CN114037125A - Numerical weather forecasting method based on all-weather observation error covariance matrix - Google Patents

Numerical weather forecasting method based on all-weather observation error covariance matrix Download PDF

Info

Publication number
CN114037125A
CN114037125A CN202111209618.5A CN202111209618A CN114037125A CN 114037125 A CN114037125 A CN 114037125A CN 202111209618 A CN202111209618 A CN 202111209618A CN 114037125 A CN114037125 A CN 114037125A
Authority
CN
China
Prior art keywords
observation
matrix
cloud
observation error
error covariance
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.)
Pending
Application number
CN202111209618.5A
Other languages
Chinese (zh)
Inventor
余意
张卫民
曹小群
冷洪泽
银福康
杨锦辉
姚强
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 University of Defense Technology
Original Assignee
National University of Defense Technology
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 University of Defense Technology filed Critical National University of Defense Technology
Priority to CN202111209618.5A priority Critical patent/CN114037125A/en
Publication of CN114037125A publication Critical patent/CN114037125A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • 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
    • G06F17/15—Correlation function computation including computation of convolution operations
    • 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
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Computational Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Computing Systems (AREA)
  • Human Resources & Organizations (AREA)
  • Software Systems (AREA)
  • Databases & Information Systems (AREA)
  • Algebra (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Game Theory and Decision Science (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Development Economics (AREA)
  • Geometry (AREA)
  • Evolutionary Computation (AREA)
  • Operations Research (AREA)
  • Computer Hardware Design (AREA)
  • Marketing (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a numerical weather forecasting method based on an all-weather observation error covariance matrix, which comprises the following steps: assimilating the infrared hyperspectral observation data, constructing an observation error covariance matrix, and performing decomposition and inversion calculation to obtain an observation error correlation matrix; realizing an observation error correlation matrix by using a cost function; calculating deviation of the simulated cloud radiance and the simulated observation radiance and the clear air radiance respectively to obtain the cloud radiance of a background field and the cloud radiance and the cloud characteristic function contained in observation; constructing expansion factors of different channels, and forming a diagonalized expansion matrix by the expansion factors; and (3) bringing the cloud characteristic function and the cloud quantity function into the cost function, calculating the newly constructed cost function and the gradient thereof in the assimilation system, solving the optimal analysis field by minimizing the solution, and performing numerical weather forecast according to the optimal analysis field. The invention can effectively utilize observation error correlation to assimilate infrared hyperspectral data aiming at all-weather.

