CN113468482A

CN113468482A - Rainstorm weather simulation forecasting method based on WRF mode

Info

Publication number: CN113468482A
Application number: CN202110761768.0A
Authority: CN
Inventors: 陈东辉; 刘长空; 王亮; 郭刚; 董欢欢; 徐涛; 任林涛
Original assignee: 31440 Unit Of Chinese Pla
Current assignee: 31440 Unit Of Chinese Pla
Priority date: 2021-07-06
Filing date: 2021-07-06
Publication date: 2021-10-01
Anticipated expiration: 2041-07-06
Also published as: CN113468482B

Abstract

The invention provides a rainstorm weather simulation forecasting method based on a WRF mode, which comprises the following steps: (1) carrying out statistical analysis on historical rainstorm weather, and carrying out weather typing; (2) selecting a plurality of cases from each type of rainstorm, and carrying out sensitivity test on different rainstorm simulation parameterization schemes in each type of rainstorm by adopting a WRF (weighted round robin) mode to obtain the optimal parameterization schemes for simulating various types of rainstorm; (3) and (4) performing trial forecast evaluation, and comparing and analyzing a trial forecast result with the existing numerical forecast product to obtain a rainstorm weather simulation parameterization scheme. A WRF mode is adopted to carry out parametric scheme sensitivity test research on different types of rainstorm simulation in the northeast region, and then forecast evaluation inspection is carried out to provide the optimal parametric scheme for numerical simulation of various types of rainstorm in the northeast region.

Description

Rainstorm weather simulation forecasting method based on WRF mode

Technical Field

The invention relates to the technical field of weather forecast, in particular to a rainstorm weather simulation forecasting method based on a WRF mode.

Background

In recent years, there has been considerable research into the climatic characteristics of precipitation anomalies in the northeast region in summer, and it is believed that regional and extensive storms in the northeast region are primarily the result of the interaction of the west wind zone, subtropical zone and tropical circulation systems.

The rainstorm forecast mainly comprises two forms of numerical simulation research and collective forecast, and the numerical forecast achieves quite accurate degree for both situation field forecast and rainfall forecast; the method for forecasting the rainstorm with the integrated power factor in the integrated forecast can release the action heat for the numerical forecast result on the basis of the mode forecast, can correct the mode rainfall forecast correspondingly, and has obvious advantages in the aspect of forecasting the rainstorm falling area.

However, the capability of the existing numerical forecasting mode for forecasting sudden rainstorm weather is weak, the accurate description of the mode physical process and the coordination performance of the power framework are not enough to reflect the dynamic thermal process of the actual rainstorm system development, and the requirements for refinement, fixed point, quantification, no gap and the like of the rainstorm forecast are difficult to realize.

The number of members in ensemble prediction is limited, so that the probability density function of the current atmospheric state cannot be completely and exactly described, but only samples obtained by sampling from the current atmospheric probability density function can be represented, and therefore, it is very critical how to sample the probability density function which can reflect the current atmospheric state as much as possible.

The forecasting forms cannot distinguish a rainstorm weather system, and the forecasting accuracy cannot be guaranteed.

Disclosure of Invention

Aiming at the technical problems, the invention aims to provide a rainstorm weather simulation forecasting method based on a WRF mode, solves the problems that a rainstorm weather system cannot be distinguished, the forecasting accuracy is low and the like, and provides a foundation for the fine forecasting of rainstorm in northeast China.

The invention provides a rainstorm weather simulation forecasting method based on a WRF mode, which comprises the following steps:

(1) carrying out statistical analysis on historical rainstorm weather, and carrying out weather typing;

(2) selecting a plurality of cases from each type of rainstorm, and carrying out sensitivity test on different rainstorm simulation parameterization schemes in each type of rainstorm by adopting a WRF (weighted round robin) mode to obtain the optimal parameterization schemes for simulating various types of rainstorm;

(3) and (4) performing trial forecast evaluation, and comparing and analyzing a trial forecast result with the existing numerical forecast product to obtain a rainstorm weather simulation parameterization scheme.

Further, in the step (1), the historical rainstorm weather includes 199 cases of regional rainstorm and large-scale rainstorm occurring in the northeast region of 1981 and 2017.

Further, in the step (1), the rainstorm type obtained by weather classification comprises: shear type rainstorms, cyclone type rainstorms, cold vortex type rainstorms, and typhoon type rainstorms.

Further, in the step (2), a plurality of cases are selected in each type of heavy rain by analyzing the weather situation in the 500hPa weather map.

Further, in the step (2), the global forecast data for driving the WRF mode adopts forecast data issued by the united states atmospheric environment forecast center.

Further, in the step (2), the WRF mode adopts a double nesting scheme, the center positions of two nesting regions are both located at (46.7 degrees N, 125.0 degrees E), the outer regions are 110-140 degrees E, 30-60 degrees N, and the resolution is 27 km; the internal region is: 115-135E, 34-54N, the resolution is 9km, the vertical direction is divided into 28 layers, and the top of the mode layer is 50 hPa.

Further, in the step (2), the parameterization scheme includes a micro-physics scheme, a long and short wave radiation scheme, a land surface process, a boundary layer scheme and a cloud parameterization scheme, and two schemes are respectively selected in each sensitivity test to form 32 scheme combinations.

Further, in the step (3), the existing numerical forecast product includes a T639 global ensemble forecasting system and a mid-european weather forecast center model.

Further, in the step (3), the inspection data includes hourly 0.1 ° × 0.1 ° precipitation data, 0.1 ° gpm precipitation data and site precipitation data provided by the chinese meteorological science data sharing service network and fused with the CMORPH precipitation product.

Further, in the step (3), the WRF forecast precipitation field is interpolated on a grid which is matched with the live data for scoring inspection.

