtain an improved mixing length;

[0034] The minimum turbulent diffusion coefficient and the turbulent Prandtl number are improved to obtain the improved minimum turbulent diffusion coefficient and the turbulent Prandtl number;

[0035] An improved boundary layer scheme is obtained based on the improved mixing length, the improved minimum turbulent diffusion coefficient and the improved turbulent Prandtl number;

[0036] The prediction effect of the radiation fog process is evaluated according to the improved boundary layer scheme to obtain an evaluation result.

[0037] Furthermore, the various boundary layer observation data include:

[0038] Meteorological tower data, ground weather station data, microwave radiometer data, wind profiler radar data, and sounding data.

[0039] The ultimate purpose of the present invention is to observe and obtain the vertical profile and time evolution characteristics of key turbulence parameters (turbulent diffusion coefficient, mixing length, turbulence intermittent intensity index, etc.) under stable boundary layer conditions based on multi-source observation means such as 15 layers of gradient meteorological elements (temperature, humidity, wind) of the meteorological tower and 5 layers of eddy-related turbulence elements (turbulent kinetic energy). Based on this fact, by improving the turbulence parameterization expression of the boundary layer scheme in the numerical weather and air quality model, the numerical model for the stable boundary layer weather process (such as fog, heavy pollution process, etc.) The accuracy of numerical forecasts is improved. On this basis, the boundary layer structure profile, macroscopic characteristics, and turbulence key parameter observation results under stable stratification conditions obtained based on multi-source observation means are used to comprehensively test the simulation effect of the improved boundary layer scheme, and evaluate the model improvement effect.

[0040] The boundary layer parameterization schemes for numerical weather and air quality models are mainly responsible for solving the closure problem of vertical turbulent flux terms, arbitrary scalar or momentum The vertical turbulent flux term can generally be expressed as:

[0041]

[0042] It can be seen from the above formula that the calculation of heat, water vapor and momentum vertical turbulent exchange flux is based on the heat and water vapor turbulent exchange coefficient K h (i.e. vorticity diffusion coefficient) or momentum turbulence exchange coefficient K m (i.e. vorticity viscosity coefficient) and the vertical gradient of the corresponding variable The product of is determined by γ, which is the inverse gradient term. According to the different closed framework principles, boundary layer schemes can generally be divided into first-order closure, 1.5-order closure or higher-order closure. The closure order is determined by the highest-order prediction equation introduced. The first-order closure describes the turbulent flux term as a function of known quantities, and the 1.5-order closure introduces a prediction equation for some second-order moments (such as turbulent kinetic energy) in the turbulent flux calculation equation. In the first-order nonlocal boundary layer scheme, the momentum turbulent diffusion coefficient (K m ) can be expressed as:

[0043]

[0044] Where k is the von Karman constant, w s is the characteristic velocity of the mixing layer, h is the boundary layer height, and p is the morphological parameter (generally taken as 2). On this basis, the heat turbulent diffusion coefficient (K h ) can be obtained from the momentum turbulent diffusion coefficient (K m ) and the turbulent Prandtl number P r Sure:

[0045]

[0046]

[0047] P r0 =1.272,ε=0.1

[0048] In the 1.5-order local closed boundary layer scheme, the turbulent kinetic energy related term q (q = (2·TKE) 1 / 2 =QKE 1 / 2 ), the turbulent mixing length scale (i.e., mixing length, l), and the momentum and thermal stability functions (S m ,S h ) to calculate the heat turbulence diffusion coefficient (K h ) and momentum turbulent diffusion coefficient (K m ) value:

[0049] K h,m =lqS h,m (5)

[0050] From the above formula, it can be seen that the mixing length l has an important influence on the parameterization calculation of turbulent flux. In the boundary layer parameterization scheme, the turbulent mixing length scale is jointly affected by the surface friction, the boundary layer turbulence intensity and the buoyancy (thermal) effect. In the calculation formula, l is determined by three components, namely: the surface friction length scale l s , turbulence length scale l t and the buoyancy length scale l b The value of the turbulent mixing length scale l is equal to the harmonic mean of the above three components:

[0051]

[0052] Among them, the surface friction length scale l s , turbulence length scale l t and the buoyancy length scale l b They can be expressed as:

[0053]

[0054]

[0055]

[0056]

[0057] Among them, α 1 is the empirical constant (=0.23), N is the Brunt-Vaisala frequency q c is the turbulent velocity scale (=[(g / θ 0 )<ω'θ' v >glt ] 1 / 3 ), z i is the boundary layer height, and Δz is the entrainment layer thickness. BL is the BouLac mixing length, which characterizes the vertical upward or downward displacement (l) of an air mass under the influence of buoyancy (related to the potential temperature θ) after it obtains a certain amount of initial turbulent kinetic energy (e). up and l down ), l up and l down Defined as:

[0058]

[0059]

[0060] l BL By l up and l down The geometric mean is determined by:

[0061] l BL =(l up 2 +l down 2 ) 1 / 2 (13)

[0062] In the parameterization calculation of the mixing length of the current boundary layer scheme, only the atmospheric stability, the average turbulent scale of the boundary layer, and the influence of buoyancy (thermal turbulence) are considered, but the role of wind shear (mechanical turbulence) is ignored. Wind shear is the main source of turbulent energy in the stable boundary layer, and the mixing length is a key parameter that directly characterizes the mixing ability of turbulence. It is very necessary to introduce the contribution of wind shear in its parameterization calculation.

[0063] Specific improvements to the turbulence parameterization of the boundary layer scheme include:

[0064] (1) Introducing the wind shear effect into the boundary layer scheme turbulent mixing length expression (Eqs. 11 and 12) Term, where e is the turbulent kinetic energy term, S(z') is the vertical wind shear, and C 0 is an empirical parameter):

[0065]

[0066]

[0067]

[0068] Among them, the empirical parameter C 0The value of can be obtained based on the fitting of meteorological tower gradient observations, large eddy simulation or calculation results of empirical formulas in literature.

[0069] Based on equation (14-16), l upnew and l downnew The geometric mean of the mixing length l can be obtained by comprehensively considering the buoyancy-shear effect. BS (Buoyancy-Shearbased mixing length, referred to as BS mixing length):

[0070] l BS =(l upnew 2 +l downnew 2 ) 1 / 2 (17)

[0071] On this basis, the improved mixing length l that comprehensively considers the buoyancy-shear effect is BS Introduced into the expression of turbulent mixing length l (Eq. 6), and further modified the original surface friction length scale l s , turbulence length scale l t and the buoyancy length scale l b The three-term coupling method corrects the calculation of the main length scale l from the harmonic mean of the three quantities to the surface friction length scale l s and the turbulence length scale l t The harmonic mean and buoyancy-shear mixing length l BS The smaller value of , that is:

[0072]

[0073] The modified three-term coupling method of turbulent mixing length can better reflect the vertical profile characteristics of turbulence under stable atmospheric stratification conditions, and realizes the display of the influence of wind shear on mechanical turbulence in the parameterized calculation of turbulent mixing length (i.e., buoyancy-shear mixing length l BS Item). The improved turbulent mixing length calculation formula (Equation 18) replaces the original turbulent mixing length calculation formula (Equation 6) and is substituted into the turbulent diffusion coefficient and vertical turbulent exchange flux calculation formula (Equations 1 and 5), and the improved turbulent diffusion coefficient and vertical turbulent exchange flux expressions are obtained:

[0074] K new = l new Q h,m (19)

[0075]

[0076] Therefore, it is possible to take into account the effect of wind shear mechanical turbulence in the calculation of turbulent diffusion coefficient and turbulent vertical exchange flux.

[0077] (2) Introducing the large eddy simulation mixing length l in the parameterized calculation of the turbulent mixing length in the boundary layer scheme LES and the weight function P related to the model grid resolution TKE , realizing the adaptive transition from mesoscale to large eddy scale in the parameterized calculation of turbulent mixing process in numerical model.

[0078] In large eddy simulation (LES) scale, the Deardorff (1974) mixing length is usually used:

[0079]

[0080] In order to introduce the large eddy scale mixing length l into the parameterization expression of turbulent mixing length in boundary layer scheme, LES The influence of the model level grid size is further defined as a weight function P TKE :

[0081]

[0082] Among them, dxdh is a function related to the horizontal grid scale of the numerical model and the boundary layer height. On this basis, based on the weight function P TKE Implement the original boundary layer scheme (mesoscale) mixing length l meso (i.e., Equation 6) and the large eddy mixing length l LES The combination of (Equation 18):

[0083] l SA =P TKE l meso +(1-P TKE ) LES (twenty three)

[0084] The improved scale-adaptive mixed length l SA Substituting the (mesoscale) mixing length l in Equation 5, we have:

[0085] K h,m = l SA Q h,m (twenty four)

[0086] Thus, the parameterized calculation of turbulent mixing length is realized under different horizontal grid resolution conditions, which is obtained by changing the mesoscale mixing length l meso Mixing length l LES Adaptive calculation and transition.