Description

Numerical weather forecasting method based on all-weather observation error covariance matrix
Technical Field
The invention belongs to the technical field of numerical weather forecast, and particularly relates to a numerical weather forecast method based on an all-weather observation error covariance matrix.
Background
Numerical Weather Prediction (NWP) is a typical partial differential equation initial boundary problem. With the gradual perfection of the numerical model forecasting mode, the accuracy of the initial value condition is more and more considered as an important aspect in the field of numerical weather forecasting, and the accuracy directly influences the success or failure of numerical forecasting. Meanwhile, with the development of observation technology, the amount and types of observation information are increasing, and how to effectively utilize the observation information and provide more accurate initial values becomes the core of numerical weather forecast research. Data assimilation improves the initial condition of the pattern by fusing the various known information available (including observations, patterns and corresponding error statistics, etc.). The meteorological satellite observation of infrared hyperspectrum and the like can realize the ground detection under All-weather (All-sky) conditions from the outside of the atmosphere, is not limited by the type of the ground surface, has the characteristics of large information quantity, high space-time resolution and the like, and can effectively fill the 'information blind area' which is difficult to detect in the conventional observation. Numerical forecasting has improved significantly since satellite data was processed using variational assimilation, taking the four-dimensional variational data assimilation (4DVAR) system, which is a business of the mid-european weather forecast center (ECMWF), as an example, 90% of the observed information is derived from satellite data.
The observation error covariance matrix plays an important role in the variation assimilation system, and determines the importance of observation information and the diffusion mode of the information in space and between different variables. The more non-zero elements of the observation error covariance matrix, the more information that can be extracted from the observation, in the sense that the observation error covariance matrix should be as full a matrix as possible. In general, the variance of diagonal elements constituting the observation error covariance mainly represents the instrument error, and the part of information is easy to acquire; off-diagonal elements characterize the correlation between observations, which is estimated a posteriori and is relatively difficult to estimate accurately. Background error covariance matrices have been developed to a considerable degree of complexity since the start of data assimilation, but the observation error covariance is extremely simple, usually ignoring off-diagonal elements and considering only the variance made up of diagonal elements. For example, when infrared hyperspectral satellite data are assimilated, it is simply assumed that errors between observations are uncorrelated, field-of-view thinning is correspondingly performed and non-adjacent channels are selected, and the error variance is generally given according to statistical experience.
In order to make up for the deficiency that the covariance matrix of the observation errors does not take into account the correlation of the errors, the observation errors are usually artificially amplified, i.e. an observation error expansion technique, so that the weights of the observations in the analysis are consistent with the true weights. However, the observation errors of the satellite data have diversity, including not only the instrument errors but also the representative errors, and there is a correlation between the observations. These correlations include not only spatial correlations, but especially strong channel correlations for thousands of channels of infrared hyperspectrum. Research shows that the error correlation degree among the channels changes along with atmospheric conditions, the correlation degree is small under Clear-sky (Clear-sky) conditions, and the channel correlation degree is more obvious under all-weather cloud and rain conditions. At present, research is primarily tried to consider the correlation among channels for infrared hyperspectral data in an observation error covariance matrix, and the correlation coefficient and the error variance of the channels are diagnosed by using posterior information, but the real-time change of the observation error along with the cloud water under all-weather conditions is not fully considered.