According to the rainstorm weather simulation forecasting method based on the WRF mode, parametric scheme sensitivity test research is carried out on different types of rainstorm simulation in the northeast region by adopting the WRF mode, and then forecast evaluation inspection is carried out, so that an optimal parametric scheme for numerical simulation of the rainstorm in the northeast region and each type of rainstorm is given. The problem that a rainstorm weather system cannot be distinguished in the existing weather forecasting process can be effectively solved, the forecasting accuracy can be greatly improved, and the rainstorm can be finely forecasted in the northeast region. A solid foundation is laid for issuing rainstorm disaster early warning information in advance and researching rainstorm characteristic rules in northeast regions. The method has the advantages that the rainstorm intensity, the falling area and the occurrence time in the northeast are subjected to fine forecast guarantee, and preparation is made for issuing early warning information of the rainstorm disaster in advance and reducing the influence caused by the rainstorm disaster to the maximum extent.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.

FIG. 1 is a schematic flow chart of a rainstorm weather simulation forecasting method according to the present invention;

FIG. 2 is a research area diagram of the rainstorm weather simulation forecasting method of the present invention;

FIG. 3 is a 500hPa weather chart for a shear type storm at typical individual screening;

FIG. 4 is a 500hPa weather chart for a typical case screening of cyclonic stormwater;

FIG. 5 is a 500hPa weather chart of a typical individual screening of a cold vortex storm;

FIG. 6 is a 500hPa weather chart of a typhoon type rainstorm during typical individual screening;

FIG. 7 is a graph showing the predicted effect score of SH _1 in example 1;

FIG. 8 is a graph of the predicted deviation of SH _1 in example 1;

FIG. 9 is a cumulative precipitation distribution map for SH _1 in example 1;

FIG. 10 is a graph showing the predicted effect score of SH _2 in example 1;

FIG. 11 is a graph of the predicted deviation of SH _2 in example 1;

FIG. 12 is a graph of the cumulative precipitation profile for SH _2 in example 1;

FIG. 13 is a graph showing the predicted effect score of SH _3 in example 1;

FIG. 14 is a graph of the predicted deviation of SH _3 in example 1;

FIG. 15 is a graph of the cumulative precipitation distribution of SH _3 in example 1;

FIG. 16 is a graph showing the score of the forecast effect for each rainstorm with a precipitation threshold of 50 mm.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. The components of embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.

Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Some embodiments of the invention are described in detail below with reference to the accompanying drawings. The embodiments described below and the features of the embodiments can be combined with each other without conflict.

The rainstorm weather simulation forecasting method based on the WRF mode mainly carries out statistical analysis on rainstorm disastrous weather in the northeast region from 1981 to 2017, summarizes and summarizes the types and the characteristics of the rainstorm disastrous weather in the region, further carries out simulation research on various rainstorm weather historical samples by applying the mesoscale WRF mode, and calculates various localized physical parameterization schemes with the best rainstorm forecasting effect in the northeast region through designing historical sample simulation tests.

The technical route of the research of the rainstorm weather simulation forecasting method is shown in figure 1.

The method mainly comprises the following steps:

(3) and (4) performing test forecast evaluation, comparing and analyzing a test forecast result with the existing numerical forecast product, and respectively performing quantitative inspection on the mode calculation efficiency, the operation time and the forecast effect to obtain a rainstorm weather simulation parameterization scheme.

The area researched by the rainstorm weather simulation forecasting method based on the WRF mode is shown in fig. 2, wherein the color filling represents the terrain height, and the dotted distribution is the station precipitation distribution position.

The WRF mode adopts a double nesting scheme, the center positions of two layers of nesting areas are both positioned at (46.7 degrees N and 125.0 degrees E), the outer area is 110-140 degrees E and 30-60 degrees N, and the resolution is 27 km; the internal region is: 115-135E, 34-54N, the resolution is 9km, the vertical direction is divided into 28 layers, and the top of the mode layer is 50 hPa.

Through sensitivity test research, an optimal parameterization scheme for simulating different types of rainstorms is formed.

The rainstorm weather simulation forecasting method based on the WRF mode mainly comprises the following steps:

statistical analysis is carried out on rainstorm disastrous weather in the northeast region of China since 1981, the types and the characteristics of the rainstorm weather are summarized, and the weather classification is carried out on 199 regional rainstorms and large-scale rainstorms which appear in the northeast region of 1981-2017. The rainstorm weather conditions in the northeast region can be divided into 4 major categories, namely shear type rainstorm, cyclone type rainstorm, cold vortex type rainstorm and typhoon type rainstorm;

selecting 2-3 typical examples of each type of rainstorm, performing sensitivity test research on a rainstorm simulation parameterization scheme in each example by adopting a WRF mode, respectively selecting 2 schemes from 5 main parameterization schemes, namely a micro-physical scheme, a long and short wave radiation scheme, a land surface process, a boundary layer scheme and a cloud accumulation parameterization scheme, combining 32 schemes in total, performing sensitivity test, and finally providing an optimal parameterization scheme for simulating various types of rainstorms;

and (4) performing trial forecast evaluation and inspection, comparing and analyzing the trial forecast result TS score with the T639 global ensemble forecast system and the European mid-term weather forecast center mode, and respectively performing quantitative inspection on the mode calculation efficiency, the operation time and the forecast effect.