[0087] On this basis, we further consider the influence of the horizontal grid scale (i.e., scale adaptivity) on non-local turbulence. Based on the method of constructing a weight function P that depends on the horizontal grid scale, we introduce a weight function P that depends on the horizontal grid scale in the calculation of non-local turbulent fluxes:

[0088]

[0089]

[0090] Where, the subscript SA represents the result after considering the scale-awareness improvement, the superscript NL represents the mass flux term (i.e., non-local term), P <ω'q'> and P <ω'θ'> Represent the weighting functions for water vapor (q) and heat (θ) mass fluxes respectively:

[0091]

[0092]

[0093] and That is, the water vapor (q) and heat (θ) mass flux (non-local flux) terms after the improvement considering the scale adaptive mechanism.

[0094] On this basis, in order to expand the influence of the analytical part in the calculation of turbulent mixing length (i.e. reduce the influence of the parameterized part), a new scale weight function P is given. TKE2 , that is, the weight of the direct analytical part in the turbulent mixing length is increased:

[0095]

[0096] Substituting equation (29) into equation (23), we can obtain the scale-adaptive turbulent mixing length expression after improving the direct analytical turbulence influence weight.

[0097] (3) Adjust the value of the minimum turbulent diffusion coefficient Kzmin of the boundary layer scheme to improve the ability of the air quality model to simulate atmospheric pollutant concentrations under stable boundary layer conditions.

[0098] In the boundary layer scheme, the calculation equation for the heat turbulence diffusion coefficient is:

[0099]

[0100] In the formula, K zminThe preset lower limit (i.e., minimum threshold) of the turbulent diffusion coefficient in the air quality model is used to avoid the abnormal result of "no turbulence atmosphere" (i.e., K = 0) in the calculation of the strong stable boundary layer. zmin =0.01m 2 ·s -1 ) is adjusted to 1m 2 ·s -1 , which can help improve the numerical simulation capability of air quality models for atmospheric pollutant concentrations under stable boundary layer conditions.

[0101] (4) Adjust the boundary layer scheme turbulent Prandtl number P r The value of (i.e., P in Equation 3) r values), improve the boundary layer meteorological elements (temperature, humidity, wind) and pollutant concentration (PM 2.5 ) vertical profile simulation capability. By setting different turbulent Prandtl numbers P r Numerical sensitivity tests were conducted to evaluate the effects of different turbulent Prandtl numbers P on the meteorological elements of the stable boundary layer and the radiosonde observations of pollutant concentrations (under heavy pollution weather). r The influence of the value on the simulation accuracy of the vertical profile of the meteorological and pollutant concentrations in the stable boundary layer, and the acquisition of the turbulent Prandtl number P of the numerical model that is more suitable for simulating the stable boundary layer during heavy pollution processes r The numerical model's ability to simulate the vertical profiles of stable boundary layer meteorology and pollutant concentrations is improved by taking values.

[0102] Specifically, the vertical profile of the stable boundary layer meteorological elements (corresponding to Figure 1 Observational facts on the temporal and spatial evolution of the macroscopic characteristics and key parameters of turbulence:

[0103] Multiple fog and haze low visibility weather processes in Tianjin from 2016 to 2019 were selected as typical cases of stable boundary layer. Using the 15-story gradient meteorology, eddy-related turbulent fluxes, visibility, ground meteorological station and field sounding observations of the 255m meteorological tower in Tianjin, the spatiotemporal variation characteristics of key turbulent parameters (turbulent diffusion coefficient K, mixing length L and intermittent intensity index LIST) under stable stratification conditions, as well as the vertical profiles of boundary layer meteorological elements (boundary layer temperature, humidity, wind speed, wind direction and other meteorological elements including the vertical profile distribution structure of pollutant concentrations such as PM2.5 and PM10) and macroscopic characteristics of weather processes (horizontal fog area range, vertical fog layer thickness, pollution area range, vertical thickness of pollution layer, formation and dissipation time of fog or pollution process and boundary layer height under fog or heavy pollution weather (typical process of stable boundary layer)) were analyzed. The results show that in terms of time variation trend, the diffusion coefficient K and mixing length L both show a trend of high during the day and low at night, indicating that the turbulent mixing effect is stronger during the day (convective boundary layer) than at night (stable boundary layer). From a vertical perspective, the diffusion coefficient K and mixing length L increase first and then decrease with increasing altitude. The maximum value in the vertical direction often occurs at an altitude of about 120m. At this altitude, the maximum value of the diffusion coefficient K during the day can reach about 160m2 / s, while K can be reduced to <0.1m under stable conditions at night. 2 / s( Figure 2 ).in, Figure 2 (a) The temporal variation trend of the turbulent diffusion coefficient K observed at different heights (40m, 80m, 120m, 160m, and 200m) of the meteorological tower during the fog process in December 2016 (the shadow in the figure represents the fog occurrence period). Figure 2 (b) The statistical results of the turbulent diffusion coefficient K obtained by observing at different heights of the meteorological tower during four fog processes between 2016 and 2019 (the orange curve in the figure represents the average value profile of K at different heights).

