CN110968926B - Method for predicting atmospheric parameters based on improved background error covariance matrix - Google Patents

Method for predicting atmospheric parameters based on improved background error covariance matrix Download PDF

Info

Publication number
CN110968926B
CN110968926B CN201811147232.4A CN201811147232A CN110968926B CN 110968926 B CN110968926 B CN 110968926B CN 201811147232 A CN201811147232 A CN 201811147232A CN 110968926 B CN110968926 B CN 110968926B
Authority
CN
China
Prior art keywords
atmospheric
parameters
covariance matrix
forecast
background error
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201811147232.4A
Other languages
Chinese (zh)
Other versions
CN110968926A (en
Inventor
何杰颖
张升伟
董晓龙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
National Space Science Center of CAS
Original Assignee
National Space Science Center of CAS
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by National Space Science Center of CAS filed Critical National Space Science Center of CAS
Priority to CN201811147232.4A priority Critical patent/CN110968926B/en
Publication of CN110968926A publication Critical patent/CN110968926A/en
Application granted granted Critical
Publication of CN110968926B publication Critical patent/CN110968926B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Operations Research (AREA)
  • Probability & Statistics with Applications (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Algebra (AREA)
  • Evolutionary Biology (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention provides a method for predicting atmospheric parameters based on an improved background error covariance matrix, which comprises the following steps: step 1) inputting satellite observation data into a WRF model; outputting atmospheric parameter variables for weather forecast at the appointed time and at the appointed spatial resolution grid point; step 2) calculating the variance of the variables of the same kind of parameters in the atmospheric parameters and the covariance among different parameters; step 3) drawing a relation curve of the atmospheric parameter variable and the forecast duration, and obtaining a time period with linear correlation between the atmospheric parameter variable and the forecast duration from the relation curve; step 4) constructing a background error covariance matrix, wherein main diagonal elements are the variation of variances of the same parameter variables in the atmosphere parameters along with the forecast time, and non-diagonal elements are the variation of covariance among different parameter variables in the atmosphere along with the forecast time; and 5) inputting satellite observation data and a background error covariance matrix into a WRF (Wireless sensor Filter) equalization model, so as to obtain a predicted value of the atmospheric parameter in the prediction duration.