The invention selects forecast data of Global Spectrum Mode (GSM) to be used for a mode drive field, which comprises the following specific steps:

forecasting data issued by a Global Forecasting System (GFS) of the American atmospheric environment forecasting center is selected to carry out a mode driving field, the triangular cutoff wave number of a global spectrum mode is 254, the global Gaussian grid is 768 multiplied by 384 which is approximately equal to 0.5 degrees multiplied by 0.5 degrees, 64 sigma layers are vertically layered, the height from the ground to about 2.7hPa is up, and the time interval is 6 h;

selecting precipitation data of a conventional observation station and an automatic meteorological station in the northeast of China from 1981 to the present as historical reference data;

the method takes the hourly 0.1 degree multiplied by 0.1 degree precipitation data, 0.1 degree gpm precipitation data, site precipitation data and the like fused with the CMORPH precipitation products and provided by the Chinese meteorological science data sharing service network as the detection data.

The following describes the detection data of a selected typical example:

the rainstorm simulation test data of 8/1/2008, 7/20/2010, 7/16/2013 and 9/2010 are hourly 0.1 degree multiplied by 0.1 degree rainfall data which are provided by a Chinese meteorological science data sharing service network and are fused with CMORPH rainfall products;

the inspection data selected for the rainstorm cases of 8 months and 3 days in 2017 is 0.1-degree gpm precipitation data;

the selected test data of heavy rain in 21/8/1997, 21/6/1981, 16/6/1984 and 16/8/1994 are the station precipitation data.

The WRF mode parameterization scheme is debugged, 2 schemes are respectively selected from 5 main parameterization schemes of a micro-physics scheme, a long and short wave radiation scheme, a land surface process, a boundary layer scheme and a cloud parameterization scheme, and a sensitivity test is carried out by combining 32 schemes which are 5 powers of 2.

From the rainstorm types obtained by the early statistical classification, typical representative examples of various types of rainstorm weather are screened out from each type of rainstorm through analyzing the weather conditions in a 500hPa weather map, 2-3 typical examples are selected for each type of rainstorm, sensitivity research of different parameterization schemes is carried out on the typical examples, and the optimal parameterization scheme capable of simulating the type of rainstorm is finally calculated through repeated debugging.

The screening process and standard of each typical example of the type of rainstorm mainly refer to the 500hPa weather chart and the weather situation in fig. 3-6, which specifically includes:

in the selection of the typical example of shear type stormy weather, low grooves appear in a 500hPa weather chart (35-55 degrees N, 110-140 degrees E), the whole middle latitude area of the continental east Asia is in a two-groove one-ridge type, two grooves are respectively arranged in the northeast China area of West Siberian and China, and one ridge is arranged near the Beigal lake, wherein the lowest potential height of the groove area of the northeast Asia is below 5800gpm, the groove line is positioned at the first line of the desert river, the red peak and the Beijing, the groove line continuously develops and moves to the northeast Asia within 12h in the future (the right part in the picture 3), and at the moment, the groove line is positioned at the first line of the Rema, the Changchun and the Dandong, the typical example of shear type stormy weather is selected when the above standard is met.

In a typical example of a cyclonic stormy weather, in the 500hPa weather chart (50 ° -60 ° N, 105 ° -125 ° E), the closed low center of the bagal lake occurs in the eastern to northeastern inner Mongolia, and the central air pressure is less than 550hPa, and the trough line extending from the low center is located in the Manchuria, the red peak and the first line of the rising sun; within 12h in the future (see right part of fig. 4), the low pressure center continues to move to the northeast, the gutter line moves to the trenebel, vinpoch, and dandong lines, and the central air pressure continues to be maintained below 550hPa, which is a typical case for cyclonic stormy weather.

In the typical example selection of the cold vortex type rainstorm weather, a closed contour line appears in a 500hPa weather chart (45-55 degrees N, 105-125 degrees E), the central air pressure value is below 560hPa, a cold center or an obvious cold groove is matched, the life history is maintained for at least 3d, and the low-pressure center continuously moves to the northeast area within 12h in the future (see the right part in figure 5), and the low-pressure center is defined as the typical example of the cold vortex type rainstorm weather when the strength is maintained unchanged.

In the typical example selection of the typhoon type rainstorm weather, in a 500hPa weather chart, the typhoon center is located in the yellow sea area, and within 12h in the future (see the right part in fig. 6), the typhoon continues to land near the continental area in the north, at the moment, the airflow in the middle latitude west wind is in the range of 35-55 degrees N and 100-140 degrees E, after the landing of the typhoon system is weakened, the typhoon system is captured by the middle latitude west wind system, and under the action of a west wind zone, the typhoon type rainstorm weather continues to move to the northeast, so that the typical example of the typhoon type rainstorm weather is selected.

It should be noted that, in the invention, the WRF forecast precipitation field is interpolated on the grid matched with the live data for scoring inspection, only the d02 area is inspected at this time, and the precipitation on the ocean is not inspected and analyzed.

In the invention, the micro-physical scheme of the WRF mode selects a Thompson grapnel scheme and a WSM 6-class grapnel scheme, and the modes are respectively represented by

numbers

8 and 6; the long-short wave radiation scheme adopts an RRTM long wave radiation scheme and a Goddard short wave radiation scheme, and the long-short wave radiation scheme and the Goddard short wave radiation scheme are respectively represented by

numbers

4 and 5 in the modes; the land process adopts a unified Noah scheme and a RUC scheme, which are respectively represented by the numbers 2 and 3 in the mode; the boundary layer scheme adopts a YSU scheme and a Mellor-Yamada-Janjic (eta) TKE scheme, and the numbers are respectively represented by 1 and 2 in the mode; meanwhile, when the boundary layer scheme is a YSU scheme, the near-ground scheme is a Monin-Obukhov scheme, and when the boundary layer scheme selects a Mellor-Yamada-Janjic (eta) TKE scheme, the near-ground layer scheme is correspondingly a MYJ Monin-Obukhov scheme; the cloud parameterization scheme adopts a Betts-Miller-Janjic scheme and a Tiedtke scheme, and is respectively represented by

numbers

1 and 6 in the mode; the different parameterization schemes described above are explained below.