At present, an ECMWF (equal numerical prediction center) business data assimilation system mainly adopts a clear sky assimilation method to assimilate infrared hyperspectral data, an observation error covariance model used in the data assimilation system is simple, and the error level of the infrared hyperspectral data in a cloud and rain area under all-weather conditions is difficult to accurately describe. The numerical simulation application value of the infrared hyperspectral observation data is fully exerted, and an all-weather data assimilation system is constructed by means of an advanced observation error covariance model. The error model is required to change along with the cloud water condition in real time and is closely related to the cloud water substance in the field of view.
The existing four-dimensional variation assimilation method for the infrared hyperspectral data only can assimilate a channel which is not influenced by clouds above a clear sky view field and a full cloud view field cloud top due to the fact that an observation error covariance model is simple to a certain extent, available effective observation is very limited (the number of the available effective observation is less than 10% approximately), and the application value of the infrared hyperspectral data is not fully exerted.
Disclosure of Invention
In view of the above, the invention provides a numerical weather forecasting method based on an all-weather observation error covariance matrix, and provides an infrared hyperspectral all-weather observation error covariance model based on a cloud scene to solve the problems that an observation error covariance model adopted by a current business variation assimilation system cannot accurately describe the infrared hyperspectral observation error channel correlation and the observation error variance level under all-weather conditions, and an infrared hyperspectral observation error covariance channel correlation technology based on a cloud characteristic function and an observation error variance expansion technology based on cloud amount;
the invention discloses a numerical weather forecasting method based on an all-weather observation error covariance matrix, which comprises the following steps of:
s1: the method comprises the steps of selecting infrared hyperspectral observation data for assimilation for a period of time, constructing an observation error covariance matrix for errors of satellite observation data, and performing decomposition and inversion calculation on the observation error covariance matrix to obtain an observation error correlation matrix;
s2: the observation error correlation matrix is realized by a cost function;
s3: calculating deviation of the simulated cloud radiance and the simulated observation radiance and the clear air radiance respectively to obtain the cloud radiance of a background field and the cloud radiance and the cloud characteristic function contained in observation;
s4: constructing expansion factors of different channels, and forming a diagonalized expansion matrix by the expansion factors;
s5: and substituting the cloud characteristic function and the cloud amount function into the cost function, calculating the newly constructed cost function and the gradient thereof in an assimilation system, solving an optimal analysis field by a minimalization solution, and performing numerical weather forecast according to the optimal analysis field.
Further, the step of S1 includes the following sub-steps:
s11: selecting a period of infrared hyperspectral observation data for assimilation, diagnosing and analyzing errors of satellite observation data, and utilizing a covariance matrix of posterior observation errors
Figure BDA0003308356950000031
Calculating the covariance matrix of the observation errors by a calculation formula, and the covariance matrix of the posterior observation errors
Figure BDA0003308356950000032
The calculation is as follows:
Figure BDA0003308356950000033
wherein E [ alpha ], [ beta ], [ alpha ], [ beta ]]The expression is used for solving the mathematical expectation,
Figure BDA0003308356950000034
difference values for observation and analysis;
s12: respectively counting the covariance of each channel combination to obtain an observation error covariance matrix, wherein the observation error covariance R (i, j) of the ith channel and the jth channel is as follows:
Figure BDA0003308356950000035
s13: the elements of the observation error covariance matrix R are constructed from the observation error covariance R (i, j) between each channel.
Figure BDA0003308356950000041
S14: carrying out block diagonalization decomposition on the infrared hyperspectral observation data error covariance matrix;
s15: diagonalizing a block to observe an error covariance matrix RkDiagonalization conversion is carried out, and an empirical orthogonal expansion method is adopted to realize the diagonalization conversion, and the method comprises the following steps:
Rk＝∑C∑
wherein, sigma is an observation error standard deviation matrix belonging to a diagonal matrix, and the diagonal element is the observation error standard deviation sigma of the ith channeliC is a correlation coefficient matrix between observation error channels;
s16: for the decomposed observation error covariance matrix RkCarry out inversion
Figure BDA0003308356950000042
The inversion formula is as follows:
Figure BDA0003308356950000043
wherein, Λ is the characteristic value λ of the matrix C of the correlation coefficients among the observation error channelsjA diagonal matrix formed by E is an eigenvector E of a matrix C of correlation coefficients between observation error channelsjA composed orthogonal matrix.
Further, the cost function J includes an observation item JoAnd background field item JbExpressed as:
J＝Jo+Jb
cost function observation item JoExpressed as:
Figure BDA0003308356950000044
wherein d is the observed increment, σoWhen the cost function is solved in an incremental mode for observing the standard deviation of the error, the observation item JoThe gradient of (d) is:
Figure BDA0003308356950000045
wherein HTIs the transpose of a tangent observation operator with a column vector of hiFor measuring the sensitivity of the observation to changes in atmospheric conditions, σoStandard deviation of observation error.
Further, a cloud feature function CcldThe calculation is as follows:
Figure BDA0003308356950000046
wherein Hclr(xb) Is cloud emissivity, Hcld(xb) The radiance is clear sky.
Further, the expansion factors S of the different channels j are constructedjFrom SiForming a diagonalized expansion matrix S, expanding the observation errors:
R＝ES0.5ΛS0.5ET≡ESΛET，
where E is the eigenvector E of the matrix of correlation coefficients between the channels of observation errorsjA composed orthogonal matrix.
Further, the first eigenvalue expansion factor is S1The coefficient of expansion of the other characteristic value being set to a constant, i.e. SjThe value-taking strategy is expressed as:
Figure BDA0003308356950000051
wherein S is1Expressed as:
Figure BDA0003308356950000052
a is the coefficient of the minimum expansion scale factor under the clear sky condition, b is the coefficient of the maximum expansion scale factor under the cloud condition, CaAs a function of the cloud cover.
Further, the cloud quanta function CaBy simulating cloud cover CmAnd observing the cloud cover CoCalculate to obtain the tableShown below:
Cm＝Bcld-Bclr
Co＝Ocld-Bclr
Ca＝(Cm+Co)/2
wherein B iscldLight temperature representing RTTOV vs. simulation of cloudy profile, BclrLight temperature, O, representing RTTOV vs. simulation of cloudy contourscldIt indicates a cloudy bright temperature was observed.
The invention has the following beneficial effects:
an observation error covariance model based on cloud characteristic function observation error correlation is researched aiming at an upper troposphere water vapor channel in satellite infrared hyperspectral observation data, characteristic deviation and characteristic Jacobian ratio related to observation error covariance matrix error are obtained through the construction of a cloud characteristic function in a visual field, and the observation error correlation is effectively utilized to assimilate the infrared hyperspectral data.
The method is based on a cloud cover observation error expansion technology, establishes a function corresponding relation of observation error variance changing along with the cloud cover, explores different characteristic channels, adopts different cloud cover expansion values and observation error covariance characteristic value threshold truncation expansion schemes, and provides an effective and practical way for realizing different degrees of influence of the channels with different error levels on a cost function.
Drawings
FIG. 1 is a flow chart of a numerical weather forecasting method based on an all-weather observation error covariance matrix.
Detailed Description
The invention is further described with reference to the accompanying drawings, but the invention is not limited in any way, and any alterations or substitutions based on the teaching of the invention are within the scope of the invention.
As shown in FIG. 1, the method for numerical weather forecast based on all-weather observation error covariance matrix of the present invention comprises the following steps:
s1: and selecting infrared hyperspectral observation data for assimilation for a period of time, constructing an observation error covariance matrix for errors of the satellite observation data, and decomposing and inverting the observation error covariance matrix to obtain an observation error correlation matrix.
Under all-weather observation conditions, the infrared hyperspectrum is influenced by cloud water, the error level is complex relative to clear sky observation, and the construction of an observation error mode is very key for assimilating all-weather infrared hyperspectral observation data. Firstly, the posterior information of an assimilation system is utilized to diagnose an infrared hyperspectral observation error covariance matrix under a clear air condition, and a basic observation error covariance matrix is constructed for introducing influence of cloud water. The method comprises the following specific steps:
s11: selecting a period of infrared hyperspectral observation data for assimilation, diagnosing and analyzing errors of satellite observation data, and utilizing a covariance matrix of posterior observation errors