Description

Method for predicting atmospheric parameters based on improved background error covariance matrix
Technical Field
The invention relates to the field of atmospheric prediction and data assimilation systems, in particular to a method for predicting atmospheric parameters based on an improved background error covariance matrix.
Background
The prior information in the data assimilation system generally originates from prediction, the background error covariance is related to the prediction value and the true value, the overall assimilation effect is affected in the assimilation system, and the prior information is the expected value of the difference between the two values, namely the product of the difference between the two values and the transposed difference between the two values. A more accurate background error covariance matrix is beneficial to improving the data assimilation effect, because the assimilation depends on priori knowledge, and uncertainty of the priori knowledge is generally expressed by a probability density function, is generally simplified into a Gaussian form, and errors can be transferred to a real-time observation information domain through a control variable or the variable itself to influence the balance of iterative algorithms in an assimilation system. The static atmosphere variable is embedded in the background error covariance matrix, so that the assimilation process is greatly influenced, and the static matrix with larger error can amplify noise in an unmeasurable variable mode, so that the assimilation stability and the assimilation applicability are influenced.
The methods currently in common use are based on 24-hour or 12-hour analysis data and are not applicable in extreme weather assimilation systems that change rapidly in short time. The state updating analysis method needs dense actual measurement data, and the aggregation method needs independent aggregate Kalman matrix or B matrix to disturb the model. In the WRFDA mode developed by NCAR, control variables with different expression forms can be generated by using an open-source executable file, so that the operation of the WRFDA is satisfied, but the mode does not deeply consider the hydrogel particles so far, and is not suitable for extreme weather conditions of high-speed space-time variation.
Disclosure of Invention
The invention aims to solve the problems that the background error covariance matrix commonly used at present does not deeply predict the parameters of hydraulic condensate particles in the atmosphere, is not suitable for an extreme weather assimilation system which rapidly changes in short time and cannot predict extreme weather conditions of high-speed space-time change.
In order to achieve the above object, the present invention provides a method for improving background error covariance matrix to predict atmospheric parameters, comprising:
step 1) inputting satellite observation data into a WRF model; outputting atmospheric parameter variables for weather forecast at the appointed time and at the appointed spatial resolution grid point;
step 2) calculating the variance of the variables of the same kind of parameters in the atmospheric parameters and the covariance among different parameters;
step 3) drawing a relation curve of the atmospheric parameter variable and the forecast duration, and obtaining a time period with linear correlation between the atmospheric parameter variable and the forecast duration from the relation curve;
step 4) constructing a background error covariance matrix, wherein main diagonal elements are the variation of variances of the same parameter variables in the atmosphere parameters along with the forecast time, and non-diagonal elements are the variation of covariance among different parameter variables in the atmosphere along with the forecast time;
and 5) inputting satellite observation data and a background error covariance matrix into a WRF (Wireless sensor Filter) equalization model, so as to obtain a predicted value of the atmospheric parameter in the prediction duration.
As an improvement of the above method, the step 1) includes:
step 1-1), taking 0,6, 12 and 18 hours of analysis data as background field data, inputting the background field data into a WRF model, and selecting a regional grid point as 15 km;
step 1-2) outputting atmospheric parameters at a specified time and at a specified spatial resolution grid point, wherein the atmospheric parameters comprise thermodynamic parameters and hydraulic condensate parameters; the thermodynamic parameters include temperature and humidity, the hydraulic parameters include cloud, rain, snow, ice, and aragonite;