Wherein, the Thompson graupel scheme improves the earlier Reisner scheme, and adopts cooper formula to replace the ice crystal nucleation process of the Fletehe curve; the Walkoet is utilized in the automatic conversion process instead of the original Kessler process; aragonite replaces the exponential distribution process with a generalized gamma distribution; the drag coefficient of the aragonite is determined by the mixture ratio of the aragonite rather than by a constant; according to the 3 rd modification, the freezing growth of snow exceeds the sublimation growth of snow before the freezing of snow is converted to aragonite; using a drag coefficient of a dimensional distribution of snow subject to temperature; the distribution of the rain scale drag coefficient is subject to the magnitude of the rain mixing ratio and can therefore be used to calculate the rain and small rain drop landing process.

The WSM 6-class grapnel protocol adds the predictor variable, aragonite, and some processes associated with it, to 6 predictors of water phase species, making the microphysical process more complex. The saturation adjustment of the WSM 6-class graupel scheme separately processes the ice and water saturation process according to the schemes of Dudhia and Hong et al, optimizes the calculated magnitude, reduces the sensitivity of the scheme to the time step of the mode, and is suitable for high-resolution simulation.

Radiation transmission for the RRTM long wavelength radiation scheme the flux and cooling rate for the atmospheric long wavelength spectral domain (10-3000cm-1) were calculated using the K correlation method (the relevant K method). Molecular species considered for the model include water vapor, ozone, carbon dioxide, methane, nitrogen dioxide, and halocarbons. The K distribution is obtained directly from the LBRTM line-by-line mode, which provides the absorption coefficient required by the RRTM, and a look-up table is preset to accurately represent the long wave process due to absorption of the molecular species described above.

The Goddard short wave radiation scheme includes an atmospheric circulation mode, a meso-scale mode, and a cloud mode. It calculates the solar radiation flux due to the absorption of water vapor, ozone, carbon dioxide, oxygen, clouds and aerosols, as well as due to the scattering of clouds, aerosols and various gases.

The unified Noah scheme may operate alone as a one-dimensional single-point mode or may be coupled to an atmospheric mode. The mode has 4 layers of soil (0.1, 0.4, 1.0 and 2.0m), the soil heat conduction equation is adopted for calculating the soil temperature, the Richard equation is adopted for calculating the soil humidity, and the simple water balance method is adopted for calculating the runoff. The method adopts a finite difference space splitting method and a Crank-Nicholson time integration scheme when numerical integration is carried out on a control equation. The unified Noah scheme can forecast the influence of soil icing and snow accumulation, improves the capability of processing urban ground, and considers the property of a ground emitter.

The RUC scheme contains 6 soil layers and 2 snow layers, and carefully considers the soil icing process, the temperature and density difference of uneven snow and snow, the influence of frozen soil and snow cover in the energy and moisture transmission process, the vegetation effect and canopy water.

The YSU scheme represents non-locally induced flux with an inverse gradient term, explicitly dealing with the wraparound layer on top of the planet boundary layer. Entrainment was expressed as a quantity proportional to the surface buoyancy flux from the study of the large scale vortex mode.

The Mellor-Yamada-Janjic (eta) TKE scheme replaces the 2.5 order turbulent closed model of Mellor-Yamada with a boundary layer and turbulent parameterization in free atmosphere, which predicts turbulent kinetic energy with local vertical mixing. The scheme calls a SLAB (thin layer) mode to calculate the temperature of the ground; exchange coefficients were calculated using similar theory before SLAB and vertical flux was calculated using the implicit diffusion scheme after SLAB. When the boundary layer scheme is the scheme, the near-ground layer scheme generally corresponds to the MYJ monon-Obukhov scheme.

The Betts-Miller-Janjic scheme performs relaxation adjustment on the thermal profile at a given time interval, and the convective mass flux can consume a certain effective buoyancy force during the relaxation time. The scheme determines the buffering time and the convection profile according to the cloud effect representing the characteristics of the drainage basin, and modifies the trigger mechanism to adapt to higher horizontal resolution.

The Tiedtke scheme is a mass flux type scheme, and under the condition of removing a time scale, the shallow convection effective potential energy and momentum transmission are considered, so that the phenomena of deep convection, a Xinfeng yun area and subtropical organized convection can be effectively described.

In the WRF mode, FNL data is used as the driving field, and when the selected instance is too early, the driving field data is ERA-Interim data. By adopting a double nesting scheme, the central positions of two layers of nesting areas are both located at (46.7 degrees N, 125.0 degrees E), the resolution of a coarse grid is 27km, the resolution of a fine grid is 3km, the map projection adopts a Labert projection, 27 unequally-spaced sigma layers are arranged in the vertical direction, the top of a mode layer is selected to be 50hPa, the integration step length is 120s, and 12h before the mode integration is taken as spin-up time.

The settings of the specific test parameterization scheme of the invention are shown in the following table:

in four different types of rainstorm weather, 2-3 typical examples of each type of rainstorm are selected, and 9 rainstorm processes are selected in total, wherein the specific conditions are shown in the following table:

in the trial forecasting evaluation, the evaluation is mainly performed for a forecasting field and an observation field, and the following statistics are used to evaluate the forecasting effect. The following are the calculation formulas for hit rate forecast score (TS), Missing report rate (MR), empty report rate (FAR), and forecast deviation (FBI), respectively:

the definitions of a, b, c and d in the formulas (1) to (4) are shown in the following tables:

in the formula (1), TS is used to evaluate the forecast effect of a precipitation event satisfying a certain precipitation threshold, the value range is 0-1, when the precipitation event forecast is accurate (b is 0, c is 0), the value is equal to 1, and the forecast effect is the best; when the precipitation event forecast is inaccurate (a is 0), its value is 0, and there is no forecasting skill.