Figure BDA0003308356950000067
The calculation formula calculates the covariance matrix of the observation error of the GIIRS (interferometric atmospheric vertical detector).
Figure BDA0003308356950000061
Wherein E [ alpha ], [ beta ], [ alpha ], [ beta ]]The expression is used for solving the mathematical expectation,
Figure BDA0003308356950000062
representing the difference between the observed and background fields (information delta or referred to as observation delta),
Figure BDA0003308356950000063
wherein xbAs background field, y as satellite GIIRS observation, H as observation operator, if bias correction is taken into account
Figure BDA0003308356950000064
Where b is a bias correction parameter.
Figure BDA0003308356950000065
For observation and analysisThe difference value is obtained by comparing the difference value,
Figure BDA0003308356950000066
wherein δ xaTo analyze the increments. Assuming that the error is Gaussian, the observed and background errors are uncorrelated, and the weight of the observation in the analysis is the same as when the true error signature is used, then
Figure BDA0003308356950000071
S12: their covariance is counted separately for each channel combination. Wherein, the observation error covariance R (i,) of the ith channel and the jth channel is:
Figure BDA0003308356950000072
s13: the elements of the observation error covariance matrix R are constructed from the observation error covariance R (i, j) between each channel.
Figure BDA0003308356950000073
S14: and performing block diagonalization decomposition by using the infrared hyperspectral observation data error covariance matrix diagnosed in the step S12. Wherein, the error covariance block of the water vapor channel selected for research is assumed to be Rk(considering the observation error correlation, the observation error correlation is a square matrix of n × n dimensions), and represents the covariance matrix of the observation errors of n channels in the k-th observation set of the whole infrared hyperspectral channel set, and if 7 GIIRS water vapor channels are planned to be selected, n is 7.
S14: for the block diagonalized observation error covariance matrix R obtained at S13kAnd carrying out diagonalization conversion, and realizing by methods such as empirical orthogonal expansion and the like.
Rk＝∑C∑
Wherein, Sigma is an observation error standard deviation matrix belonging to a diagonal matrix, and diagonal elements are the observation error standard deviation sigma of the ith channeliAnd C is a matrix of correlation coefficients between observation error channels.
S15: r after decomposition of S14kCarry out inversion
Figure BDA0003308356950000074
The inversion formula is as follows:
Figure BDA0003308356950000075
Figure BDA0003308356950000077
the coefficient matrix needs to be matrix decomposed again, then
Figure BDA0003308356950000076
Expressed as:
Figure BDA0003308356950000081
where Λ is the eigenvalue λ of the C matrixjA diagonal matrix is formed, E is a characteristic vector E of the C matrixjA composed orthogonal matrix.
S2: realization of observation error correlation matrix in cost function
The data assimilation cost function J includes an observation item JoAnd background field item JbExpressed as:
J＝Jo+Jb
when channel correlation is not considered, R is a diagonal matrix considering variance only, and a cost function observation item JoExpressed as:
Figure BDA0003308356950000082
wherein d is an observation increment, and when the cost function is solved in an increment mode, the gradient of the observation item is as follows:
Figure BDA0003308356950000083
wherein HTIs the transpose of a tangential observation operator, also called Jacobian matrix, whose column vector is hiAnd is typically used to measure the sensitivity of the observation to changes in atmospheric conditions.
When channel correlation is considered, R matrix inversion cannot be simply expressed as inversion of a variance diagonal matrix, and form inversion after observation error correlation coefficient matrix decomposition needs to be used, and a cost function observation item is expressed as:
Figure BDA0003308356950000084
wherein the content of the first and second substances,
Figure BDA0003308356950000085
in order to be a transpose of the feature vector,
Figure BDA0003308356950000086
is a characteristic deviation term, λjIs the eigenvalue of the matrix of the correlation coefficients between the observed error channels.
The gradient corresponding to the observation term is:
Figure BDA0003308356950000087
HTejis a characteristic jacobian and is used to measure the sensitivity of the observation to changes in atmospheric conditions.
S3: calculating deviation of the simulated cloud radiance and the simulated observation radiance and the deviation of the simulated observation radiance and the simulated clear air radiance, and obtaining the cloud radiance of the background field, the cloud radiance contained in observation and a cloud characteristic function.