step 1-3) setting atmospheric parameters of the forecast sample
Figure SMS_1
i=1..7, superscript-f is denoted as forecast samples, each +.>
Figure SMS_2
The values of 7 parameters including temperature, humidity and the water condensation density of cloud, rain, snow, ice and aragonite particles, wherein the water condensation density ρ of cloud, rain, snow, ice and aragonite particles h Values were obtained by the following steps:
step 1-3-1) setting the mixing ratio of the condensate parameters to be Q h ，Q h Respectively adopt Q cl 、Q r 、Q s 、Q i 、Q g Is a numerical value of (2); wherein Q is cl Represents the mixing ratio of cloud water, Q r Represents the mixing ratio of cloud and rain, Q s Represents the snow mixing ratio, Q i Represents the ice mixing ratio, Q g Representing the aragonite mixing ratio;
step 1-3-2) calculating the different hydrogel densities ρ of the cloud, rain, snow, ice and aragonite particles, respectively h ，ρ h Respectively adopt ρ cl ，ρ r ，ρ s ，ρ i ，ρ g To represent the values of cloud density, rain density, snow density, ice density and aragonite density, in g/m, respectively 3 ；
Figure SMS_3
Wherein Q is v For the water-vapor mixing ratio,ρ v is the density of water vapor.
As an improvement of the above method, said step 2) comprises:
step 2-1) calculating background errors of the k atmosphere parameters
Figure SMS_4
Figure SMS_5
Figure SMS_6
A is the background error of the ith parameter variable in the atmospheric parameters at the moment t, and a is expressed as an analysis sample; Δt is expressed as time increment, +.>
Figure SMS_7
An analysis sample representing the ith atmospheric parameter at the initial moment, a->
Figure SMS_8
Is a forecast sample of the ith atmospheric parameter after the deltat moment, i is more than or equal to 1 and less than or equal to k;
step 2-2) calculating the variance of the background error of the atmospheric parameters
Figure SMS_9
Figure SMS_10
Step 2-3) calculating covariance of background errors of different atmospheric parameters
Figure SMS_11
Figure SMS_12
Wherein, j is more than or equal to 1 and less than or equal to k, i is not equal to j.
As an improvement of the above method, the atmospheric parameter variable versus the forecast duration of step 3) includes a thermodynamic parameter variable versus the forecast duration and a hydrogel parameter variable versus the forecast duration.
As an improvement of the above method, the step 2) variance and covariance are expressed as a function σx of Δt i 2 (Δt) and σx i x j (Δt)：
Figure SMS_13
/>
Figure SMS_14
Where deltat is expressed as the length of the forecast,
Figure SMS_15
is offset.
As a modification of the above method, the predicted time period of the thermodynamic parameter variable in the step 3) is 9 hours, and the predicted time period of the hydraulic parameter variable is 4.5 hours.
As an improvement of the above method, the variation of the atmospheric parameter over the forecast duration follows a brownian motion in time.
As an improvement of the above method, the background error covariance matrix in the step 4) dynamically changes along with the forecast duration, and the update speed is consistent with the update speed of the satellite transmission observation data.
As an improvement of the above method, the step 5) specifically includes:
step 5-1) inputting satellite observation data and a background error matrix into a WRF assimilation model to obtain a cost function j (x):
Figure SMS_16
wherein x is an atmospheric parameter variable, x b Is the atmospheric parameter of the known background field, y is satellite observation data, H is observationOperator, R is the observation error matrix, T is the transposition, B is the background error covariance matrix;
step 5-2) deriving the cost function j (x), and when j' (x) =0, converging the background error covariance matrix to obtain a predicted value of the atmospheric parameter.
The invention has the advantages that:
the method for predicting the atmospheric parameters by using the improved background error covariance matrix provided by the invention utilizes the improved background error covariance matrix to enable all the atmospheric state parameters to converge in the prediction time, solves the current situation that the current general background error covariance matrix predicts the hydraulic parameters in a divergent state in a data assimilation mode, and has good effect in an atmospheric data assimilation system through verification.
Drawings