In the formula (2), MR represents the proportion of the region which is missed in the actual precipitation region occupying all the actual precipitation regions, the value range is 0-1, and the smaller the value is, the better the value is.

In the formula (3), the FAR accounts for the proportion of the area without actual precipitation in the forecast precipitation area to the total forecast precipitation area, the value range is 0-1, and the smaller the value is, the better the value is.

In the formula (4), the FBI is mainly used for measuring the forecast deviation of the mode to a certain magnitude of precipitation, and the score is numerically equal to the ratio of the total lattice points meeting a certain precipitation threshold value in the forecast region to the total lattice points corresponding to the actual precipitation, so as to reflect the total forecast effect of the precipitation. When no empty report or missing report exists (b is 0, c is 0), the value is equal to 1, and the forecast is completely accurate; when the missing report is more than the empty report times (c is more than b), the value is less than 1, and the report is insufficient; when the empty report is more than the report missing times (b is more than c), the value is more than 1, and the report is over-forecast.

The SH _1, SH _2, CY _2 and VO _2 rainstorm test inspection data are the precipitation fusion products of an automatic station and CMORPH hours, the resolution ratio of which is 0.1 degree multiplied by 0.1 degree, the inspection data for SH _3 rainstorm is 0.1 degree gpm precipitation data, and the station precipitation data for CY _1, VO _1, TY _1 and TY _2 rainstorm. The WRF forecast precipitation field is interpolated on a grid which is matched with live data, comparison and analysis are further carried out, only the d02 area is detected at this time, and precipitation on the ocean is not detected and analyzed.

Example 1

In this embodiment, sensitivity test research is performed on a simulation parameterization scheme of shear type rainstorm according to the rainstorm weather simulation forecasting method based on the WRF mode, and a description is performed.

In this embodiment, 3 cases of 1/8/2010, 20/2010, and 3/8/2017 are selected, respectively.

The SH _1 is a case of 8/1/2008, and rainfall observation data of the case is a fusion product of an automatic station and CMORPH hour rainfall, and FNL data is used as a WRF mode driving field. The simulation time is 0000UTC from 31/7/8/1/1200 UTC in 2008, and the first 12h is Spin-up time.

In order to compare the simulation effect of each parameterized scheme on the example, the relevant statistics are calculated according to the formulas (1) to (3), and the selected precipitation thresholds are 10mm, 25mm and 50mm respectively, which correspond to medium rain, heavy rain and heavy rain respectively.

Referring to fig. 7, the abscissa is 32 scenarios, the first to third columns of pictures represent TS score, MR score and FAR score, respectively, and the first to third rows represent precipitation forecast scores corresponding to 10mm, 25mm and 50mm of precipitation threshold, respectively. Different bars represent different mean forecast scores. The box represents the optimal parameterization scheme for this case.

It can be seen that the rainstorm TS scores of the 32 schemes in this simulation are all lower than 0.26, which is less than the forecast scores of T639, forecaster and EC modes, wherein the scores of the schemes 1, 7, 17, 18 and 21 are higher. The scores for protocol 7, protocol 8 and protocol 11 were lower in the MR score. The 32 solutions in FAR score were overall higher than the predictor, T639 and EC mode predictor scores, with lower scores for solution 1, solution 7, solution 18 and solution 21. In summary, although the rainstorm TS score of the scenario 18 is not the highest, the null report rate and the missing report rate of the scenario 18 are the lowest, so the simulation result of the scenario 18 in SH _1 is the best.

For visually analyzing the total simulation effect of rainfall meeting a certain rainfall threshold in the 32 scheme forecast areas in SH _1, the FBI value is calculated according to the formula (4), and the selected rainfall thresholds are 10mm, 25mm and 50mm respectively, which correspond to medium rain, heavy rain and heavy rain respectively.

Referring to fig. 8, (a) the forecast deviation corresponding to 10mm is taken for the precipitation threshold, (b) the forecast deviation corresponding to 25mm is taken for the precipitation threshold, (c) the forecast deviation corresponding to 50mm is taken for the precipitation threshold, and the middle horizontal line represents that the FBI value is 1.

Compared with actual precipitation, in the medium rain forecast, underestimation exists in 10 scheme simulation forecast results, the FBI values of the scheme 10 and the scheme 28 are close to 1, and overestimation exists in 20 scheme simulation forecast results, wherein underestimation exists in the scheme 18; in heavy rain and rainstorm forecasting, the over-estimation phenomenon exists in the simulation forecasting results of 32 schemes.

The shear type northeast rainstorm process in SH _1 mainly focuses on 1200UTC from 31 days of 7 months to 1200UTC from 1 day of 8 months in 2008. fig. 9 shows the 0-12h cumulative precipitation, 12-24h cumulative precipitation and 0-24h cumulative precipitation simulated by the optimal scheme (scheme 18) of the 32 schemes. (a1) - (a3) is live precipitation, and (b1) - (b3) are forecast results corresponding to the optimal parameterization scheme of the example, and the first column to the third column respectively represent 0-12h accumulated precipitation, 12-24h accumulated precipitation and 0-24h accumulated precipitation.

As can be seen, the live precipitation falling area is in a block shape, the rainstorm centers in 24h accumulated precipitation are mainly distributed in Liaoning and Jilin, and the maximum precipitation can reach more than 60 mm. In the simulation result of the scheme 18, the rainstorm center of 24h accumulated precipitation extends to the northeast of Heilongjiang south, and the maximum precipitation can reach more than 80 mm. In contrast to the live situation, the rainstorm centre position of the scenario 18 simulation result is substantially consistent with the observation, but the local area rainfall is too great, the rainstorm centre range extends in the northeast-southwest direction, and the rainstorm intensity is higher than the live situation. The precipitation simulation result of the scheme 18 for the northern part of Heilongjiang is greatly different from the actual situation.