For the calculation of the cloud characteristic function and the cloud cover, the cloud radiance H is respectively calculated for the channels by using an RTTOV (remote time-of-flight) rapid radiation transmission modeclr(xb) And clear sky radiance Hcld(xb). With the most affected cloud, i.e. the weighting function being located at the mostThe bottom layer channel is a research object, then the simulated cloud radiance and the simulated observation radiance are respectively deviated from the clear sky radiance, namely the simulated clear sky radiance is taken as a reference basis, the cloud radiance of the background field and the cloud radiance included in observation are obtained, finally weighted averaging is carried out, and the cloud characteristic function C is obtainedcld：
Figure BDA0003308356950000091
When the channel correlation is considered, the channel correlation,
Figure BDA0003308356950000092
expressed as normalized feature information augmentative term, using CcldInstead, it can be implemented to consider cloud-feature based error correlation in the observed error model.
S4: constructing expansion factors of different channels, and forming a diagonalized expansion matrix by the expansion factors;
considering the influence of cloud cover on the standard deviation of observation errors, the expansion factors S of different channels j need to be constructedjFrom SiA diagonalized expansion matrix S is constructed.
Figure BDA0003308356950000093
Thus, the observation error dilation can be expressed as:
R＝ES0.5ΛS0.5ET≡ESΛET
Rkthe eigenvalue expansion of the matrix draws up a strategy of adopting different eigenvalue types, and the expansion factor of the first eigenvalue is S1The expansion coefficient of other characteristic values is set to 1 (or other constants are expanded, and an empirical value is obtained through testing according to experiments), namely SjThe value-taking strategy of (a) can be expressed as:
Figure BDA0003308356950000094
wherein S is1Expressed as:
Figure BDA0003308356950000095
a is the minimum expansion scale factor coefficient (for example, the value is 0.2) under the clear sky condition, b is the maximum expansion scale factor coefficient (for example, the value is 3.2) under the cloud condition, and CaAs a function of the cloud cover.
Cloud function CaBy simulating cloud cover CmAnd observing the cloud cover CoCalculated as follows:
Cm＝Bcld-Bclr
Co＝Ocld-Bclr
Ca＝(Cm+Co)/2
wherein B iscldLight temperature representing RTTOV vs. simulation of cloudy profile, BclrLight temperature, O, representing RTTOV vs. simulation of cloudy contourscldIt indicates a cloudy bright temperature was observed. Cloud cover function C when using radiance bright temperature to represent cloud coveraAnd cloud characteristic function CcldHas equivalence.
S5: and substituting the cloud characteristic function and the cloud amount function into the cost function, calculating the newly constructed cost function and the gradient thereof in an assimilation system, solving an optimal analysis field by a minimalization solution, and performing numerical weather forecast according to the optimal analysis field.
And applying the cloud characteristic functions and the cloud quantum functions obtained by the calculation in the steps S3 and S4 to the cost function observation items in the step S2 to realize an observation error covariance model and corresponding assimilation cost functions based on the cloud scene.
And calculating a newly constructed cost function and the gradient thereof in an assimilation system, minimizing a solution to solve an optimal analysis field, and verifying the convergence of the cost function.
The invention has the following beneficial effects:
aiming at an upper troposphere water vapor channel in the infrared hyperspectral observation data of the FY-4A satellite, an observation error covariance model based on the correlation of the observation error of the cloud characteristic function is researched, the characteristic deviation and the characteristic Jacobian ratio related to the error of the observation error covariance matrix are obtained through the construction of the cloud characteristic function in the visual field, and the correlation of the observation error is effectively utilized to assimilate the infrared hyperspectral data.
The method is based on a cloud cover observation error expansion technology, establishes a function corresponding relation of observation error variance changing along with the cloud cover, explores different characteristic channels, adopts different cloud cover expansion values and observation error covariance characteristic value threshold truncation expansion schemes, and provides an effective and practical way for realizing different degrees of influence of the channels with different error levels on a cost function.
The above embodiment is an embodiment of the present invention, but the embodiment of the present invention is not limited by the above embodiment, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be regarded as equivalent replacements within the protection scope of the present invention.