FIG. 1 is a schematic diagram of a WRF mode input/output of the prior art;
FIG. 2 is a graph of NCEP6 hour analysis data provided by the national atmospheric research center data research division;
FIG. 3 is a schematic diagram of a background error covariance matrix selection region of the present invention;
FIG. 4 (a) is a diagram illustrating the background error covariance matrix data source according to the invention;
FIG. 4 (b) is a schematic diagram of background error covariance matrix data calculation according to the invention;
FIG. 5 (a) is a graph showing correlation between the condensate parameters and the forecast time according to the present invention;
FIG. 5 (b) is a graph showing the correlation between thermodynamic parameters and forecast time according to the present invention;
FIG. 6 is a flow chart of a method of predicting atmospheric parameters based on an improved background error covariance matrix in accordance with the invention.
Detailed Description
The invention will now be described in detail with reference to the drawings and specific examples.
The invention provides a method for predicting atmospheric parameters by improving a background error covariance matrix, which aims at an observation data assimilation mode with high space-time resolution in a region, wherein the assimilation mode is input as microsatellite observation data, the time resolution is 10-15 minutes, the spatial resolution is 10km, and the increment between an observation bright temperature value and an advanced radiation transmission mode simulation bright temperature value is calculated at any time by utilizing Kalman filtering.
The background error covariance matrix is a matrix which dynamically changes along with the forecast duration, is different from the static concept in a time window in the existing assimilation mode, and has the update speed consistent with the input satellite observation data, so that the assimilation system better realizes the compromise optimization of the background error and the observation error, and the data assimilation effect is improved.
The invention provides a method for predicting atmospheric parameters by improving a background error covariance matrix, which specifically comprises the variances and covariances of atmospheric temperature, humidity, clouds, rain, snow, ice and aragonite particles.
With the re-analysis grid point data of one of the data sets FNL provided by the national environmental prediction center (NCEP) provided by the national institute of atmospheric research, as input, every 6 hours (1 degree by 1 degree or 2.5 degrees by 2.5 degrees), 32 layers were set up from 1000hpa (mb) to 10hpa (mb) from the ground surface up to the high altitude. Parameters include surface air pressure, sea surface air pressure, potential height, temperature, sea surface temperature, soil condition, ice coverage, relative humidity, u, v wind speed, vertical motion, vorticity, ozone and other assimilated parameters, and high space-time resolution atmospheric parameters are output through a WRF model. The forecasting time interval is 15 minutes, and the grid point of the selected area is 15 kilometers; different area ranges and grid point sizes can be set according to the needs; by calculating the parameter characteristics of the forecast data and the analysis data of the same time difference, the change response of the atmospheric thermodynamic and hydraulic condensate parameters along with time under the extreme weather condition is found, and the change of each parameter in different time scale ranges is proved to follow the Brownian motion in time, see formulas (1) - (5), and then the background error covariance matrix is calculated according to the relation.
First, calculate the background error of k atmospheric parameters
Figure SMS_17
Figure SMS_17
1≤i≤k：
Figure SMS_18
Figure SMS_19
The background error of the ith parameter variable in the atmospheric parameters at the moment t is represented by a superscript f as a forecast sample and a as an analysis sample; Δt is expressed as time increment, +.>
Figure SMS_20
An analysis sample representing the ith atmospheric parameter at the initial time,
Figure SMS_21
is a forecast sample of the ith atmospheric parameter after the delta t moment;
second, calculate the variance of the background error of the atmospheric parameters
Figure SMS_22
Figure SMS_22
1≤i≤k：
Figure SMS_23
Again, the covariance of the background errors of the different atmospheric parameters is calculated
Figure SMS_24
1≤i≤k,1≤j≤k,i≠j：
Figure SMS_25