An example of 20 days in 7 months in 2010 is called SH _2 for short, precipitation observation data is a fusion product of an automatic station and CMORPH hour precipitation, and FNL data is used as a WRF mode driving field. The simulation time is from 1200UTC at 19 days 7-7 months to 0000UTC at 21 days 7-7 months in 2010, and the first 12h is Spin-up time.

In order to visually analyze the simulation effect of each parameterization scheme on the example, the related statistics are calculated according to the formulas (1) to (3), and the selected precipitation threshold values are 10mm, 25mm and 50mm respectively, which correspond to medium rain, heavy rain and heavy rain respectively.

As can be seen in fig. 10, there are 11 scenarios with rainstorm TS scores greater than 0.32, higher than forecaster, T639 and EC mode forecast scores. The rainstorm MR, FAR scores for the 32 scenarios were overall lower than the predictor, T639, and EC mode prediction scores, with the lowest MR, FAR score for scenario 14. In summary, although the rainstorm TS score of the scheme 14 is not the highest, the null report rate and the missing report rate of the scheme 14 are the lowest, so the simulation result of the scheme 14 in the SH _2 is the optimal.

For visually analyzing the total simulation effect of rainfall meeting a certain rainfall threshold in the 32 scheme forecast areas in SH _2, the FBI value is calculated according to the formula (4), and the selected rainfall thresholds are 10mm, 25mm and 50mm respectively, which correspond to medium rain, heavy rain and heavy rain respectively.

Referring to fig. 11, (a) the forecast deviation corresponding to 10mm is taken for the precipitation threshold, (b) the forecast deviation corresponding to 25mm is taken for the precipitation threshold, (c) the forecast deviation corresponding to 50mm is taken for the precipitation threshold, and the middle horizontal line represents that the FBI value is 1.

Compared with actual rainfall, the simulation forecast results of 32 schemes have an overestimation phenomenon in the medium rain forecast and the heavy rain forecast; in the rainstorm forecast, 6 scheme simulation forecast results have underestimation phenomena, and 26 scheme simulation forecast results have overestimation phenomena, wherein the scheme 14 has overestimation phenomena.

The shear type northeast rainstorm process in SH _2 mainly focuses on 20 days 0000UTC to 21 days 0000UTC of 7 months in 2010, and the 0-12h accumulated precipitation, the 12-24h accumulated precipitation and the 0-24h accumulated precipitation are simulated by the optimal scheme (scheme 14) in the 32 schemes shown in FIG. 12. (a1) - (a3) is live precipitation, and (b1) - (b3) are forecast results corresponding to the optimal parameterization scheme of the example, and the first column to the third column respectively represent 0-12h accumulated precipitation, 12-24h accumulated precipitation and 0-24h accumulated precipitation.

It can be seen that the actual precipitation falling area is in the southwest-northeast direction, the rainstorms are mainly distributed in Liaoning, Jilin and Heilongjiang, and the maximum precipitation can reach 60 mm. Compared with the actual situation, the scheme 14 can basically simulate the main rain zone and the trend of the precipitation, and the rainstorm center is close to the actual situation but the rainstorm intensity and the range are larger. The scheme 14 has poor effect of simulating the precipitation condition in the north of Heilongjiang.

SH _2 is an example of the data of 8, 3 and 2017, wherein the rainfall observation data is GPM observation data, and FNL data is used as a WRF mode driving field. The GPM observation generally shows stronger precipitation compared with the automatic station and CMORPH hour precipitation fusion product, so that the report missing rate of the current case is higher. The simulation time is 8 months, 2 days, 0000UTC to 8 months, 3 days, 1200UTC in 2017, and the first 12h is Spin-up time.

As can be seen in fig. 13, scenario 17 has a rainstorm TS score of about 0.21, higher than the other 31 scenarios, but lower than the forecaster, T639 and EC mode forecast scores. The rainstorm MR score for scenario 17 was the lowest, approximately 0.74, higher than the predictor, EC mode prediction score, and lower than the T639 prediction score. The rainstorm FAR score for scenario 17 was lower, lower than the predictor, T639, and EC mode prediction scores. In combination with the above considerations, in SH _3, although scheme 17 has a higher null report rate compared to other schemes, scheme 17 results optimally as a whole.

For visually analyzing the total simulation effect of rainfall meeting a certain rainfall threshold in the 32 scheme forecast areas in SH _3, the FBI value is calculated according to the formula (4), and the selected rainfall thresholds are 10mm, 25mm and 50mm respectively, which correspond to medium rain, heavy rain and heavy rain respectively.

Referring to fig. 14, (a) the forecast deviation corresponding to 10mm is taken for the precipitation threshold, (b) the forecast deviation corresponding to 25mm is taken for the precipitation threshold, (c) the forecast deviation corresponding to 50mm is taken for the precipitation threshold, and the middle horizontal line represents that the FBI value is 1.

Compared with actual precipitation, in the medium rain forecast, the underestimation phenomenon exists in the simulation forecast results of 5 schemes, the FBI value of the scheme 4 is close to 1, the overestimation phenomenon exists in the simulation forecast results of 26 schemes, and the overestimation phenomenon exists in the scheme 17; in the heavy rain and rainstorm forecasting, the simulation forecasting results of the 32 schemes have underestimation phenomena, and accord with the characteristic that the precipitation in GPM observation data is stronger, wherein the FBI value of the scheme 17 is closer to 1 than other schemes.

The shear type northeast rainstorm process in SH _3 mainly focuses on 8/month 2/day 1200UTC to 8/month 3/day 1200UTC in 2017, and the 0-12h accumulated precipitation, the 12-24h accumulated precipitation and the 0-24h accumulated precipitation simulated by the optimal scheme (scheme 17) in the 32 schemes are shown in FIG. 15. (a1) - (a3) is live precipitation, and (b1) - (b3) are forecast results corresponding to the optimal parameterization scheme of the example, and the first column to the third column respectively represent 0-12h accumulated precipitation, 12-24h accumulated precipitation and 0-24h accumulated precipitation.

It can be seen that the rainstorm center of live rainfall is in the northeast-southwest direction and is mainly located in Liaoning, Jilin, southeast inner Mongolia and south black dragon river, and the maximum 24-hour cumulative rainfall can reach more than 90 mm. Compared with the actual situation, the main position of the rainstorm center of the simulation result of the scheme 17 is consistent with that of observation, but the simulated rainstorm center is relatively divergent and relatively low in strength, and relatively weak rainstorm centers exist in the middle and north of Heilongjiang.

In the embodiment, 2 precipitation observation data are selected for 3 cases of shear type rainstorm, namely GPM observation data and a product of precipitation fusion of an automatic station and CMORPH hours. The observation data of GPM selected for SH _3 is generally stronger in precipitation than that of automatic stations selected for SH _1 and SH _2 and CMORPH hour precipitation fusion products, so that the SH _3 report missing rate is higher.

Sensitivity test research is respectively carried out on 3 northeast tangent variant rainstorm cases, the tangential variant rainstorm simulation result is found to be ideal, the test is more accurate in simulation of a rainfall precipitation area and a rainstorm center, and the results of simulation and forecast of the tangential variant rainstorm in the northeast region are better in the scheme 14, the scheme 17 and the scheme 18. The parametric settings for the 3 schemes are shown in the table below.

The schemes 17 and 18 adopt the micro-physical, long and short wave, land and boundary layer parameterization schemes in accordance, and the scheme 14 is greatly different from the schemes 17 and 18. By analyzing fig. 10, it can be seen that scenario 18 also responded well to the 20-day 20-7-2010 rainstorm case. The numerical simulation of the switched rainstorm in the northeast region is shown in the specification, a WSM 6-class slope scheme is selected for the micro-physical process parameterization, an RRTM scheme is selected for the long and short wave radiation parameterization, a unified Noah scheme is selected for the land process parameterization, a YSU scheme is selected for the boundary layer parameterization, and the cloud parameterization scheme is selected according to specific conditions, so that an ideal simulation result can be obtained.

Example 2

This example was conducted by conducting a sensitivity test study on a simulation parameterization scheme of a cyclone type rainstorm according to the rainstorm weather simulation forecasting method based on the WRF mode in the present invention, and the study procedure was the same as the method procedure in example 1.

In this example, 2 cases of 8/21 days in 1997 and 16/7/2013 were selected.

Sensitivity test research is carried out on 2 cyclone type rainstorm cases in the northeast area respectively, the simulation of a rainstorm falling area is found to be ideal, and the result of the cyclone type rainstorm simulation in the northeast area is better in the scheme 8. The parametric settings for the 2 schemes are shown in the table below.

This shows that for the numerical simulation of the cyclone type rainstorm in the northeast region, the Thompson grapnel scheme is selected for the micro-physical process parameterization, the RRTM scheme is selected for the long and short wave radiation parameterization, the RUC scheme is selected for the land process parameterization, the Mellor-Yamada-Janjic (eta) TKE scheme is selected for the boundary layer parameterization, and the Tiedtke scheme is selected for the cloud parameterization, so that a better simulation effect can be achieved.

Example 3

In this embodiment, sensitivity test research is performed on a simulation parameterization scheme of the cold vortex type rainstorm according to the rainstorm weather simulation forecasting method based on the WRF mode, and the research process is the same as the method steps in embodiment 1.

In this example, 2 cases of 6/21/2010 and 9/2010 were selected.

Sensitivity test research is carried out on 2 cases of the cold vortex type rainstorm in the northeast region respectively, the fact that the falling region for simulating the cold vortex type rainstorm is ideal is found, and the result of simulating the cold vortex type rainstorm in the northeast region by the scheme 10 and the scheme 11 is better. The parametric settings for the 2 schemes are shown in the table below.

The optimal schemes corresponding to the 2 cases are consistent except that the boundary layer parameterization schemes are different. This shows that for the numerical simulation of the cold vortex storm in the northeast region, the Thompson grapnel scheme is selected for the micro-physical process parameterization, the Goddard scheme is selected for the long and short wave radiation parameterization, the unified Noah scheme is selected for the land process parameterization, the Tiedtke scheme is selected for the cloud parameterization, and the boundary layer parameterization scheme can be selected according to specific conditions, so that a better simulation effect can be obtained.

Example 4

In this embodiment, according to the rainstorm weather simulation forecasting method based on the WRF mode, the sensitivity test study is performed on the typhoon type rainstorm simulation parameterization scheme, and the study process is the same as the method steps in embodiment 1.

In this example, 2 cases of 6/16 th in 1984 and 16/8/1994 were selected.

Sensitivity test research is carried out on 2 typhoon type rainstorm cases in the northeast region respectively, the strength of simulating the typhoon type rainstorm is found to be ideal, and the simulation result of the scheme 15 and the scheme 7 on the typhoon type rainstorm in the northeast region is better. The parametric settings for the 2 schemes are shown in the table below.