Claims (7)

1. A numerical weather forecasting method based on an all-weather observation error covariance matrix is applied to an assimilation system and is characterized by comprising the following steps of:
s1: the method comprises the steps of selecting infrared hyperspectral observation data for assimilation for a period of time, constructing an observation error covariance matrix for errors of satellite observation data, and performing decomposition and inversion calculation on the observation error covariance matrix to obtain an observation error correlation matrix;
s2: the observation error correlation matrix is realized by a cost function;
s3: calculating deviation of the simulated cloud radiance and the simulated observation radiance and the clear air radiance respectively to obtain the cloud radiance of a background field and the cloud radiance and the cloud characteristic function contained in observation;
s4: constructing expansion factors of different channels, and forming a diagonalized expansion matrix by the expansion factors;
s5: and substituting the cloud characteristic function and the cloud amount function into the cost function, calculating the newly constructed cost function and the gradient thereof in an assimilation system, solving an optimal analysis field by a minimalization solution, and performing numerical weather forecast according to the optimal analysis field.
2. The all-weather observation error covariance matrix-based numerical weather forecasting method of claim 1, wherein the step of S1 comprises the sub-steps of:
s11: the method comprises the following steps of selecting infrared hyperspectral observation data for assimilation for a period of time, and diagnosing and analyzing errors of the satellite observation data; error epsilon of the observation datao＝y-ytWherein y represents an infrared hyperspectral observed value measured by the instrument, ytShowing an infrared hyperspectral truth value of the atmosphere;
calculating a posterior observation error covariance matrix
Figure FDA0003308356940000011
The specific formula is as follows:
Figure FDA0003308356940000012
wherein E [ alpha ], [ beta ], [ alpha ], [ beta ]]The expression is used for solving the mathematical expectation,
Figure FDA0003308356940000013
difference values for observation and analysis;
s12: respectively counting the covariance of each channel combination to obtain an observation error covariance matrix R, wherein the observation error covariance R (i, j) of the ith channel and the jth channel is as follows:
Figure FDA0003308356940000014
wherein the content of the first and second substances,
Figure FDA0003308356940000015
representing the difference between the observed and background fieldsN is the number of channels;
s13: constructing elements of an observation error covariance matrix R from the observation error covariance R (i, j) between each channel;
Figure FDA0003308356940000021
s14: carrying out block diagonalization decomposition on the infrared hyperspectral observation data error covariance matrix;
s15: diagonalizing a block to observe an error covariance matrix RkDiagonalization conversion is carried out, and an empirical orthogonal expansion method is adopted to realize the diagonalization conversion, and the method comprises the following steps:
Rk＝ΣCΣ
wherein, Σ is an observation error standard deviation matrix, belongs to a diagonal matrix, and the diagonal element is the observation error standard deviation σ of the ith channeliC is a correlation coefficient matrix between observation error channels;
s16: for the decomposed observation error covariance matrix RkCarry out inversion
Figure FDA0003308356940000025
The inversion formula is as follows:
Figure FDA0003308356940000022
wherein, Λ is the characteristic value λ of the matrix C of the correlation coefficients among the observation error channelsjA diagonal matrix formed by E is an eigenvector E of a matrix C of correlation coefficients between observation error channelsjA composed orthogonal matrix.
3. The method of numerical weather forecasting based on all-weather observation error covariance matrix of claim 1, wherein the cost function J comprises an observation item JoAnd background field item JbExpressed as:
J＝Jo+Jb
cost function observation item JoExpressed as:
Figure FDA0003308356940000023
wherein d is the observed increment, σoWhen the cost function is solved in an incremental mode for observing the standard deviation of the error, the observation item JoThe gradient of (d) is:
Figure FDA0003308356940000024
wherein HTIs the transpose of a tangent observation operator with a column vector of hiFor measuring the sensitivity of the observation to changes in atmospheric conditions, σoStandard deviation of observation error.
4. The method for numerical weather forecasting based on all-weather observation error covariance matrix as claimed in claim 1, wherein the cloud eigenfunction CcldThe calculation is as follows:
Figure FDA0003308356940000031
wherein Hclr(xb) Is cloud emissivity, Hcld(xb) The radiance is clear sky.
5. The method of numerical weather prediction based on all-weather observation error covariance matrix of claim 1, wherein the dilation factor S of different channels j is constructedjFrom SiForming a diagonalized expansion matrix S, expanding the observation errors:
R＝ES0.5ΛS0.5ET≡ESΛET，
where E is the eigenvector E of the matrix of correlation coefficients between the channels of observation errorsjOfAn orthogonal matrix.
6. The method of numerical weather forecasting based on all-weather observation error covariance matrix of claim 1, wherein the first eigenvalue inflation factor is S1The coefficient of expansion of the other characteristic value being set to a constant, i.e. SjThe value-taking strategy is expressed as:
Figure FDA0003308356940000032
wherein S is1Expressed as:
Figure FDA0003308356940000033
a is the coefficient of the minimum expansion scale factor under the clear sky condition, b is the coefficient of the maximum expansion scale factor under the cloud condition, CaAs a function of the cloud cover.
7. The method of numerical weather prediction based on all-weather observation error covariance matrix of claim 6, wherein the cloud cover function CaBy simulating cloud cover CmAnd observing the cloud cover CoCalculated as follows:
Cm＝Bcld-Bclr
Co＝Ocld-Bclr
Ca＝(Cm+Co)/2
wherein B iscldLight temperature representing RTTOV vs. simulation of cloudy profile, BclrLight temperature, O, representing RTTOV vs. simulation of cloudy contourscldIt indicates a cloudy bright temperature was observed.
CN202111209618.5A 2021-10-18 2021-10-18 Numerical weather forecasting method based on all-weather observation error covariance matrix Pending CN114037125A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111209618.5A CN114037125A (en) 2021-10-18 2021-10-18 Numerical weather forecasting method based on all-weather observation error covariance matrix

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111209618.5A CN114037125A (en) 2021-10-18 2021-10-18 Numerical weather forecasting method based on all-weather observation error covariance matrix

Publications (1)

Publication Number Publication Date
CN114037125A true CN114037125A (en) 2022-02-11

Family

ID=80135252

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111209618.5A Pending CN114037125A (en) 2021-10-18 2021-10-18 Numerical weather forecasting method based on all-weather observation error covariance matrix

Country Status (1)

Country Link
CN (1) CN114037125A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115165786A (en) * 2022-07-06 2022-10-11 国家卫星气象中心(国家空间天气监测预警中心) Equivalent clear sky radiation correction method for infrared hyperspectral atmosphere detector based on imager

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115165786A (en) * 2022-07-06 2022-10-11 国家卫星气象中心(国家空间天气监测预警中心) Equivalent clear sky radiation correction method for infrared hyperspectral atmosphere detector based on imager
CN115165786B (en) * 2022-07-06 2023-09-19 国家卫星气象中心(国家空间天气监测预警中心) Equivalent clear sky radiation correction method for infrared hyperspectral atmospheric detector based on imager

Similar Documents

Publication Publication Date Title
Liu et al. Uncertainty analysis of eleven multisource soil moisture products in the third pole environment based on the three-corned hat method
Bi et al. An improved particle filter algorithm based on ensemble Kalman filter and Markov chain Monte Carlo method
Loew et al. A dynamic approach for evaluating coarse scale satellite soil moisture products
Ingram et al. Sensitivity of iterative spectrally smooth temperature/emissivity separation to algorithmic assumptions and measurement noise
Román-Cascón et al. Correcting satellite-based precipitation products through SMOS soil moisture data assimilation in two land-surface models of different complexity: API and SURFEX
CN114966692B (en) Transformer-based InSAR technology frozen soil area multivariable time sequence deformation prediction method and device
Gruber et al. The potential of 2D Kalman filtering for soil moisture data assimilation
Thomas et al. Data assimilation for convective-cell tracking on meteorological image sequences
Aires et al. Statistical approaches to assimilate ASCAT soil moisture information—I. Methodologies and first assessment
Yao et al. Assessment and simulation of global terrestrial latent heat flux by synthesis of CMIP5 climate models and surface eddy covariance observations
CN109212631B (en) Satellite observation data three-dimensional variation assimilation method considering channel correlation
CN113468799A (en) Method and system for acquiring near-ground PM2.5 concentration in static meteorological satellite observation
CN113466143A (en) Soil nutrient inversion method, device, equipment and medium
CN115859789A (en) Method for improving inversion accuracy of polar atmosphere temperature profile
Feng et al. Performance of the improved two-source energy balance model for estimating evapotranspiration over the heterogeneous surface
Merchant et al. Bias correction and covariance parameters for optimal estimation by exploiting matched in-situ references
CN114037125A (en) Numerical weather forecasting method based on all-weather observation error covariance matrix
Hu et al. The data-driven solution of energy imbalance-induced structural error in evapotranspiration models
Semane et al. Comparison of Desroziers and three-cornered hat methods for estimating COSMIC-2 bending angle uncertainties
Mangiarotti et al. Evolutionary bi-objective optimization of a semi-arid vegetation dynamics model with NDVI and σ0 satellite data
Thao et al. Comparison of regression algorithms for the retrieval of the wet tropospheric path
Legrand et al. Diagnosing non-Gaussianity of forecast and analysis errors in a convective-scale model
De Lannoy et al. Adaptive soil moisture profile filtering for horizontal information propagation in the independent column-based CLM2. 0
CN114047563A (en) All-weather assimilation method of infrared hyperspectrum
Tian et al. Improving MODIS aerosol estimates over land with the surface BRDF reflectances using the 3-D discrete cosine transform and RossThick-LiSparse models

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