According to the above formula, the variance and covariance are expressed as a function σx of Δt i 2 (Δt) and σx i x j (Δt)：
Figure SMS_26
Figure SMS_27
Where deltat is expressed as a forecast duration, which is an integer multiple of deltat,
Figure SMS_28
is biased;
Δt＝15mins (6)
the linear correlation of the atmospheric parameter variable with the forecast duration is obtained, and the background error variance and covariance decrease with the shortening of the time increment deltat.
As shown in fig. 1, using global 6-hour 1 o The grid analysis data FNL is used as an initial field driving weather forecast mode WRF to obtain weather forecast at a specified moment and at a specified spatial resolution grid point, cloud and rain atmospheric environment data such as temperature, humidity, air pressure, water vapor content and the like output by forecast are extracted to be used as data sources of a background error covariance matrix, and the data can be directly extracted according to longitude and latitude and height of a storage area.
As shown in table 1:
TABLE 1
Figure SMS_29
As shown in fig. 3, the selected area is an atlantic area frequent by hurricanes, domain is 25-35 degrees N, -70-60 degrees W, the time span is 2015-2017, 10 typical typhoons are covered, and the data is a FNL atmospheric parameter data set issued by global initial analysis data NCEP.
As shown in fig. 4 (a), the WRF mode was used to predict the atmospheric parameters for the next 9 hours in combination with the background field data of 0,6, 12, and 18 hours, and the input background field data was input at 15 minute intervals. And extracting the atmospheric parameters needed in the background error covariance matrix for output data, and performing correlation calculation, wherein:
the hydrogel parameters include density parameters of the cloud, rain, snow, ice, and aragonite particles;
mixing ratio of hydrogel parameters (Q h ) Respectively using Q cl 、Q r 、Q s 、Q i 、Q g A representation; wherein Q is c Represents the mixing ratio of cloud water, Q r Represents the mixing ratio of cloud and rain, Q s Represents the snow mixing ratio, Q i Represents the ice mixing ratio, Q g Indicating the mixing ratio of the shot.
The specific transformation extraction formula is shown in the specification of FIG. 4 (b) and formulas (7), (8) and (9), the density ρ of the condensate h The unit is g/m 3 ，ρ h Respectively adopt ρ cl ，ρ r ，ρ s ，ρ i ，ρ g Is a numerical value of (2);
Figure SMS_30
/>
Figure SMS_31
Figure SMS_32
here, M h Any amount of hydraulic condensate in unit volume V;
M d dry air mass per unit volume V;
M v the mass of water vapor per unit volume V.
As shown in fig. 4 (a), the initial field is selected for 3 days, 4 times a day, a forecast time interval is 15 minutes, a total forecast duration is 9 hours, 36 forecast times are provided, each forecast time has 12 forecast samples, and the forecast samples serve as a basic data source for verifying the brown operation rule.
As shown in fig. 5, hydrogel parameters such as Cloud (Cloud), rain (Rain), snow (Snow), ice (Ice), and aragonite (granpel), and thermodynamic parameters such as atmospheric Temperature (Temperature) and Vapor (Vapor), all have different degrees of linearity with the forecast duration. The abscissa is the forecast time, t=1 represents the first forecast time, t=2 represents the second forecast time, and so on, and the ordinate is the variance. It can be seen that the correlation between the hydrogel parameters and the forecast time is stronger, the quasi-linear relationship gradually weakens and disappears along with the increase of time, the background field and the observation field of the whole system are integrated, the forecast time of the hydrogel parameters is set to be 4.5 hours, and the thermodynamic parameters can meet the quasi-linear relationship within 9 hours.
The improved background error covariance matrix provided by the invention utilizes a WRF mode, performs Brownian motion rule analysis on atmospheric parameters based on forecast field data and forecast time of different starting times at 15 minutes as time intervals, and calculates parameter variances and covariance in the background error matrix, namely each element of the matrix by using the rule, thereby obtaining a background error covariance matrix dynamically changing along with satellite observation data input, wherein the background error covariance matrix is a matrix, a symmetric matrix and a positive matrix, the diagonal is not 0, and the dimension is equal to the number of variables.