[0104] Under stable stratification conditions, atmospheric turbulence in the near-surface layer often exhibits strong intermittency. The intermittent nature of atmospheric turbulence can lead to an overestimation of turbulence parameters, that is, an overestimation of the atmospheric diffusion capacity in the numerical model. This study uses the Hilbert-Huang transform (HHT) technique to separate and detect intermittent signals of atmospheric turbulence in the boundary layer, and quantitatively characterizes the strength of turbulence intermittency based on the LIST index. The results show that the intensity of turbulence intermittency also shows a temporal evolution characteristic of high during the day and low at night. The minimum value of the LIST index can reach about 0.60, indicating that the model overestimates nighttime turbulence most seriously, with an overestimation of about 40%.

[0105] Specifically, the simulation performance of the stable boundary layer (fog) process is evaluated by different boundary layer schemes and different horizontal grid resolutions:

[0106] (1) Evaluation of different boundary layer schemes on fog simulation performance

[0107] Thirteen heavy fog cases in North China from 2016 to 2022 (including 7 land fog and 6 sea fog) were selected to compare and evaluate the pros and cons of five boundary layer parameterization schemes in scoring the forecasting ability of heavy fog processes. The selected boundary layer schemes include three non-local closure schemes: the YSU scheme, the ACM2 scheme (asymmetric convection scheme) and the SH scheme; and two local closure schemes: the MYNN scheme and the MYJ scheme. The scoring indicators for the inspection of horizontal fog areas include: hit rate POD, false alarm rate FAR and TS score; the scoring indicators for the inspection of ground meteorological elements during the fog process include: root mean square error RMSE and consistency index IoA. The results show that the simulation of heavy fog processes in North China is very sensitive to the selection of different boundary layer schemes ( Figure 3 ), Figure 3 Sea fog is orange, and land fog is blue. Among them, (MYNN: Mellor-YamadaNakanishi and Niino scheme; YSU: Yonsei University scheme; MYJ: Mellor-Yamada-Jajic scheme; SH: Shin-Hong scheme) for land fog, the non-local scheme YSU and the local schemes MYNN and MYJ have relatively good simulation capabilities; for sea fog, the simulation performance of the non-local scheme YSU is significantly better than other schemes. The main reason for the low accuracy of fog process forecasting comes from the fog formation and dissipation stages. In the early stage of fog formation, the coverage area is prone to obvious over-reporting or under-reporting; after sunrise, the simulated fog dissipation process is generally early and fast, resulting in a large number of under-reporting; for the mature stage of fog development, the forecast skill score is generally relatively high. For the same boundary layer scheme, there are also obvious differences in the simulation capabilities for different types of fog processes. Overall, the model has a relatively good forecasting ability for advection fog, but a relatively poor forecasting ability for radiation fog.

[0108] (2) The impact of different grid resolutions on fog prediction capabilities

[0109] Based on the mesoscale numerical weather model WRF, the comprehensive effects of different horizontal grid resolutions from ten kilometers to one hundred meters (15km, 5km, 3km, 2.5km, 1km, 500m) on the forecast skill scores of four heavy fog processes in North China from 2016 to 2019, as well as the horizontal\vertical fog layer development and the vertical profiles of meteorology\turbulence in the fog were analyzed. The results show that the change of horizontal resolution from ten kilometers to one hundred meters will have a significant impact on the fog forecast skill score, but with the increase of resolution, the model's forecast skill for heavy fog is not improved in one direction. Among the six resolutions, the 5km resolution has the best fog simulation performance, followed by 2.5km, while the simulation effect score of the 3km to 500m high-resolution nested inner grid is generally lower than that of the outer mother grid, that is, although the high-resolution nested grid simulation improves the grid resolution in key areas, it fails to improve the fog forecast skill score. With the increase of horizontal resolution, the simulated fog area tends to decrease, and this effect is more obvious in the fog formation stage. For the vertical development of the fog layer, the improvement of horizontal resolution is conducive to the simulation of the vertical mixing development of the fog layer. The ground relative humidity and vertical inversion structure simulated at 5km resolution are consistent with the actual situation, which may be the reason why the fog forecast score of 5km resolution is better than that of other resolutions. The turbulent diffusion effect simulated under fine horizontal resolution is generally stronger than that of coarse resolution. Therefore, the vertical mixing of the simulated fog layer at high resolution is often more uniform, which is more conducive to the vertical thickening of the fog layer, while the simulated inversion intensity at low resolution is often stronger, the low-level fog water is denser and the thickness of the fog layer is lower.