The 2 schemes are consistent except that the long-short wave radiation parameterization scheme is selected differently. The numerical simulation of typhoon rainstorms in the northeast region is shown in the specification, a Thompson grapnel scheme is selected for the micro-physical process parameterization, an RUC scheme is selected for the land process parameterization, a Mellor-Yamada-Janjic (eta) TKE scheme is selected for the boundary layer parameterization, a Betts-Miller-Janjic scheme is selected for the cloud parameterization, and the long-short wave radiation parameterization scheme is judged according to specific conditions, so that a better simulation effect can be obtained.

By performing the parameterization scheme sensitivity tests on 9 cases corresponding to the 4 types of northeast rainstorms, namely the northeast tangential rainstorm, the cyclone rainstorm, the cold vortex rainstorm and the typhoon rainstorm, it can be found that the optimal parameterization schemes corresponding to the rainstorms have large differences. In the service operation process, the specific type of rainstorm is not known, and a general parameterization scheme needs to be provided for rainstorm simulation in the northeast region in order to select the parameterization scheme as soon as possible for numerical simulation.

Fig. 16 shows the rainfall threshold values with a TS score of 50mm rainstorm cases, and the abscissa shows the schemes, (a) rainstorm cases of 8/1 th in 2008, (b) rainstorm cases of 7/20 th in 2010, (c) rainstorm cases of 8/3 th in 2017, (d) rainstorm cases of 8/21 th in 1997, (e) rainstorm cases of 7/16 th in 2013, (f) rainstorm cases of 6/21 th in 1981, (g) rainstorm cases of 8/9 th in 2010, (h) rainstorm cases of 6/16 th in 1984, and (i) rainstorm cases of 8/16 th in 1994. The horizontal line in the middle represents the forecaster mean forecast score, the horizontal line below represents the T639 mode mean forecast score, and the horizontal line above represents the EC mode mean forecast score. The box represents the selection of the optimal scheme for simulating the rainstorm numerical value in the northeast region.

When protocol 7 was selected for each case, the TS scores were at a higher level compared to the other protocols in each case, except for the rainstorm cases at 7/16/2013 and at 21/6/1981. The overall TS score of the 7-month 16-day rainstorm cases in 2013 is far lower than the TS score of other cases. The 21 st rainstorm cases occurred at the earliest time in 1981, 6.1, and the data quality was poor for simulation and evaluation. The representative of the rainstorm cases of 7 and 16 months in 2013 and the rainstorm cases of 21 months 6 and 1981 are poor.

In view of the above considerations, the recommended scheme 7 is a general parameterized scheme for rainstorm numerical simulation in the northeast region. When the rainstorm in the northeast region needs to be numerically simulated and the rainstorm type in the northeast region is unknown, the Thompson grapnel scheme is selected for the micro-physical process parameterization, the RRTM scheme is selected for the long and short wave radiation parameterization, the RUC scheme is selected for the land process, the Mellor-Yamada-Janjic (eta) TKE scheme is selected for the boundary layer parameterization, and the Betts-Miller-Janjic scheme is selected for the cloud accumulation parameterization, so that a proper simulation effect can be obtained.

By adopting the rainstorm weather simulation forecasting method based on the WRF mode, the rainstorm in the northeast region can be finely forecasted, a unified rainstorm weather simulation parameterization scheme suitable for the northeast region can be obtained after simulation analysis research is carried out on different rainstorm types, and the forecasting accuracy is effectively improved.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A rainstorm weather simulation forecasting method based on a WRF mode is characterized by comprising the following steps:

2. The WRF mode-based stormwater weather simulation forecasting method as claimed in claim 1, wherein in the step (1), the historical stormwater weather includes 199 regional stormwater and large-scale stormwater cases in the northeast region of 1981 and 2017.

3. The WRF mode-based rainstorm weather simulation forecasting method according to claim 2, wherein in the step (1), the rainstorm types obtained through weather classification include: shear type rainstorms, cyclone type rainstorms, cold vortex type rainstorms, and typhoon type rainstorms.

4. The WRF mode-based stormwater weather simulation forecasting method as claimed in claim 1, wherein in the step (2), a plurality of cases are selected for each type of stormwater by analyzing weather conditions in a 500hPa weather map.

5. The WRF mode-based rainstorm weather simulation forecasting method according to claim 4, wherein, in the step (2), the global forecast data driving the WRF mode uses forecast data issued by the united states atmospheric environment forecasting center.

6. The method for simulating and forecasting stormy weather based on WRF mode as claimed in claim 5, wherein in the step (2), the WRF mode adopts a double nesting scheme, the center positions of two nested areas are both located at (46.7 degrees N and 125.0 degrees E), the outer areas are 110-140 degrees E and 30-60 degrees N, and the resolution is 27 km; the internal region is: 115-135E, 34-54N, the resolution is 9km, the vertical direction is divided into 28 layers, and the top of the mode layer is 50 hPa.

7. The WRF mode-based rainstorm weather simulation forecasting method according to claim 6, wherein in the step (2), the parameterization schemes comprise a micro-physics scheme, a long and short wave radiation scheme, a land surface process, a boundary layer scheme and a cumulus parameterization scheme, and two schemes are respectively selected from each sensitivity test to form 32 scheme combinations.

8. The WRF mode-based rainstorm weather simulation forecasting method according to claim 1, wherein in the step (3), the existing numerical forecasting products include T639 global ensemble forecasting system and european mid-term weather forecasting center mode.

9. The WRF model-based rainstorm weather simulation forecasting method according to claim 8, wherein in the step (3), the inspection data includes hourly 0.1 ° x 0.1 ° precipitation data, 0.1 ° gpm precipitation data and site precipitation data provided by a chinese weather science data sharing service network, in which an automatic station in china is fused with a CMORPH precipitation product.

10. The WRF mode-based rainstorm weather simulation forecasting method according to claim 9, wherein in the step (3), the WRF forecast rainfall field is interpolated on a grid which is matched with live data for scoring test.