As shown in table 2, the main diagonal element is the variation relationship of the atmospheric parameter variance with the forecast duration, that is, the variance shown in the table, and the off-diagonal element is the variation relationship of the covariance of the related element with the forecast duration, that is, the covariance shown by the diagonal, wherein the variance and the covariance are both functions of time and satisfy the brown theory.
And analyzing the covariance of each atmosphere parameter of the same forecast duration and the variation of the covariance along with the forecast duration according to different starting times and forecast moments, so as to analyze the relation between the background error covariance matrix and the forecast duration.
The background error covariance matrix is applied to a satellite microwave assimilation system, a group of background error covariance matrices with smaller actual atmospheric state errors are provided every 15 minutes, and under the condition that the observation errors are unchanged, the prediction errors are closer to the true values, so that atmospheric parameters converged in the prediction duration are obtained, namely, the atmospheric parameters are more consistent with atmospheric thermodynamic and hydraulic parameters.
TABLE 2
Figure SMS_33
/>
As shown in fig. 6, a method flow of predicting an atmospheric parameter based on an improved background error covariance matrix according to the invention is provided, comprising:
step 1) inputting satellite observation data into a WRF model; outputting atmospheric parameter weather forecast at a specified time and at a specified spatial resolution grid point;
step 2) extracting atmospheric parameters output by weather forecast, and calculating to obtain variable variance and covariance of the condensate parameters;
step 3) drawing a relation curve of atmospheric thermodynamics and the parameter variable of the condensate and the forecast duration by taking the designated moment as an abscissa and the variance of the parameter variable of the condensate as an ordinate;
step 4) obtaining from the relation curve that the atmospheric condensate parameter variance and covariance are functions of time in a certain time period, and the variation complies with the Brownian motion in time in different time scale ranges; thus obtaining a certain time period with linear correlation between the parameter variables of the atmospheric thermodynamics and the hydraulic condensate and the forecast duration;
step 5) taking the atmospheric condensate parameter variances at different moments within a certain time period with linear correlation as main diagonal elements, and constructing a background error covariance matrix by taking the covariance of the correlation elements as off-diagonal elements;
and 6) applying the background error covariance matrix in the satellite microwave assimilation system, providing a group of background error covariance matrices with smaller actual atmospheric state errors every 15 minutes, and under the condition that the observation errors are unchanged, enabling the difference between the forecast value and the true value to be the local minimum, so as to obtain the atmospheric parameters converged in the forecast duration, and enabling the cost function to be the minimum, namely more consistent with the atmospheric thermodynamic and hydraulic parameters.
The cost function of step 6) can be expressed as:
Figure SMS_34
the right side of the equation consists of two parts, and the background error and the observation error jointly determine a cost function; in the actual atmospheric parameter forecasting process, carrying out a derivation operation on the background error matrix and observation data according to a cost function formula;
where x is an atmospheric state variable, x b Is the atmospheric state parameter of the known background field, y is satellite observation data, H is an observation operator, R -1 Is an observation error matrix, and B is an error covariance matrix. And deriving a cost function formula j (x), and converging the background error covariance matrix when j' (x) =0 to obtain a predicted value of the atmospheric state parameter.
Finally, it should be noted that the above embodiments are only for illustrating the technical solution of the present invention and are not limiting. Although the present invention has been described in detail with reference to the embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made thereto without departing from the spirit and scope of the present invention, which is intended to be covered by the appended claims.