[0110] Furthermore, the key turbulence parameters include:

[0111] Turbulent diffusion coefficient, mixing length, and turbulent intermittency intensity index.

[0112] Specifically, the minimum turbulent diffusion coefficient Kzmin is the minimum threshold of the diffusion coefficient K defined in the model to avoid the abnormal result of "no turbulence atmosphere" (i.e. K = 0) in the calculation of the strong stable boundary layer. This threshold will have a significant impact on the simulation of the stable boundary layer at night. A model sensitivity test was carried out using a heavy pollution process in December 2017 as an example. During this process, Tianjin experienced a long-term inversion process with a height of more than 600m and an inversion intensity of 2.9℃·100m. -1 The continuous existence of the inversion layer inhibits the vertical diffusion of pollutants. Based on the model process analysis technology, it is estimated that turbulent mixing reduces the PM2.5 concentration to -30 μg·m during the day. -3 ·h -1 , significantly weakened at night, below -5 μg·m -3 ·h -1 , the ratio is 6:1( Figure 4In order to explore the impact of the change of the minimum turbulent diffusion coefficient in the boundary layer parameterization scheme on the simulation results of near-surface pollutant concentrations, this study evaluated the effects of different minimum turbulent diffusion coefficient values ​​on the surface PM in urban and suburban areas of Tianjin during the pollution process based on the WRF / Chem model. 2.5 Impact of concentration simulation results and comparison with observation results. During the pollution process, the minimum turbulent diffusion coefficient has a greater impact on the simulation of near-ground pollutant concentrations. The increase in the minimum turbulent diffusion coefficient will lead to a decrease in near-ground pollutant concentrations, and this impact has obvious daily variation characteristics, which is more obvious at night. This is mainly because the ground and the lower atmosphere gradually cool down after sunset, the buoyancy source that provides turbulent energy gradually disappears, the stability of the atmospheric stratification is enhanced, and most of the turbulent energy in the nighttime mixing layer comes from wind shear. Therefore, the turbulent mixing at night is weak (approaching the minimum threshold of the mode) and is often obviously intermittent. Therefore, the minimum turbulent diffusion coefficient has a greater impact on the vertical turbulent mixing intensity at night. The increase in the minimum turbulent diffusion coefficient enhances the vertical turbulent mixing at night, and the vertical distribution of pollutants in the atmospheric boundary layer is more uniform, near-ground PM 2.5 The increase in the minimum turbulent diffusion coefficient will also lead to a decrease in daytime PM 2.5 The concentration decreases during the day, but the change is significantly smaller than that at night. This is mainly due to the buoyancy caused by surface heating during the day, which generates turbulence. The turbulence intensity at this stage is usually much higher than the minimum value set in the model. The difference in pollutant concentration during the day mainly comes from the residual difference in pollutant concentration during the previous night.

[0113] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0114] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A turbulence parameterization method for improving stable boundary layer simulation, characterized in that: include: Obtain various boundary layer observation data; Based on the various boundary layer observation data, determine the vertical profile and time evolution characteristics of key turbulent parameters under stable boundary layer conditions according to multi-source observation methods; determining a plurality of boundary layer scenarios based on the vertical profile and the time evolution characteristics; improving the mixing length according to wind shear to obtain an improved mixing length; The minimum turbulent diffusion coefficient and the turbulent Prandtl number are improved to obtain the improved minimum turbulent diffusion coefficient and the turbulent Prandtl number; An improved boundary layer scheme is obtained based on the improved mixing length, the improved minimum turbulent diffusion coefficient and the improved turbulent Prandtl number; The prediction effect of the radiation fog process is evaluated according to the improved boundary layer scheme to obtain an evaluation result.

2. The silicon detector component of the novel packaging structure according to claim 1 is characterized in that: The various boundary layer observation data include: Meteorological tower data, ground weather station data, microwave radiometer data, wind profiler radar data, and sounding data.

3. The silicon detector component of the novel packaging structure according to claim 1 is characterized in that: The key turbulence parameters include: Turbulent diffusion coefficient, mixing length, and turbulent intermittency intensity index.

Images