Claims (9)

1. A method of predicting an atmospheric parameter based on an improved background error covariance matrix, comprising:
step 1) inputting satellite observation data into a WRF model; outputting atmospheric parameter variables for weather forecast at the appointed time and at the appointed spatial resolution grid point;
step 2) calculating the variance of the variables of the same kind of parameters in the atmospheric parameters and the covariance among different parameters;
step 3) drawing a relation curve of the atmospheric parameter variable and the forecast duration, and obtaining a time period with linear correlation between the atmospheric parameter variable and the forecast duration from the relation curve;
step 4) constructing a background error covariance matrix, wherein main diagonal elements are the variation of variances of the same parameter variables in the atmosphere parameters along with the forecast time, and non-diagonal elements are the variation of covariance among different parameter variables in the atmosphere along with the forecast time;
and 5) inputting satellite observation data and a background error covariance matrix into a WRF (Wireless sensor Filter) equalization model, so as to obtain a predicted value of the atmospheric parameter in the prediction duration.
2. The method for predicting atmospheric parameters based on an improved background error covariance matrix according to claim 1, wherein said step 1) comprises:
step 1-1), taking 0,6, 12 and 18 hours of analysis data as background field data, inputting the background field data into a WRF model, and selecting a regional grid point as 15 km;
step 1-2) outputting atmospheric parameters at a specified time and at a specified spatial resolution grid point, wherein the atmospheric parameters comprise thermodynamic parameters and hydraulic condensate parameters; the thermodynamic parameters include temperature and humidity, the hydraulic parameters include cloud, rain, snow, ice, and aragonite;
step 1-3) setting atmospheric parameters of the forecast sample
Figure FDA0004110384020000011
Superscript-f is denoted as forecast samples, each +.>
Figure FDA0004110384020000012
The values of 7 parameters including temperature, humidity and the water condensation density of cloud, rain, snow, ice and aragonite particles, wherein the water condensation density ρ of cloud, rain, snow, ice and aragonite particles h Values were obtained by the following steps:
step 1-3-1) setting the mixing ratio of the condensate parameters to be Q h ，Q h Respectively adopt Q cl 、Q r 、Q s 、Q i 、Q g Is a numerical value of (2); wherein Q is cl Represents the mixing ratio of cloud water, Q r Represents the mixing ratio of cloud and rain, Q s Represents the snow mixing ratio, Q i Represents the ice mixing ratio, Q g Representing the aragonite mixing ratio;
step 1-3-2) calculating the different hydrogel densities ρ of the cloud, rain, snow, ice and aragonite particles, respectively h ，ρ h Respectively adopt ρ cl ，ρ r ，ρ s ，ρ i ，ρ g To represent the values of cloud density, rain density, snow density, ice density and aragonite density, in g/m, respectively 3 ；
Figure FDA0004110384020000013
Wherein Q is v For the water-steam mixing ratio ρ v Is the density of water vapor.
3. The method for predicting atmospheric parameters based on an improved background error covariance matrix according to claim 1, wherein said step 2) comprises:
step 2-1) calculating background errors of the k atmosphere parameters
Figure FDA0004110384020000021
Figure FDA0004110384020000022
Figure FDA0004110384020000023
A is the background error of the ith parameter variable in the atmospheric parameters at the moment t, and a is expressed as an analysis sample; Δt is expressed as time increment, +.>
Figure FDA0004110384020000024
An analysis sample representing the ith atmospheric parameter at the initial moment, a->
Figure FDA0004110384020000025
Is a forecast sample of the ith atmospheric parameter after the deltat moment, i is more than or equal to 1 and less than or equal to k; />
Step 2-2) calculating the variance of the background error of the atmospheric parameters
Figure FDA0004110384020000026
Figure FDA0004110384020000027
Step 2-3) Calculating covariance of background errors of different atmospheric parameters
Figure FDA0004110384020000028
Figure FDA0004110384020000029
Wherein, j is more than or equal to 1 and less than or equal to k, i is not equal to j.
4. The method for predicting atmospheric parameters based on the improved background error covariance matrix according to claim 3, wherein the atmospheric parameter variable versus the predicted time period in step 3) comprises a thermodynamic parameter variable versus the predicted time period and a hydraulic parameter variable versus the predicted time period.
5. The method of predicting atmospheric parameters based on an improved background error covariance matrix as recited in claim 4, wherein said step 2) variance and covariance are expressed as a function σx of Δt i 2 (Δt) and σx i x j (Δt)：
Figure FDA00041103840200000210
Figure FDA00041103840200000211
Where deltat is expressed as the length of the forecast,
Figure FDA00041103840200000212
is offset.
6. The method for predicting atmospheric parameters based on an improved background error covariance matrix according to claim 4, wherein the predicted time period of the thermodynamic parameter variable of step 3) is 9 hours, and the predicted time period of the hydraulic parameter variable is 4.5 hours.
7. The method of predicting an atmospheric parameter based on an improved background error covariance matrix as recited in claim 6, wherein the change in the atmospheric parameter over a prediction period follows brownian motion in time.
8. The method for predicting atmospheric parameters based on the improved background error covariance matrix according to claim 1, wherein the background error covariance matrix of step 4) dynamically changes with the prediction duration, and the update speed is consistent with the update speed of the satellite transmission observation data.
9. The method for predicting atmospheric parameters based on the improved background error covariance matrix according to any one of claims 1-8, wherein said step 5) comprises:
step 5-1) inputting satellite observation data and a background error matrix into a WRF assimilation model to obtain a cost function j (x):
Figure FDA0004110384020000031
wherein x is an atmospheric parameter variable, x b Is the atmospheric parameter of the known background field, y is satellite observation data, H is an observation operator, R is an observation error matrix, T is a transposition, and B is a background error covariance matrix;
step 5-2) deriving the cost function j (x), and when j' (x) =0, converging the background error covariance matrix to obtain a predicted value of the atmospheric parameter.
CN201811147232.4A 2018-09-29 2018-09-29 Method for predicting atmospheric parameters based on improved background error covariance matrix Active CN110968926B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811147232.4A CN110968926B (en) 2018-09-29 2018-09-29 Method for predicting atmospheric parameters based on improved background error covariance matrix

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811147232.4A CN110968926B (en) 2018-09-29 2018-09-29 Method for predicting atmospheric parameters based on improved background error covariance matrix

Publications (2)

Publication Number Publication Date
CN110968926A CN110968926A (en) 2020-04-07
CN110968926B true CN110968926B (en) 2023-05-02

Family

ID=70027269

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811147232.4A Active CN110968926B (en) 2018-09-29 2018-09-29 Method for predicting atmospheric parameters based on improved background error covariance matrix

Country Status (1)

Country Link
CN (1) CN110968926B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111881590A (en) * 2020-07-30 2020-11-03 中国科学院空天信息创新研究院 Spatial analysis method for concentration of atmospheric particulate matter
CN114048433B (en) * 2021-10-26 2022-06-21 南京大学 Mixed assimilation system and method based on ensemble Kalman filtering framework
CN116975523B (en) * 2023-09-22 2023-12-12 南京气象科技创新研究院 Data assimilation background error covariance characteristic statistical method for strong convection weather typing

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4880440B2 (en) * 2006-12-22 2012-02-22 財団法人電力中央研究所 Snow accretion prediction method and snow accretion prediction program
CN106339568B (en) * 2015-07-08 2019-04-05 中国电力科学研究院 A kind of numerical weather forecast method based on mixing ambient field
CN107273995A (en) * 2016-04-08 2017-10-20 株式会社日立制作所 Urban Air Pollution Methods
CN107991722A (en) * 2017-12-25 2018-05-04 北京墨迹风云科技股份有限公司 Method for building up, Forecasting Methodology and the prediction meanss of weather prediction model

Also Published As

Publication number Publication date
CN110968926A (en) 2020-04-07

Similar Documents

Publication Publication Date Title
Buehner et al. Intercomparison of variational data assimilation and the ensemble Kalman filter for global deterministic NWP. Part II: One-month experiments with real observations
Villafuerte et al. Significant influences of global mean temperature and ENSO on extreme rainfall in Southeast Asia
Kumar et al. Simulations over South Asia using the Weather Research and Forecasting model with Chemistry (WRF-Chem): set-up and meteorological evaluation
James et al. Observation system experiments with the hourly updating Rapid Refresh model using GSI hybrid ensemble–variational data assimilation
Ota et al. Ensemble-based observation impact estimates using the NCEP GFS
Caron et al. Implementation of deterministic weather forecasting systems based on ensemble–variational data assimilation at Environment Canada. Part II: The regional system
Cheruy et al. Combined influence of atmospheric physics and soil hydrology on the simulated meteorology at the SIRTA atmospheric observatory
Kumar et al. Assimilation of INSAT‐3D hydro‐estimator method retrieved rainfall for short‐range weather prediction
Wong et al. Trend of daily rainfall and temperature in Peninsular Malaysia based on gridded data set
Necker et al. A convective‐scale 1,000‐member ensemble simulation and potential applications
Ingleby et al. Improved variational analyses using a nonlinear humidity control variable
Brugnara et al. High-resolution analysis of daily precipitation trends in the central Alps over the last century.
CN110968926B (en) Method for predicting atmospheric parameters based on improved background error covariance matrix
Auger et al. AROME–NWC: A new nowcasting tool based on an operational mesoscale forecasting system
Wang et al. A parameterized method for air-quality diagnosis and its applications
Jones et al. Assimilation of satellite infrared radiances and Doppler radar observations during a cool season observing system simulation experiment
Schiro et al. Tropical continental downdraft characteristics: mesoscale systems versus unorganized convection
Afanasev et al. Analysis of interpolation methods to map the long-term annual precipitation spatial variability for the Republic of Bashkortostan, Russian Federation
Yang et al. Evaluating the impact of the COSMIC RO bending angle data on predicting the heavy precipitation episode on 16 June 2008 during SoWMEX-IOP8
Sobash et al. Next-day prediction of tornadoes using convection-allowing models with 1-km horizontal grid spacing
Prasad et al. Improvements in medium range weather forecasting system of India
Zhu et al. A regional GSI-based ensemble Kalman filter data assimilation system for the rapid refresh configuration: Testing at reduced resolution
Squitieri et al. On the forecast sensitivity of MCS cold pools and related features to horizontal grid spacing in convection-allowing WRF simulations
Kim et al. Effect of assimilating Himawari-8 atmospheric motion vectors on forecast errors over East Asia
Gastaldo et al. Assimilation of radar reflectivity volumes in a pre‐operational framework

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant