WO2020096141A1 - Method for assimilating radar melting layer elevation data on basis of ultra-short forecast model, and recording medium and apparatus for performing same - Google Patents

Method for assimilating radar melting layer elevation data on basis of ultra-short forecast model, and recording medium and apparatus for performing same Download PDF

Info

Publication number
WO2020096141A1
WO2020096141A1 PCT/KR2019/002464 KR2019002464W WO2020096141A1 WO 2020096141 A1 WO2020096141 A1 WO 2020096141A1 KR 2019002464 W KR2019002464 W KR 2019002464W WO 2020096141 A1 WO2020096141 A1 WO 2020096141A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
observation data
radar
short
altitude
Prior art date
Application number
PCT/KR2019/002464
Other languages
French (fr)
Korean (ko)
Inventor
민기홍
배정호
Original Assignee
경북대학교 산학협력단
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 경북대학교 산학협력단 filed Critical 경북대학교 산학협력단
Publication of WO2020096141A1 publication Critical patent/WO2020096141A1/en

Links

Images

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01SRADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S13/00Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
    • G01S13/88Radar or analogous systems specially adapted for specific applications
    • G01S13/95Radar or analogous systems specially adapted for specific applications for meteorological use
    • G01S13/951Radar or analogous systems specially adapted for specific applications for meteorological use ground based
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01WMETEOROLOGY
    • G01W1/00Meteorology
    • G01W1/14Rainfall or precipitation gauges
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01WMETEOROLOGY
    • G01W2201/00Weather detection, monitoring or forecasting for establishing the amount of global warming
    • YGENERAL 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
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02ATECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Definitions

  • the present invention relates to a method for assembling altitude data of a radar fusion layer based on a very short-term forecasting model, and a recording medium and apparatus for performing the same, more specifically, a very short-term to improve the initial field of a numerical forecasting model using observation data of a weather radar. It relates to a method of assimilation of altitude data in a radar melting layer based on a forecast model, and a recording medium and apparatus for performing the same.
  • a weather radar is a device that calculates the size of a radio wave signal that is reflected or scattered from a meteorological target by firing electromagnetic waves, and can monitor a large area (effective observation radius: 240 km) very quickly (every 10 minutes). It is one of the most efficient remote sensing equipment for calculating a large amount of precipitation.
  • the reflectivity which is a radio wave signal reflected or scattered from the target, is related to the size distribution of water droplets present in the pulse volume emitted from the weather radar.
  • the radar reflectivity obtained from a weather radar when snowfall particles fall at a high altitude of 0 ° C or lower and pass through the isothermal layer at 0 ° C, the outer surface of the snowfall particle starts to melt and the reflectance value rapidly increases in this layer. This is due to the increased dielectric constant and size effect.
  • the altitude decreases almost all snowfall particles melt to form water droplets, and the reflectance value decreases again. As such, the reflectance value increases rapidly and then decreases again, until a certain reflectivity appears, and the area is bright band (Bright Band). It is called an area, and the isothermal layer at 0 °C is called the altitude at which snowfall particles begin to melt, that is, the altitude of the melted layer.
  • One aspect of the present invention is an ultra-short-term forecasting model that calculates the altitude and temperature of the melted layer from reflectivity data obtained from a weather radar, and preprocesses the altitude and temperature of the melted layer to be applied to data assimilation to improve the initial field of the ultra-short-term forecasting model.
  • a method for assimilation of advanced data based on a radar fusion layer and a recording medium and apparatus for performing the same.
  • the method of assimilation of the elevation data of the radar melting layer based on the ultra-short-term forecasting model comprises collecting radar observation data from a weather radar system and calculating observation data including altitude and temperature of the melting layer. Pre-processing to be applicable to data assimilation, calculating analytic increments by applying the 3D variable data assimilation method between the background fields of the ultra-short-term forecast model and the initial field of the ultra-short-term forecast model using the analysis increments And generating and predicting precipitation.
  • collecting the radar observation data from the weather radar system and calculating the observation data including the altitude and temperature of the melting layer includes: detecting a bright band from the radar observation data, and converting the bright band into the elevation of the melting layer It may include a step of calculating and comparing the altitude of the fusion layer with the Lewinsonde data to calculate the temperature of the fusion layer.
  • collecting radar observation data from the weather radar system and calculating observation data including altitude and temperature of the melting layer may include processing the observation data in a buffer form and storing the observation data in a database.
  • the pre-processing so that the observation data can be applied to data assimilation includes interpolating the altitude of the melting layer included in the observation data into a grid of a numerical prediction model and spatial information corresponding to the elevation of the melting layer. It may include the step of extracting the background field variable of the numerical prediction model for.
  • the step of pre-processing the observation data to be applied to the data assimilation is a total of 24 vertical standard altitude arrays by adding four arrays vertically from 20 vertical standard altitude arrays of 1000 hPa to 1 hPa sections to 750 hPa to 550 hPa sections.
  • the step of calculating an analysis increment by applying a three-dimensional variable data assimilation method between the background fields of the very short-term forecast model is the equation of the observed data and the early short-term forecast model through an external iterative cycle and an internal iterative cycle. Calculating the cost function and slope for the background field, and using the slope value between the initial field of the early short-term forecast model before the observation data is applied and the initial field of the early short-term forecast model when the observed data is applied. And calculating an analytical increment.
  • the step of generating an initial field of an early short-term forecast model and predicting precipitation using the analysis increment may include applying the analysis increment to a vertical temperature profile of the initial field of the early short-term forecast model.
  • the computer may be a computer-readable recording medium in which a computer program is recorded for performing the method for assimilation of the advanced data for the radar fusion layer based on the ultra-short-term forecast model.
  • the radar fusion layer elevation data assimilation device based on the ultra-short-term forecast model for solving the above problems is an observation data collection unit that collects radar observation data from a weather radar system and calculates observation data including altitude and temperature of the fusion layer.
  • An observation data pre-processing unit that pre-processes the observation data to be applied to the data assimilation, a variable data assimilation unit that calculates an analysis increment by applying a three-dimensional variable data assimilation method between the background fields of the very short-term forecast model, and the It includes an initial field prediction unit that generates an initial field of the ultra-short-term forecast model using the incremental analysis and predicts precipitation.
  • the observation data collection unit detects a bright band from the radar observation data, calculates the bright band at the altitude of the fusion layer, and compares the altitude of the fusion layer with the Lewinsonde data to calculate the melting layer temperature. Can be.
  • observation data collection unit may process the observation data in a buffer form and store it in a database.
  • observation data pre-processing unit interpolates the altitude of the melting layer included in the observation data into the grid of the numerical prediction model, and sets the background field variable of the numerical prediction model for spatial information corresponding to the elevation of the melting layer. Can be extracted.
  • observation data pre-processing unit adds four arrays vertically in the range of 20 vertical standard altitudes in the range of 1000 hPa to 1 hPa in the range of 750 hPa to 550 hPa to generate a total of 24 vertical standard altitude arrays, and in the vertical standard altitude array
  • an observation data variable is extracted, and the same variable as the observation data variable can be extracted from the background field of the very short-term forecast model as the background field variable.
  • variable data assimilation unit calculates a cost function and a slope for the background of the observation data and the very short-term forecast model by using an external iterative cycle and an internal iterative cycle.
  • An analysis increment may be calculated between the initial field of the early short-term forecast model before application and the initial field of the early short-term forecast model when the observation data is applied.
  • the initial field prediction unit may apply the analysis increment to the vertical temperature profile of the initial field of the ultra-short forecast model.
  • the spatial and temporal resolution is low, while in the case of radar reflectivity data, the spatial and temporal resolution is high, so it is possible to calculate the altitude and temperature of the melting layer more accurately.
  • the initial temperature information of the upper, middle and lower layers of the temperature field simulated in the numerical prediction model is obtained. Can be corrected.
  • FIG. 1 is a block diagram of a radar fusion layer advanced data assimilation device based on an ultra-short-term forecast model according to an embodiment of the present invention.
  • FIG. 2 is a view for explaining a method of calculating the altitude and temperature of the melting layer in the observation data collection unit shown in FIG. 1.
  • 3 is a view for explaining the characteristics of the bright band appearing in the radar reflectivity data.
  • FIG. 4 is a diagram for explaining a 3D variable data assimilation method.
  • FIG. 5 is a flowchart of a method for assimilation of an advanced data of a radar fusion layer based on an ultra-short-term forecast model according to an embodiment of the present invention.
  • 6A to 7D are graphs showing bright band altitudes calculated from radar observation data in a case of stratified clouds.
  • 8A and 8B are graphs showing the vertical temperature profile of the initial dynamic field generated according to the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention in the rainy season case.
  • 9A to 11E are graphs showing changes in dynamic field according to an assimilation method of the radar melting layer elevation data based on the ultra-short-term forecast model of the present invention in the rainy season front-line case.
  • 12A to 12D are graphs comparing the precipitation simulation result between the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention and the precipitation simulation result between the normative experiments.
  • FIG. 1 is a block diagram of a radar fusion layer advanced data assimilation device based on an ultra-short-term forecast model according to an embodiment of the present invention.
  • the radar fusion layer elevation data assimilation device based on the ultra-short-term forecasting model is an apparatus for improving the initial field of the ultra-short-term forecasting model using the observation data of the weather radar system 1 , It is connected to the weather radar system 1 by wired or wireless resources, refers to the weather radar system 1, or may include some or all of the functions of the weather radar system 1.
  • VDAPS Very Short Time Range Data Assimilation and Prediction System
  • ODB Observation DataBase
  • OPS Observation Processing System
  • VAR three dimensional VARiational data assimilation system
  • prediction forecast
  • the apparatus 100 of the present invention calculates altitude and temperature information of the melted layer using the observation data of the weather radar system 1, and applies the altitude and temperature information of the melted layer to an ultra-short-term forecast model to preprocess and assimilate data.
  • the initial dynamic field of the very short-term forecast model can be corrected by performing.
  • the apparatus 100 of the present invention may include an observation data collection unit 110, an observation data pre-processing unit 130, a variable data assimilation unit 150, and an initial field prediction unit 170.
  • the configuration of the observation data collection unit 110, the observation data pre-processing unit 130, the variable data assimilation unit 150, and the initial field prediction unit 170 shown in FIG. 1 is formed of an integrated module, or composed of one or more modules Can be. However, on the contrary, each configuration may be made of a separate module.
  • the device 100 according to the present invention may have mobility or be fixed.
  • the device 100 according to the present invention may be in the form of a server or an engine, a device, an apparatus, a terminal, a user equipment (UE), a mobile station (MS), It can also be called another term, such as a wireless device or a handheld device.
  • UE user equipment
  • MS mobile station
  • the device 100 may execute or manufacture various software based on an operating system (OS), that is, a system.
  • OS operating system
  • the operating system is a system program to enable the software to use the hardware of the device, such as Android OS, iOS, Windows Mobile OS, Sea OS, Symbian OS, and BlackBerry OS Mobile computer operating system and Windows, Linux, Unix, MAC , AIX, HP-UX, and computer operating systems.
  • the apparatus 100 may be installed and executed software (application) for assembling high-level radar melting layer data based on the ultra-short-term forecast model, observation data collection unit 110, observation data pre-processing unit 130, variation
  • the configuration of the data assimilation unit 150 and the initial field prediction unit 170 may be controlled by software executed in the device 100.
  • the observation data collection unit 110 may perform a physical process of the observation data database (ODB) step in the very short-term forecast model.
  • ODB observation data database
  • observation data from the synoptic scale eg, AWS, maritime buoy, wind profiler and Lewinsonde
  • observation data such as temperature, relative humidity, rain humidity, and wind information
  • BUFR Binary Universal Form for the Representation of meteorological data
  • the observation data collection unit 110 may collect radar observation data obtained from the weather radar system 1 to calculate the altitude and temperature of the melting layer and store it in a database.
  • the observation data collection unit 110 may collect radar reflectivity data from the weather radar system 1 and detect a bright band in the radar reflectivity data.
  • the weather radar system 1 When using the weather radar system 1 to observe a layer in which the water body changes from ice particles to water particles, abnormally high radar reflectivity may appear.
  • the ice particles pass through the isothermal layer at 0 ° C., the surface of the particles gradually starts to melt, and the particles formed with the water film layer on the outside promote adhesion by adhesion, and exhibit high reflectivity by increasing the dielectric constant and size.
  • the volume decreases and the reflection decreases again. That is, the radar reflectivity data observed from the 0 ° C isothermal layer, that is, the melting layer, increases rapidly, and then decreases again.
  • the reflectance value increases from the point where the reflectance value increases rapidly, then decreases again, resulting in constant reflectivity. Up to the point representing the value can be defined as a bright band.
  • the observation data collection unit 110 can detect a bright band from the radar reflectivity data, and calculate the altitude and temperature of the melting layer using the bright band.
  • liquid particles exist below the altitude of the melting layer, which affects the amount of ground precipitation. Therefore, if the initial field of the numerical forecasting model is set by applying the altitude of the melting layer, it will be possible to more accurately estimate the above-ground precipitation.
  • FIG. 2 is a view for explaining a method of calculating the altitude and temperature of the fusion layer in the observation data collection unit shown in FIG. 1, and
  • FIG. 3 is a view for explaining characteristics of bright bands appearing in the radar reflectivity data.
  • the observation data collection unit 110 may generate an average reflectivity vertical profile using radar reflectivity data.
  • Observation data collection unit 110 is a logarithmic mean reflectivity (log (Z H )), primary differential (f '(h)), secondary differential (f''(h)) Curvature (C (h)), primary derivative of curvature (C '(h)), can generate a vertical profile.
  • the observation data collection unit 110 may detect the top, top, and bottom of the bright band using the five profiles required for the detection of the bright band.
  • the observation data collection unit 110 may detect the altitude of the maximum reflectance value in the reflectance data as the peak of the bright band (BB PEAK ).
  • the observation data collection unit 110 can detect the altitude of the point where the secondary derivative value is “+” and the primary derivative value is changed from “+” to “-” in the reflectivity data as the highest point of the bright band (BB TOP ). have.
  • the observation data collection unit 110 may detect the altitude of the point where the secondary derivative value is "-" in the reflectivity data and the primary derivative value is changed from “-" to "+” as the lowest point of the bright band (BB BOTTOM ). .
  • the observation data collection unit 110 may detect the altitude corresponding to the bright band region in the radar reflectivity data, and calculate it as the altitude of the melting layer. In addition, the observation data collection unit 110 may calculate the temperature of the elevation of the melting layer by comparing the elevation of the melting layer with the corresponding elevation in the Lewinsonde data.
  • the observation data collection unit 110 may process the altitude and temperature of the melting layer calculated from the radar reflectivity data in the form of a buffer and store it in a database.
  • the observation data pre-processing unit 130 may perform a physical process of the observation data pre-processing (OPS) step in the very short-term forecast model.
  • OPS observation data pre-processing
  • observation data stored in the database can be loaded and processed to be used for data assimilation.
  • the observation data pre-processing unit 130 may retrieve the elevation of the melting layer stored in the database, interpolate it into the grid of the numerical prediction model, and extract the background field variable of the numerical prediction model for spatial information corresponding to the elevation of the melting layer.
  • observation data pre-processing unit 130 may generate a vertical standard altitude array.
  • the observation data pre-processing unit 130 may generate a total of 24 vertical standard altitude arrays by adding four arrays vertically in the 750 hPa to 550 hPa sections from the 20 vertical standard altitude arrays of the 1000 hPa to 1 hPa sections. This is to apply the bright band top-to-bottom-bottom to different vertical standard altitude arrangements.
  • the observation data pre-processing unit 130 may also modify the name list of observation data variables (varobs), background field variables (cx), and background error variables (cxbgerr) to 24 vertical standard altitude arrays.
  • the observation data pre-processing unit 130 extracts observation data variables (varobs) by interpolating observation data of altitude and temperature of the melting layer into these vertical standard altitude arrays, and observes the observation data variables (varobs) in the background of the very short-term forecast model.
  • the same variable can be extracted as a background field variable (cx).
  • variable data assimilation unit 150 may perform a physical process of a 3D variable data assimilation (VAR) step in an early short-term forecast model.
  • VAR variable data assimilation
  • variable data assimilation (VAR) stage of the very short-term forecast model an initial field can be generated by assimilation of observation data for a specific time.
  • the three-dimensional variable data assimilation method has an advantage that it has less computational cost than other data assimilation methods such as four-dimensional variable data assimilation or ensemble data assimilation.
  • Equation 1 J means a cost function (penalty) that occurs when a control variable is assimilated in 3D variable data assimilation, y is observation information, x is model analysis field information, x 0 is model background field information, and B -1 is the model error covariance, R -1 is the observation error covariance, and H (x) is the observation operator that converts the observation data into model variables.
  • VAR 3D variable data assimilation
  • FIG. 4 is a diagram for explaining a 3D variable data assimilation method.
  • Obs is a variable of observation data
  • Cx is a model background field variable
  • Cw is a potentially usable linearization state (LS) variable
  • CxPlus represents a difference between observation data and a model background field, each of which is 3D. It is a variable used in the variable data assimilation method.
  • v_hat is the result of observation data
  • CxPlus_hat is the result of the difference between the observation data and the model background field
  • Cw_hat is the slope
  • ModelOb_hat is the result of the model background field for the observation data. It is the result.
  • the first step of the 3D variable data assimilation method may read setting information necessary when performing data assimilation.
  • Read configuration information for observation data from the Observation namelist and perform data assimilation such as Minimization namelist, Data assimilation control variable (Transform namelist) and Diagnostic variable (Diagnose namelist) You can prepare for system operation by reading the necessary configuration information.
  • the second and third steps of the 3D variable data assimilation method correspond to a preparation step for loading linear field information and a step for reading linear field information.
  • the 4th and 5th steps of the 3D variable data assimilation method set memory allocation, arrangement, etc. for the analysis of the deviation field (PF) through the difference between the model's first guess and the model's background field. This corresponds to the step of calculating the deviation field.
  • an array of four control variables (AP, mu, chi, and psi) can be constructed to apply the observed data to the numerical prediction model.
  • the seventh step of the 3D variable data assimilation method it is possible to construct an array of variables necessary for data assimilation for setting information for observation data, which is an observation species set in a name list.
  • the eighth step of the 3D variable data assimilation method corresponds to the process of minimization of the observed data.
  • the cost functions (J o , J b ) of the observation data and the model background field are calculated, and the total cost function (J) can be calculated.
  • an outer loop and an inner loop may be performed to find a gradient close to 1 between the observed data and the model background field.
  • Equation 2 is used to calculate the penalty of observation data
  • equations 3 and 4 are used to calculate the slope of the Cw and ModelOb variables, respectively.
  • the observation penalty for the observed data is calculated, and the slope between the observed data and the model background field can be calculated using this, and an optimal slope close to 1 is obtained.
  • the analysis increment between the initial field of the numerical prediction model and the initial field when the observation data is applied can be finally calculated by performing the minimization process to find it.
  • the variable data assimilation unit 150 may improve the temperature distribution of the initial field according to the very short-term forecast model by inputting a high temperature of the melting layer into the very short-term forecast model using the three-dimensional variable data assimilation method. That is, the variable data assimilation unit 150 may calculate an analysis increment by performing a 3D variable data assimilation between the observed data variables (varobs) and the background field variable (cx).
  • the observed data variables (varobs) correspond to the altitude and temperature of the melting layer.
  • variable data assimilation unit 150 may perform a process of minimizing through 10 external and internal repeat cycles as shown in Table 1 below.
  • the variable data assimilation unit 150 calculates the cost function (J) for the observed data and the model background field as shown in Equations 1 and 2 through the external and internal iterative cycles, and the cost function as shown in Equations 3 and 4
  • the slope of (J) can be calculated.
  • variable data assimilation unit 150 uses the 3D variable data assimilation method to calculate the slope value as a value representing the difference between the observed data and the model background field,
  • Analysis increments can be calculated between the initial field of the numerical forecast model before the observation data are applied and the initial field when the observation data are applied.
  • the initial field predicting unit 170 may generate an initial field of the very short-term forecast model using analytical increments, which are final outputs of 3D variable data assimilation.
  • the initial field prediction unit 170 may improve the initial dynamic field by applying an analysis increment, which is the final output of the 3D variable data assimilation, to the vertical temperature profile of the initial dynamic field.
  • the initial field prediction unit 170 may perform precipitation prediction based on the initial dynamic field generated by applying an analysis increment as described above.
  • the apparatus 100 according to the present invention can calculate the temperature of the melting layer altitude by detecting a bright band in the radar reflectivity data.
  • the spatial and temporal resolution is low, while in the case of radar reflectivity data, the spatial and temporal resolution is high, so it is possible to calculate the altitude and temperature of the melting layer more accurately.
  • the apparatus 100 according to the present invention inputs the temperature data of the molten layer altitude into the temperature profile of the initial dynamic field through a three-dimensional variable data assimilation method, so that the upper, middle and lower layers of the temperature field simulated in the numerical prediction model are obtained. Initial temperature information can be corrected.
  • the apparatus 100 according to the present invention can improve precipitation prediction performance by performing precipitation prediction using an improved initial dynamic field.
  • FIG. 5 is a flowchart of a method for assimilation of an advanced data of a radar fusion layer based on an ultra-short-term forecasting model according to an embodiment of the present invention.
  • the method for assimilation of the elevation data of the radar melting layer based on the ultra-short-term forecasting model may be performed in substantially the same configuration as the apparatus 100 illustrated in FIG. 1. Therefore, the same components as the device 100 of FIG. 1 are given the same reference numerals, and repeated descriptions will be omitted.
  • the observation data collection unit 110 may collect radar observation data to calculate the altitude and temperature of the melting layer (S10).
  • the observation data collection unit 110 may detect a bright band from the radar reflectivity data, and calculate the altitude and temperature of the melting layer using the bright band.
  • the observation data collection unit 110 may generate an average reflectivity vertical profile using radar reflectivity data.
  • Observation data collection unit 110 is a logarithmic mean reflectivity (log (Z H )), primary differential (f '(h)), secondary differential (f''(h)) Curvature (C (h)), primary derivative of curvature (C '(h)), can generate a vertical profile.
  • the observation data collection unit 110 may detect the top, top, and bottom of the bright band using the five profiles required for the detection of the bright band.
  • the observation data collection unit 110 can detect the altitude of the point where the secondary derivative value is “+” and the primary derivative value is changed from “+” to “-” in the reflectivity data as the highest point of the bright band (BB TOP ). have.
  • the observation data collection unit 110 may detect the altitude of the point where the secondary derivative value is "-" in the reflectivity data and the primary derivative value is changed from “-" to "+” as the lowest point of the bright band (BB BOTTOM ). .
  • the observation data collection unit 110 may detect the altitude corresponding to the bright band region in the radar reflectivity data, and calculate it as the altitude of the melting layer. In addition, the observation data collection unit 110 may calculate the temperature of the elevation of the melting layer by comparing the elevation of the melting layer with the corresponding elevation in the Lewinsonde data.
  • the observation data collection unit 110 may process the altitude and temperature of the melting layer calculated from the radar reflectivity data in the form of a buffer and store it in a database.
  • the observation data pre-processing unit 130 may process the altitude and temperature of the melting layer calculated by collecting radar observation data (S20).
  • the observation data pre-processing unit 130 may retrieve the elevation of the melting layer stored in the database, interpolate it into the grid of the numerical prediction model, and extract the background field variable of the numerical prediction model for spatial information corresponding to the elevation of the melting layer.
  • the observation data pre-processing unit 130 may generate a vertical standard altitude array.
  • the observation data pre-processing unit 130 may generate a total of 24 vertical standard altitude arrays by adding 4 arrays vertically in the 750 hPa to 550 hPa section from 20 vertical standard altitude arrays of 1000 hPa to 1 hPa sections. . This is to apply the bright band top-to-bottom-bottom to different vertical standard altitude arrangements.
  • the observation data pre-processing unit 130 may also modify the name list of observation data variables (varobs), background field variables (cx), and background error variables (cxbgerr) to 24 vertical standard altitude arrays.
  • the observation data pre-processing unit 130 extracts observation data variables (varobs) by interpolating observation data of altitude and temperature of the melting layer into these vertical standard altitude arrays, and observes the observation data variables (varobs) in the background of the very short-term forecast model.
  • the same variable can be extracted as a background field variable (cx).
  • the variable data assimilation unit 150 may calculate an analysis increment by applying a 3D variable data assimilation method between the altitude and temperature of the melting layer and the background field of the ultra-short-term forecast model (S30).
  • the variable data assimilation unit 150 calculates the cost function (J) for the observed data and the model background field as shown in Equations 1 and 2 through the external and internal iterative cycles, and the cost function as shown in Equations 3 and 4
  • the slope of (J) can be calculated.
  • variable data assimilation unit 150 uses the 3D variable data assimilation method as a value representing the difference between the observation data and the model background field, and the initial field and observation of the numerical prediction model before the observation data are applied. Analysis increments between initial fields when data are applied can be calculated.
  • the initial field prediction unit 170 may generate an initial field of the very short-term forecast model using the incremental analysis and use it to perform precipitation prediction (S40).
  • the initial field prediction unit 170 may improve the initial dynamic field by applying an analysis increment, which is the final output of the 3D variable data assimilation, to the vertical temperature profile of the initial dynamic field of the very short-term forecast model.
  • the initial field prediction unit 170 may perform precipitation prediction based on the initial dynamic field generated by applying an analysis increment as described above.
  • the method for assimilation of the advanced data of the radar fusion layer based on the ultra-short-term forecasting model of the present invention may be implemented as an application or implemented in the form of program instructions that can be executed through various computer components to be recorded in a computer-readable recording medium.
  • the computer-readable recording medium may include program instructions, data files, data structures, or the like alone or in combination.
  • the program instructions recorded on the computer-readable recording medium are specially designed and configured for the present invention, and may be known and available to those skilled in the computer software field.
  • Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs, DVDs, and magneto-optical media such as floptical disks. media), and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like.
  • Examples of program instructions include not only machine language codes produced by a compiler, but also high-level language codes executable by a computer using an interpreter or the like.
  • the hardware device may be configured to operate as one or more software modules to perform processing according to the present invention, and vice versa.
  • FIGS. 6A to 7D are graphs showing bright band altitudes calculated from radar observation data in a stratospheric case.
  • the average altitudes of the peak, peak, and bottom of the bright band from the Gwangdeoksan (GDK) radar observation data were calculated to be 4.22 km, 3.65 km, and 3.14 km, respectively, and the temperature of the peak, peak, and lowest altitude of the bright band. was calculated as -2.1 ° C, 0.8 ° C and 4.4 ° C, respectively.
  • the average altitudes of the peaks, peaks, and troughs of the bright band from the Oseong Mountain (KSN) radar observations were calculated to be 4.35 km, 3.70 km, and 3.19 km, respectively, and the temperatures of the peak, peak, and trough altitudes of the bright bands, respectively.
  • KSN Oseong Mountain
  • the average altitudes of the peak, peak, and bottom of the bright band from the intensity (JNI) radar observations were calculated to be 4.46 km, 3.82 km, and 3.33 km, respectively, and the temperature of the peak, peak, and lowest altitude of the bright band.
  • the temperature of the peak, peak, and lowest altitude of the bright band was calculated as -2.0 ° C, 2.4 ° C and 5.1 ° C, respectively.
  • the average altitudes of the peaks, peaks, and troughs of the bright bands were calculated to be 4.64 km, 4.08 km, and 3.53 km, respectively, and the temperatures of the peaks, peaks, and trough altitudes of the bright bands, respectively.
  • the average altitude of the melting layer of Osan's Lewinzonde observation point for comparison with the observation points of Gwangdeok Mountain (GDK) and Oseong Mountain (KSN) radar was 4.18 km.
  • the average elevation of the melting layer at the Lewinzonde Gwangju observation point was 4.09 km.
  • the average elevation of the melting layer of Jeju Lewinzonde observation point for comparison with the observation point of the radar observation point (GSN) was 3.84 km.
  • the average altitudes of the peaks, peaks, and troughs of the bright bands from the Gwangdeoksan (GDK) radar observations were calculated to be 4.22 km, 3.65 km, and 3.14 km, respectively, and the temperature of the peaks, peaks, and troughs altitudes of the bright bands, respectively.
  • the average altitudes of the peaks, peaks, and troughs of the bright bands from the Oseong Mountain (KSN) radar observations were calculated to be 4.35 km, 3.70 km, and 3.19 km, respectively, and the temperatures of the peaks, peaks, and troughs altitudes of the bright bands, respectively.
  • KSN Oseong Mountain
  • the average altitudes of the peak, peak, and bottom of the bright band from the intensity (JNI) radar observations were calculated to be 4.46 km, 3.82 km, and 3.33 km, respectively, and the temperature of the peak, peak, and lowest altitude of the bright band.
  • the temperature of the peak, peak, and lowest altitude of the bright band was calculated as -2.0 ° C, 2.4 ° C and 5.1 ° C, respectively.
  • the average altitudes of the peaks, peaks, and troughs of the bright bands were calculated to be 4.64 km, 4.08 km, and 3.53 km, respectively, and the temperatures of the peaks, peaks, and troughs altitudes of the bright bands, respectively.
  • the average altitude of the melting layer of Osan's Lewinzonde observation point for comparison with the observation points of Gwangdeok Mountain (GDK) and Oseong Mountain (KSN) radar was 4.18 km.
  • the average elevation of the melting layer at the Lewinzonde Gwangju observation point was 4.09 km.
  • the average elevation of the melting layer of Jeju Lewinzonde observation point for comparison with the observation point of the radar observation point (GSN) was 3.84 km.
  • 8A and 8B are graphs showing the vertical temperature profile of the initial dynamic field generated according to the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention in the rainy season case.
  • FIG. 8A The vertical temperature profile of the initial dynamic field generated as a result of data assimilation between the elevation and temperature of the fusion layer and model background field information from radar observation data of the rainy season case is shown in FIG. 8A and It is like FIG. 8B.
  • 9A to 12D are graphs showing changes in the dynamic field according to an assimilation method of an elevation data of a radar fusion layer based on the ultra-short-term forecast model of the present invention in the rainy season front-line case.
  • the air pressure in the western coastal region decreased by more than 0.06 hPa
  • the temperature increased by more than 0.1 ° C in the midwestern region including the metropolitan area, the area near Gwangju, the southern sea and Jeju Island decreased by more than 0.2 ° C.
  • Q vapor decreased in the metropolitan area and Gyeonggi-do, increased near Jeonnam and Jeju, and q rain increased in the soiled sea.
  • q graupel increased in the vicinity of Baeknyeongdo Island, and q cf decreased.
  • the air pressure increased by 0.06 hPa or more, and decreased by 0.08 hPa or more in the south sea, and the temperature increased by 0.06 ° C or more in the inland region, as opposed to the results in the lower atmosphere.
  • Q vapor increased in the vicinity of Baeknyeongdo Island, weakly decreased in some areas of the metropolitan area and the southern inland, and q cf decreased in the vicinity of Baeknyeongdo Island.
  • the lower layer As such, as a result of data assimilation of the temperature of the fusion layer altitude according to the radar melting layer altitude data assimilation method based on the ultra-short-term forecasting model of the present invention, the lower layer, the middle layer of the atmosphere and It can be confirmed that the upper dynamic field has changed.
  • the temperature was lowered and it became relatively cold, and in order to reduce the precipitation simulated by the rainy season electric line located in the southern region, the air pressure in the central region decreased and the air pressure in the southern region increased.
  • the q vapor in the Midwest increased, making the area wet.
  • the pressure change pattern is similar to the lower layer, and q vapor decreases in the central region and increases in the southern region, and temperature increases in the central region and decreases in the southern region. From this, it can be confirmed that the middle layer of the atmosphere became dry and the southern layer became humid. In the case of the upper layer, the temperature increased with the center of the inland area warming up, and showed the opposite pattern to the atmosphere and the lower layer. As a whole, it can be confirmed that the dynamic field changed to simulate the stabilization of the atmosphere as the lower layer of the atmosphere became colder and the upper layer warmed around the inland region.
  • 12A to 12D are graphs comparing the precipitation simulation result between the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention and the precipitation simulation result between the normative experiments.
  • precipitation simulation results (EXP) according to the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention and other precipitation simulation results (CTRL) in the normative experiment to which the data assimilation method is not applied
  • CTRL precipitation simulation results

Abstract

Disclosed are a method for assimilating radar melting layer elevation data on the basis of an ultra-short forecast model, and a recording medium and an apparatus for performing same. The method for assimilating radar melting layer elevation data on the basis of an ultra-short forecast model, comprises: a step of collecting radar observation data from a meteorological radar system to produce observation data including the altitude and the temperature of a melting layer; a step of preprocessing the observation data to be applicable for data assimilation; a step of calculating an analysis increment for the observation data by applying a three-dimensional variable data assimilation method between background fields of an ultra-short forecast model; and a step of generating an initial field of the ultra-short forecast model using the analysis increment, and predicting precipitation.

Description

초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법, 이를 수행하기 위한 기록 매체 및 장치Method for assimilation of advanced data of radar melting layer based on the very short-term forecasting model, recording medium and device for performing this
본 발명은 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법, 이를 수행하기 위한 기록 매체 및 장치에 관한 것으로서, 더욱 상세하게는 기상 레이더의 관측 자료를 이용하여 수치예보모델의 초기장을 개선하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법, 이를 수행하기 위한 기록매체 및 장치에 관한 것이다. The present invention relates to a method for assembling altitude data of a radar fusion layer based on a very short-term forecasting model, and a recording medium and apparatus for performing the same, more specifically, a very short-term to improve the initial field of a numerical forecasting model using observation data of a weather radar. It relates to a method of assimilation of altitude data in a radar melting layer based on a forecast model, and a recording medium and apparatus for performing the same.
일반적으로 기상 레이더는 전자기파를 발사하여 기상학적 목표물로부터 반사 또는 산란되어 오는 전파신호의 크기를 계산하는 장비로, 넓은 영역(유효관측반경: 240km)을 매우 빠르게(10분 간격) 감시할 수 있기 때문에 넓은 영역의 강수량을 산출하는 가장 효율적인 원격탐사장비 중 하나이다.In general, a weather radar is a device that calculates the size of a radio wave signal that is reflected or scattered from a meteorological target by firing electromagnetic waves, and can monitor a large area (effective observation radius: 240 km) very quickly (every 10 minutes). It is one of the most efficient remote sensing equipment for calculating a large amount of precipitation.
목표물로부터 반사 또는 산란되어 오는 전파신호인 반사도는 기상 레이더에서 발사되는 펄스 볼륨 내에 존재하는 물방울의 크기 분포와 관계가 있다. 또한, 지상에 떨어지는 강수량도 물방울의 크기 분포의 함수이므로 레이더 반사도와 지상 강수량 간의 일정 관계식(Z-R 관계식: Z=aRb)을 사용하여 레이더 반사도로부터 지상 강수량을 추정해낼 수 있다.The reflectivity, which is a radio wave signal reflected or scattered from the target, is related to the size distribution of water droplets present in the pulse volume emitted from the weather radar. In addition, since precipitation falling on the ground is also a function of the size distribution of water droplets, it is possible to estimate the ground precipitation from the radar reflectivity using a constant relation between the radar reflectivity and the ground precipitation (ZR relation: Z = aR b ).
기상 레이더에서 획득하는 레이더 반사도에 따르면, 0℃ 이하의 높은 고도에서 강설 입자들이 낙하하여 0℃ 등온층을 통과하면 강설 입자의 겉 표면이 녹기 시작하고 이 층에서 급격하게 반사도 값이 증가하는데, 이는 입자의 유전율의 증가 및 크기 효과 때문이다. 그리고, 고도가 낮아지면서 거의 모든 강설 입자들이 융해되어 물방울을 형성하게 되고 반사도 값이 다시 감소하게 되는데, 이와 같이 반사도 값이 급격히 증가한 후 다시 감소하여 일정한 반사도가 나타나기 전까지 지역을 밝은 띠(Bright Band) 지역이라 하며, 0℃ 등온층을 강설 입자가 융해되기 시작하는 고도, 즉, 융해층 고도라고 한다. According to the radar reflectivity obtained from a weather radar, when snowfall particles fall at a high altitude of 0 ° C or lower and pass through the isothermal layer at 0 ° C, the outer surface of the snowfall particle starts to melt and the reflectance value rapidly increases in this layer. This is due to the increased dielectric constant and size effect. In addition, as the altitude decreases, almost all snowfall particles melt to form water droplets, and the reflectance value decreases again. As such, the reflectance value increases rapidly and then decreases again, until a certain reflectivity appears, and the area is bright band (Bright Band). It is called an area, and the isothermal layer at 0 ℃ is called the altitude at which snowfall particles begin to melt, that is, the altitude of the melted layer.
이러한 융해층 고도 아래에서는 액체상의 입자가 존재하여 지상 강수량에 영향을 주게 된다. 따라서, 수치예보모델 초기장에서 융해층 고도를 잘못 분석하는 경우, 구름 내 또는 하부의 온도 편차가 발생하여 수상체 유형이 바뀌고 강수 오차가 발생하게 된다.Below the altitude of the melting layer, liquid particles are present, which affect the ground precipitation. Therefore, when the altitude of the melt layer is incorrectly analyzed in the initial field of the numerical prediction model, the temperature variation in or below the cloud occurs, changing the type of the water body and causing precipitation errors.
이에 레윈죤데, 윈드프로파일러와 같은 종관규모 관측자료를 이용하여 수치예보모델 초기장의 온도 프로파일을 보정해주는 자료 동화 기술이 제안되었으나, 종관규모의 관측자료의 경우, 기상 레이더의 관측자료에 비해 시공간적으로 해상도가 낮아 융해층 고도의 정확한 분석이 어려워 강수 오차의 발생 확률이 여전히 존재한다.Therefore, data assimilation technology has been proposed to correct the temperature profile of the initial field of the numerical forecasting model by using observation data such as Lewynsonde and Wind Profiler. Due to the low resolution, it is difficult to accurately analyze the altitude of the melted layer, so the probability of precipitation error still exists.
본 발명의 일측면은 기상 레이더에서 획득하는 반사도 자료로부터 융해층 고도 및 온도를 산출하고, 융해층 고도 및 온도를 자료 동화에 적용할 수 있도록 전처리하여 초단기예보모델의 초기장을 개선하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법, 이를 수행하기 위한 기록매체 및 장치를 제공한다.One aspect of the present invention is an ultra-short-term forecasting model that calculates the altitude and temperature of the melted layer from reflectivity data obtained from a weather radar, and preprocesses the altitude and temperature of the melted layer to be applied to data assimilation to improve the initial field of the ultra-short-term forecasting model. Provided is a method for assimilation of advanced data based on a radar fusion layer, and a recording medium and apparatus for performing the same.
상기 과제를 해결하기 위한 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법은, 기상 레이더 시스템으로부터 레이더 관측 자료를 수집하여 융해층의 고도 및 온도를 포함하는 관측 자료를 산출하는 단계, 상기 관측 자료를 자료 동화에 적용할 수 있도록 전처리하는 단계, 상기 관측 자료를 초단기예보모델의 배경장 간에 3차원 변분 자료 동화 방식을 적용하여 분석 증분을 산출하는 단계 및 상기 분석 증분을 이용하여 초단기예보모델의 초기장을 생성하고 강수를 예측하는 단계를 포함한다.In order to solve the above problems, the method of assimilation of the elevation data of the radar melting layer based on the ultra-short-term forecasting model comprises collecting radar observation data from a weather radar system and calculating observation data including altitude and temperature of the melting layer. Pre-processing to be applicable to data assimilation, calculating analytic increments by applying the 3D variable data assimilation method between the background fields of the ultra-short-term forecast model and the initial field of the ultra-short-term forecast model using the analysis increments And generating and predicting precipitation.
한편, 기상 레이더 시스템으로부터 레이더 관측 자료를 수집하여 융해층의 고도 및 온도를 포함하는 관측 자료를 산출하는 단계는, 상기 레이더 관측 자료에서 밝은 띠를 탐지하는 단계, 상기 밝은 띠를 상기 융해층의 고도로 산출하는 단계 및 상기 융해층의 고도와 레윈죤데 자료를 비교하여 상기 융해층 온도를 산출하는 단계를 포함할 수 있다.On the other hand, collecting the radar observation data from the weather radar system and calculating the observation data including the altitude and temperature of the melting layer includes: detecting a bright band from the radar observation data, and converting the bright band into the elevation of the melting layer It may include a step of calculating and comparing the altitude of the fusion layer with the Lewinsonde data to calculate the temperature of the fusion layer.
또한, 기상 레이더 시스템으로부터 레이더 관측 자료를 수집하여 융해층의 고도 및 온도를 포함하는 관측 자료를 산출하는 단계는, 상기 관측 자료를 버퍼 형태로 처리하여 데이터베이스에 저장하는 단계를 포함할 수 있다.In addition, collecting radar observation data from the weather radar system and calculating observation data including altitude and temperature of the melting layer may include processing the observation data in a buffer form and storing the observation data in a database.
또한, 상기 관측 자료를 자료 동화에 적용할 수 있도록 전처리하는 단계는, 상기 관측 자료에 포함되는 상기 융해층의 고도를 수치예보모델의 격자에 내삽하는 단계 및 상기 융해층의 고도에 해당하는 공간정보에 대한 상기 수치예보모델의 배경장 변수를 추출하는 단계를 포함할 수 있다.In addition, the pre-processing so that the observation data can be applied to data assimilation includes interpolating the altitude of the melting layer included in the observation data into a grid of a numerical prediction model and spatial information corresponding to the elevation of the melting layer. It may include the step of extracting the background field variable of the numerical prediction model for.
또한, 상기 관측 자료를 자료 동화에 적용할 수 있도록 전처리하는 단계는, 1000hPa 내지 1hPa 구간의 20 개의 연직 표준 고도 배열에서 750hPa 내지 550hPa 구간에 연직으로 4 개의 배열을 추가하여 총 24 개의 연직 표준 고도 배열을 생성하는 단계, 상기 연직 표준 고도 배열에 상기 융해층의 고도를 내삽하여 관측 자료 변수를 추출하는 단계 및 상기 초단기예보모델의 배경장에서 상기 관측 자료 변수와 동일한 변수를 상기 배경장 변수로 추출하는 단계를 포함할 수 있다.In addition, the step of pre-processing the observation data to be applied to the data assimilation is a total of 24 vertical standard altitude arrays by adding four arrays vertically from 20 vertical standard altitude arrays of 1000 hPa to 1 hPa sections to 750 hPa to 550 hPa sections. Generating a step, interpolating the elevation of the melting layer in the vertical standard altitude arrangement to extract observation data variables and extracting the same variables as the observation data variables from the background field of the very short-term forecast model as the background field variables It may include steps.
또한, 상기 관측 자료를 초단기예보모델의 배경장 간에 3차원 변분 자료 동화 방식을 적용하여 분석 증분을 산출하는 단계는, 외부반복순환과 내부반복순환을 통해 수학식 상기 관측 자료 및 상기 초단기예보모델의 배경장에 대한 비용함수 및 기울기를 산출하는 단계 및 상기 기울기 값을 이용하여 상기 관측 자료가 적용되기 전의 상기 초단기예보모델의 초기장과 상기 관측 자료가 적용되었을 때의 상기 초단기예보모델의 초기장 간의 분석 증분을 산출하는 단계를 포함할 수 있다.In addition, the step of calculating an analysis increment by applying a three-dimensional variable data assimilation method between the background fields of the very short-term forecast model is the equation of the observed data and the early short-term forecast model through an external iterative cycle and an internal iterative cycle. Calculating the cost function and slope for the background field, and using the slope value between the initial field of the early short-term forecast model before the observation data is applied and the initial field of the early short-term forecast model when the observed data is applied. And calculating an analytical increment.
또한, 상기 분석 증분을 이용하여 초단기예보모델의 초기장을 생성하고 강수를 예측하는 단계는, 상기 분석 증분을 상기 초단기예보모델의 초기장의 연직 온도 프로파일에 적용하는 단계를 포함할 수 있다.In addition, the step of generating an initial field of an early short-term forecast model and predicting precipitation using the analysis increment may include applying the analysis increment to a vertical temperature profile of the initial field of the early short-term forecast model.
또한, 상기 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법을 수행하기 위한, 컴퓨터 프로그램이 기록된 컴퓨터로 판독 가능한 기록 매체일 수 있다.In addition, the computer may be a computer-readable recording medium in which a computer program is recorded for performing the method for assimilation of the advanced data for the radar fusion layer based on the ultra-short-term forecast model.
한편, 상기 과제를 해결하기 위한 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치는, 기상 레이더 시스템으로부터 레이더 관측 자료를 수집하여 융해층의 고도 및 온도를 포함하는 관측 자료를 산출하는 관측 자료 수집부, 상기 관측 자료를 자료 동화에 적용할 수 있도록 전처리하는 관측 자료 전처리부, 상기 관측 자료를 초단기예보모델의 배경장 간에 3차원 변분 자료 동화 방식을 적용하여 분석 증분을 산출하는 변분 자료 동화부 및 상기 분석 증분을 이용하여 초단기예보모델의 초기장을 생성하고 강수를 예측하는 초기장 예측부를 포함한다.On the other hand, the radar fusion layer elevation data assimilation device based on the ultra-short-term forecast model for solving the above problems is an observation data collection unit that collects radar observation data from a weather radar system and calculates observation data including altitude and temperature of the fusion layer. , An observation data pre-processing unit that pre-processes the observation data to be applied to the data assimilation, a variable data assimilation unit that calculates an analysis increment by applying a three-dimensional variable data assimilation method between the background fields of the very short-term forecast model, and the It includes an initial field prediction unit that generates an initial field of the ultra-short-term forecast model using the incremental analysis and predicts precipitation.
한편, 상기 관측 자료 수집부는, 상기 레이더 관측 자료에서 밝은 띠를 탐지하고, 상기 밝은 띠를 상기 융해층의 고도로 산출하며, 상기 융해층의 고도와 레윈죤데 자료를 비교하여 상기 융해층 온도를 산출할 수 있다.Meanwhile, the observation data collection unit detects a bright band from the radar observation data, calculates the bright band at the altitude of the fusion layer, and compares the altitude of the fusion layer with the Lewinsonde data to calculate the melting layer temperature. Can be.
또한, 상기 관측 자료 수집부는, 상기 관측 자료를 버퍼 형태로 처리하여 데이터베이스에 저장할 수 있다.In addition, the observation data collection unit may process the observation data in a buffer form and store it in a database.
또한, 상기 관측 자료 전처리부는, 상기 관측 자료에 포함되는 상기 융해층의 고도를 수치예보모델의 격자에 내삽하고, 상기 융해층의 고도에 해당하는 공간정보에 대한 상기 수치예보모델의 배경장 변수를 추출할 수 있다.In addition, the observation data pre-processing unit interpolates the altitude of the melting layer included in the observation data into the grid of the numerical prediction model, and sets the background field variable of the numerical prediction model for spatial information corresponding to the elevation of the melting layer. Can be extracted.
또한, 상기 관측 자료 전처리부는, 1000hPa 내지 1hPa 구간의 20 개의 연직 표준 고도 배열에서 750hPa 내지 550hPa 구간에 연직으로 4 개의 배열을 추가하여 총 24 개의 연직 표준 고도 배열을 생성하고, 상기 연직 표준 고도 배열에 상기 융해층의 고도를 내삽하여 관측 자료 변수를 추출하며, 상기 초단기예보모델의 배경장에서 상기 관측 자료 변수와 동일한 변수를 상기 배경장 변수로 추출할 수 있다.In addition, the observation data pre-processing unit adds four arrays vertically in the range of 20 vertical standard altitudes in the range of 1000 hPa to 1 hPa in the range of 750 hPa to 550 hPa to generate a total of 24 vertical standard altitude arrays, and in the vertical standard altitude array By interpolating the altitude of the melting layer, an observation data variable is extracted, and the same variable as the observation data variable can be extracted from the background field of the very short-term forecast model as the background field variable.
또한, 상기 변분 자료 동화부는, 외부반복순환과 내부반복순환을 통해 수학식 상기 관측 자료 및 상기 초단기예보모델의 배경장에 대한 비용함수 및 기울기를 산출하고, 상기 기울기 값을 이용하여 상기 관측 자료가 적용되기 전의 상기 초단기예보모델의 초기장과 상기 관측 자료가 적용되었을 때의 상기 초단기예보모델의 초기장 간의 분석 증분을 산출할 수 있다.In addition, the variable data assimilation unit calculates a cost function and a slope for the background of the observation data and the very short-term forecast model by using an external iterative cycle and an internal iterative cycle. An analysis increment may be calculated between the initial field of the early short-term forecast model before application and the initial field of the early short-term forecast model when the observation data is applied.
또한, 상기 초기장 예측부는, 상기 분석 증분을 상기 초단기예보모델의 초기장의 연직 온도 프로파일에 적용할 수 있다.In addition, the initial field prediction unit may apply the analysis increment to the vertical temperature profile of the initial field of the ultra-short forecast model.
상술한 본 발명에 따르면 레이더 반사도 자료에서 밝은 띠를 탐지하여 융해층 고도의 온도를 산출하는데. 종관 관측자료의 경우 시공간적으로 해상도가 낮은 반면 레이더 반사도 자료의 경우 시공간적 해상도가 높으므로 보다 정확한 융해층 고도 및 온도의 산출이 가능하다. According to the present invention described above, by detecting a bright band in the radar reflectivity data to calculate the temperature of the melting layer altitude. In the case of synoptic observation data, the spatial and temporal resolution is low, while in the case of radar reflectivity data, the spatial and temporal resolution is high, so it is possible to calculate the altitude and temperature of the melting layer more accurately.
또한, 레이더 반사도 자료로부터 산출한 융해층 고도의 온도 자료를 3차원 변분 자료 동화 방식을 통해 초기 역학장의 온도 프로파일에 입력함으로써, 수치예보모델에서 모의한 온도장의 상층, 중층 및 하층의 초기 온도 정보를 보정할 수 있다. In addition, by inputting the temperature data of the melting layer altitude calculated from the radar reflectivity data into the temperature profile of the initial dynamic field through a 3D variable data assimilation method, the initial temperature information of the upper, middle and lower layers of the temperature field simulated in the numerical prediction model is obtained. Can be corrected.
또한, 개선된 초기 역학장을 이용한 강수 예측을 수행하여 강수 예측 성능을 향상시킬 수 있다.In addition, it is possible to improve precipitation prediction performance by performing precipitation prediction using an improved initial dynamic field.
도 1은 본 발명의 일 실시예에 따른 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치의 블록도이다.1 is a block diagram of a radar fusion layer advanced data assimilation device based on an ultra-short-term forecast model according to an embodiment of the present invention.
도 2는 도 1에 도시된 관측 자료 수집부에서 융해층의 고도 및 온도를 산출하는 방법을 설명하기 위한 도면이다.2 is a view for explaining a method of calculating the altitude and temperature of the melting layer in the observation data collection unit shown in FIG. 1.
도 3은 레이더 반사도 자료에 나타나는 밝은 띠의 특징을 설명하기 위한 도면이다.3 is a view for explaining the characteristics of the bright band appearing in the radar reflectivity data.
도 4는 3차원 변분 자료 동화 방식을 설명하기 위한 도면이다.4 is a diagram for explaining a 3D variable data assimilation method.
도 5는 본 발명의 일 실시예에 따른 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법의 순서도이다.5 is a flowchart of a method for assimilation of an advanced data of a radar fusion layer based on an ultra-short-term forecast model according to an embodiment of the present invention.
도 6a 내지 도 7d는 층운형 사례에서 레이더 관측 자료로부터 산출한 밝은 띠 고도를 나타내는 그래프이다.6A to 7D are graphs showing bright band altitudes calculated from radar observation data in a case of stratified clouds.
도 8a 및 도 8b는 장마전선 사례에서 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따라 생성한 초기 역학장의 연직 온도 프로파일을 나타내는 그래프이다.8A and 8B are graphs showing the vertical temperature profile of the initial dynamic field generated according to the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention in the rainy season case.
도 9a 내지 도 11e는 장마전선 사례에서 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 역학장 변화를 나타낸 그래프이다.9A to 11E are graphs showing changes in dynamic field according to an assimilation method of the radar melting layer elevation data based on the ultra-short-term forecast model of the present invention in the rainy season front-line case.
도 12a 내지 도 12d는 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 강수 모의 결과와 규준 실험 간 강수 모의 결과를 비교한 그래프이다.12A to 12D are graphs comparing the precipitation simulation result between the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention and the precipitation simulation result between the normative experiments.
후술하는 본 발명에 대한 상세한 설명은, 본 발명이 실시될 수 있는 특정 실시예를 예시로서 도시하는 첨부 도면을 참조한다. 이들 실시예는 당업자가 본 발명을 실시할 수 있기에 충분하도록 상세히 설명된다. 본 발명의 다양한 실시예는 서로 다르지만 상호 배타적일 필요는 없음이 이해되어야 한다. 예를 들어, 여기에 기재되어 있는 특정 형상, 구조 및 특성은 일 실시예와 관련하여 본 발명의 정신 및 범위를 벗어나지 않으면서 다른 실시예로 구현될 수 있다. 또한, 각각의 개시된 실시예 내의 개별 구성요소의 위치 또는 배치는 본 발명의 정신 및 범위를 벗어나지 않으면서 변경될 수 있음이 이해되어야 한다. 따라서, 후술하는 상세한 설명은 한정적인 의미로서 취하려는 것이 아니며, 본 발명의 범위는, 적절하게 설명된다면, 그 청구항들이 주장하는 것과 균등한 모든 범위와 더불어 첨부된 청구항에 의해서만 한정된다. 도면에서 유사한 참조부호는 여러 측면에 걸쳐서 동일하거나 유사한 기능을 지칭한다.For a detailed description of the present invention, which will be described later, reference is made to the accompanying drawings that illustrate, by way of example, specific embodiments in which the invention may be practiced. These examples are described in detail enough to enable those skilled in the art to practice the present invention. It should be understood that the various embodiments of the present invention are different, but need not be mutually exclusive. For example, the specific shapes, structures, and properties described herein can be implemented in other embodiments without departing from the spirit and scope of the invention in connection with one embodiment. In addition, it should be understood that the location or placement of individual components within each disclosed embodiment can be changed without departing from the spirit and scope of the invention. Therefore, the following detailed description is not intended to be taken in a limiting sense, and the scope of the present invention, if appropriately described, is limited only by the appended claims, along with all ranges equivalent to those claimed. In the drawings, similar reference numerals refer to the same or similar functions throughout several aspects.
이하, 도면들을 참조하여 본 발명의 바람직한 실시예들을 보다 상세하게 설명하기로 한다.Hereinafter, preferred embodiments of the present invention will be described in more detail with reference to the drawings.
도 1은 본 발명의 일 실시예에 따른 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치의 블록도이다.1 is a block diagram of a radar fusion layer advanced data assimilation device based on an ultra-short-term forecast model according to an embodiment of the present invention.
본 발명의 일 실시예에 따른 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치(100, 이하 장치)는 기상레이더 시스템(1)의 관측 자료를 이용하여 초단기예보모델의 초기장을 개선하는 장치로서, 기상레이더 시스템(1)과 유선 또는 무선 자원으로 연결되거나, 기상레이더 시스템(1)을 지칭하거나, 또는, 기상레이더 시스템(1)의 일부 또는 전부의 기능을 포함할 수 있다.The radar fusion layer elevation data assimilation device based on the ultra-short-term forecasting model according to an embodiment of the present invention (100, below) is an apparatus for improving the initial field of the ultra-short-term forecasting model using the observation data of the weather radar system 1 , It is connected to the weather radar system 1 by wired or wireless resources, refers to the weather radar system 1, or may include some or all of the functions of the weather radar system 1.
초단기예보모델(VDAPS: Very Short time range Data Assimilation and Prediction System)은 기상청의 수치예보모델로 크게 관측 자료 데이터베이스(ODB: Observation DataBase), 관측 자료 전처리(OPS: Observation Processing System), 3차원 변분 자료 동화(VAR: three dimensional VARiational data assimilation system) 및 예측(forecast) 단계로 구분될 수 있다.The Very Short Time Range Data Assimilation and Prediction System (VDAPS) is a numerical forecasting model of the Korea Meteorological Agency, which largely observes the Observation DataBase (ODB), the Observation Processing System (OPS), and assimilates three-dimensional variable data. (VAR: three dimensional VARiational data assimilation system) and the prediction (forecast) stage.
본 발명의 장치(100)는 기상레이더 시스템(1)의 관측 자료를 이용하여 융해층의 고도 및 온도 정보를 산출하고, 융해층의 고도 및 온도 정보를 초단기예보모델에 적용하여 전처리 및 자료 동화 단계를 수행함으로써 초단기예보모델의 초기 역학장을 보정할 수 있다.The apparatus 100 of the present invention calculates altitude and temperature information of the melted layer using the observation data of the weather radar system 1, and applies the altitude and temperature information of the melted layer to an ultra-short-term forecast model to preprocess and assimilate data. The initial dynamic field of the very short-term forecast model can be corrected by performing.
도 1을 참조하면, 본 발명의 장치(100)는 관측 자료 수집부(110), 관측 자료 전처리부(130), 변분 자료 동화부(150) 및 초기장 예측부(170)를 포함할 수 있다. 도 1에 도시된 관측 자료 수집부(110), 관측 자료 전처리부(130), 변분 자료 동화부(150) 및 초기장 예측부(170)의 구성은 통합 모듈로 형성되거나, 하나 이상의 모듈로 이루어질 수 있다. 그러나, 이와 반대로 각 구성은 별도의 모듈로 이루어질 수도 있다.Referring to FIG. 1, the apparatus 100 of the present invention may include an observation data collection unit 110, an observation data pre-processing unit 130, a variable data assimilation unit 150, and an initial field prediction unit 170. . The configuration of the observation data collection unit 110, the observation data pre-processing unit 130, the variable data assimilation unit 150, and the initial field prediction unit 170 shown in FIG. 1 is formed of an integrated module, or composed of one or more modules Can be. However, on the contrary, each configuration may be made of a separate module.
본 발명에 따른 장치(100)는 이동성을 갖거나 고정될 수 있다. 본 발명에 따른 장치(100)는 서버(server) 또는 엔진(engine) 형태일 수 있으며, 디바이스(device), 기구(apparatus), 단말(terminal), UE(user equipment), MS(mobile station), 무선기기(wireless device), 휴대기기(handheld device) 등 다른 용어로 불릴 수 있다.The device 100 according to the present invention may have mobility or be fixed. The device 100 according to the present invention may be in the form of a server or an engine, a device, an apparatus, a terminal, a user equipment (UE), a mobile station (MS), It can also be called another term, such as a wireless device or a handheld device.
본 발명에 따른 장치(100)는 운영체제(Operation System; OS), 즉 시스템을 기반으로 다양한 소프트웨어를 실행하거나 제작할 수 있다. 운영체제는 소프트웨어가 장치의 하드웨어를 사용할 수 있도록 하기 위한 시스템 프로그램으로서, 안드로이드 OS, iOS, 윈도우 모바일 OS, 바다 OS, 심비안 OS, 블랙베리 OS 등 모바일 컴퓨터 운영체제 및 윈도우 계열, 리눅스 계열, 유닉스 계열, MAC, AIX, HP-UX 등 컴퓨터 운영체제를 모두 포함할 수 있다.The device 100 according to the present invention may execute or manufacture various software based on an operating system (OS), that is, a system. The operating system is a system program to enable the software to use the hardware of the device, such as Android OS, iOS, Windows Mobile OS, Sea OS, Symbian OS, and BlackBerry OS Mobile computer operating system and Windows, Linux, Unix, MAC , AIX, HP-UX, and computer operating systems.
본 발명에 따른 장치(100)는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화를 위한 소프트웨어(애플리케이션)가 설치되어 실행될 수 있으며, 관측 자료 수집부(110), 관측 자료 전처리부(130), 변분 자료 동화부(150) 및 초기장 예측부(170)의 구성은 장치(100)에서 실행되는 소프트웨어에 의해 제어될 수 있다.The apparatus 100 according to the present invention may be installed and executed software (application) for assembling high-level radar melting layer data based on the ultra-short-term forecast model, observation data collection unit 110, observation data pre-processing unit 130, variation The configuration of the data assimilation unit 150 and the initial field prediction unit 170 may be controlled by software executed in the device 100.
이하, 도 1에 도시된 본 발명에 따른 장치(100)의 각 구성에서의 레이더 융해층 고도 자료 동화 방법에 대하여 자세히 설명한다.Hereinafter, a method for assimilation of the radar melting layer elevation data in each configuration of the apparatus 100 according to the present invention shown in FIG. 1 will be described in detail.
관측 자료 수집부(110)는 초단기예보모델에서 관측 자료 데이터베이스(ODB) 단계의 물리 과정을 수행할 수 있다.The observation data collection unit 110 may perform a physical process of the observation data database (ODB) step in the very short-term forecast model.
초단기예보모델의 관측 자료 데이터베이스(ODB) 단계에서는 종관규모(예를 들면, AWS, 해상부이, 윈드프로파일러 및 레윈존데 등)의 관측 자료를 처리할 수 있다. 관측 자료 데이터베이스(ODB) 단계에서는 온도, 상대습도, 비습 및 바람정보 등의 관측 자료를 전세계 공용 기상자료 포맷인 버퍼(BUFR: Binary Universal Form for the Representation of meteorological data) 형태로 저장할 수 있다.  In the Observation Data Database (ODB) phase of the early short-term forecasting model, observation data from the synoptic scale (eg, AWS, maritime buoy, wind profiler and Lewinsonde) can be processed. In the Observation Data Database (ODB) stage, observation data such as temperature, relative humidity, rain humidity, and wind information can be stored in the form of a buffer (BUFR: Binary Universal Form for the Representation of meteorological data), which is a global common weather data format.
관측 자료 수집부(110)는 기상레이더 시스템(1)에서 획득하는 레이더 관측 자료를 수집하여 융해층 고도 및 온도를 산출하고, 이를 데이터베이스에 저장할 수 있다.The observation data collection unit 110 may collect radar observation data obtained from the weather radar system 1 to calculate the altitude and temperature of the melting layer and store it in a database.
구체적으로는, 관측 자료 수집부(110)는 기상레이더 시스템(1)으로부터 레이더 반사도 자료를 수집하고, 레이더 반사도 자료에서 밝은 띠(Bright Band)를 탐지할 수 있다. Specifically, the observation data collection unit 110 may collect radar reflectivity data from the weather radar system 1 and detect a bright band in the radar reflectivity data.
기상레이더 시스템(1)을 이용하여 수상체가 얼음 입자에서 물 입자로 변하는 층을 관측하는 경우, 비정상적으로 높은 레이더 반사도가 나타날 수 있다. 얼음 입자는 0℃ 등온층을 통과하면서 입자의 표면이 서서히 녹기 시작하고, 이처럼 외부에 수막층이 형성된 입자는 부착에 의한 성정을 촉진시키며, 유전율 및 크기의 증가에 의해 높은 반사도를 나타낸다. 또한, 얼음 입자가 완전한 물 입자로 변하게 되면 부피가 감소하여 반사도 또한 다시 줄어들게 된다. 즉, 0℃ 등온층, 즉, 융해층으로부터 관측되는 레이더 반사도 자료는 급격히 증가한 후 다시 감소하는 양상을 보이며, 이처럼 레이더 반사도 자료에서 반사도 값이 급격히 증가하는 지점으로부터 반사도 값이 증가하다가 다시 감소하여 일정한 반사도 값을 나타내는 지점까지를 밝은 띠로 정의할 수 있다.When using the weather radar system 1 to observe a layer in which the water body changes from ice particles to water particles, abnormally high radar reflectivity may appear. As the ice particles pass through the isothermal layer at 0 ° C., the surface of the particles gradually starts to melt, and the particles formed with the water film layer on the outside promote adhesion by adhesion, and exhibit high reflectivity by increasing the dielectric constant and size. In addition, when the ice particles turn into complete water particles, the volume decreases and the reflection decreases again. That is, the radar reflectivity data observed from the 0 ° C isothermal layer, that is, the melting layer, increases rapidly, and then decreases again. As such, in the radar reflectivity data, the reflectance value increases from the point where the reflectance value increases rapidly, then decreases again, resulting in constant reflectivity. Up to the point representing the value can be defined as a bright band.
따라서, 관측 자료 수집부(110)는 레이더 반사도 자료에서 밝은 띠를 탐지하고, 밝은 띠를 이용하여 융해층의 고도 및 온도를 산출할 수 있다. 여기서, 융해층의 고도 아래에서는 액체상의 입자가 존재하여 지상 강수량에 영향을 주게 된다. 따라서, 융해층의 고도를 적용하여 수치예보모델의 초기장을 설정하는 경우 지상 강수량을 보다 정확하게 추정할 수 있을 것이다.Therefore, the observation data collection unit 110 can detect a bright band from the radar reflectivity data, and calculate the altitude and temperature of the melting layer using the bright band. Here, liquid particles exist below the altitude of the melting layer, which affects the amount of ground precipitation. Therefore, if the initial field of the numerical forecasting model is set by applying the altitude of the melting layer, it will be possible to more accurately estimate the above-ground precipitation.
도 2는 도 1에 도시된 관측 자료 수집부에서 융해층의 고도 및 온도를 산출하는 방법을 설명하기 위한 도면이고, 도 3은 레이더 반사도 자료에 나타나는 밝은 띠의 특징을 설명하기 위한 도면이다.FIG. 2 is a view for explaining a method of calculating the altitude and temperature of the fusion layer in the observation data collection unit shown in FIG. 1, and FIG. 3 is a view for explaining characteristics of bright bands appearing in the radar reflectivity data.
도 2를 참조하면, 관측 자료 수집부(110)는 레이더 반사도 자료를 이용하여 평균 반사도 연직 프로파일을 생성할 수 있다. Referring to FIG. 2, the observation data collection unit 110 may generate an average reflectivity vertical profile using radar reflectivity data.
관측 자료 수집부(110)는 밝은 띠 탐지에 필요한 5 개의 변수인 로그 취한 평균 반사도(log(ZH)), 일차 미분(f'(h)), 이차 미분(f''(h)), 곡률(C(h)), 곡률의 일차 미분(C'(h)) 연직 프로파일을 생성할 수 있다.Observation data collection unit 110 is a logarithmic mean reflectivity (log (Z H )), primary differential (f '(h)), secondary differential (f''(h)) Curvature (C (h)), primary derivative of curvature (C '(h)), can generate a vertical profile.
관측 자료 수집부(110)는 위와 같은 밝은 띠 탐지에 필요한 5 개의 프로파일을 이용하여 밝은 띠의 상단부, 최정점 및 하단부를 탐지할 수 있다.The observation data collection unit 110 may detect the top, top, and bottom of the bright band using the five profiles required for the detection of the bright band.
도 3을 참조하면, 관측 자료 수집부(110)는 반사도 자료에서 반사도 값이 최대인 고도를 밝은 띠의 최정점(BBPEAK)으로 탐지할 수 있다. Referring to FIG. 3, the observation data collection unit 110 may detect the altitude of the maximum reflectance value in the reflectance data as the peak of the bright band (BB PEAK ).
관측 자료 수집부(110)는 반사도 자료에서 이차 미분 값이 "+"이고, 일차 미분 값이 "+"에서 "-"로 변하는 지점의 고도를 밝은 띠의 최상점(BBTOP)으로 탐지할 수 있다.The observation data collection unit 110 can detect the altitude of the point where the secondary derivative value is “+” and the primary derivative value is changed from “+” to “-” in the reflectivity data as the highest point of the bright band (BB TOP ). have.
관측 자료 수집부(110)는 반사도 자료에서 이차 미분 값이 "-" 이고, 일차 미분 값이 "-"에서 "+"로 변하는 지점의 고도를 밝은 띠의 최하점(BBBOTTOM)으로 탐지할 수 있다.The observation data collection unit 110 may detect the altitude of the point where the secondary derivative value is "-" in the reflectivity data and the primary derivative value is changed from "-" to "+" as the lowest point of the bright band (BB BOTTOM ). .
이처럼, 관측 자료 수집부(110)는 레이더 반사도 자료에서 밝은 띠 영역에 해당하는 고도를 탐지할 수 있으며, 이를 융해층의 고도로 산출할 수 있다. 또한, 관측 자료 수집부(110)는 융해층의 고도와 레윈죤데 자료에서의 해당 고도를 비교하여 융해층 고도의 온도를 산출할 수 있다. As such, the observation data collection unit 110 may detect the altitude corresponding to the bright band region in the radar reflectivity data, and calculate it as the altitude of the melting layer. In addition, the observation data collection unit 110 may calculate the temperature of the elevation of the melting layer by comparing the elevation of the melting layer with the corresponding elevation in the Lewinsonde data.
관측 자료 수집부(110)는 위와 같이 레이더 반사도 자료로부터 산출한 융해층의 고도 및 온도를 버퍼 형태로 처리하여 데이터베이스에 저장할 수 있다.The observation data collection unit 110 may process the altitude and temperature of the melting layer calculated from the radar reflectivity data in the form of a buffer and store it in a database.
관측 자료 전처리부(130)는 초단기예보모델에서 관측 자료 전처리(OPS) 단계의 물리 과정을 수행할 수 있다. The observation data pre-processing unit 130 may perform a physical process of the observation data pre-processing (OPS) step in the very short-term forecast model.
초단기예보모델의 관측 자료 전처리(OPS) 단계에서는 데이터베이스에 저장된 관측 자료를 불러들여 자료 동화에 사용할 수 있도록 가공할 수 있다.In the pre-processing of observation data (OPS) of the very short-term forecast model, observation data stored in the database can be loaded and processed to be used for data assimilation.
관측 자료 전처리부(130)는 데이터베이스에 저장된 융해층의 고도를 불러들여 수치예보모델의 격자에 내삽하고, 융해층의 고도에 해당하는 공간정보 대한 수치예보모델의 배경장 변수를 추출할 수 있다.The observation data pre-processing unit 130 may retrieve the elevation of the melting layer stored in the database, interpolate it into the grid of the numerical prediction model, and extract the background field variable of the numerical prediction model for spatial information corresponding to the elevation of the melting layer.
구체적으로는, 관측 자료 전처리부(130)는 연직 표준 고도 배열을 생성할 수 있다. Specifically, the observation data pre-processing unit 130 may generate a vertical standard altitude array.
초단기예보모델의 OpsMod_Sonde.f90 모듈에서는 레윈존데 관측 자료를 처리하는데, 이때, 1000hPa 내지 1hPa 구간 내에서 20 개의 연직 표준 고도 배열을 적용한다.In the OpsMod_Sonde.f90 module of the early short-term forecast model, Lewinsonde observation data is processed, and 20 vertical standard altitude arrangements are applied within a range of 1000 hPa to 1 hPa.
관측 자료 전처리부(130)는 이러한 1000hPa 내지 1hPa 구간의 20 개의 연직 표준 고도 배열에서 750hPa 내지 550hPa 구간에 연직으로 4 개의 배열을 추가하여 총 24 개의 연직 표준 고도 배열을 생성할 수 있다. 이는 밝은 띠의 최상점-최정점-최하점을 각각 다른 연직 표준 고도 배열에 적용하기 위함이다.The observation data pre-processing unit 130 may generate a total of 24 vertical standard altitude arrays by adding four arrays vertically in the 750 hPa to 550 hPa sections from the 20 vertical standard altitude arrays of the 1000 hPa to 1 hPa sections. This is to apply the bright band top-to-bottom-bottom to different vertical standard altitude arrangements.
관측 자료 전처리부(130)는 관측 자료 변수(varobs), 배경장 변수(cx) 및 배경 오차 변수(cxbgerr)의 네임리스트 또한 24 개의 연직 표준 고도 배열로 수정할 수 있다.The observation data pre-processing unit 130 may also modify the name list of observation data variables (varobs), background field variables (cx), and background error variables (cxbgerr) to 24 vertical standard altitude arrays.
관측 자료 전처리부(130)는 융해층의 고도 및 온도의 관측 자료를 이러한 연직 표준 고도 배열에 내삽하여 관측 자료 변수(varobs)를 추출하고, 초단기예보모델의 배경장에서 관측 자료 변수(varobs)와 동일한 변수를 배경장 변수(cx)로 추출할 수 있다.The observation data pre-processing unit 130 extracts observation data variables (varobs) by interpolating observation data of altitude and temperature of the melting layer into these vertical standard altitude arrays, and observes the observation data variables (varobs) in the background of the very short-term forecast model. The same variable can be extracted as a background field variable (cx).
변분 자료 동화부(150)는 초단기예보모델에서 3차원 변분 자료 동화(VAR) 단계의 물리 과정을 수행할 수 있다.The variable data assimilation unit 150 may perform a physical process of a 3D variable data assimilation (VAR) step in an early short-term forecast model.
초단기예보모델의 3차원 변분 자료 동화(VAR) 단계에서는 정해진 특정시간에 대해 관측 자료를 동화하여 초기장을 생성할 수 있다. 3차원 변분 자료 동화 방식은 4차원 변분 자료 동화 또는 앙상블 자료 동화와 같은 다른 자료 동화 방식에 비해 적은 계산 비용을 갖는다는 장점이 있다. In the 3D variable data assimilation (VAR) stage of the very short-term forecast model, an initial field can be generated by assimilation of observation data for a specific time. The three-dimensional variable data assimilation method has an advantage that it has less computational cost than other data assimilation methods such as four-dimensional variable data assimilation or ensemble data assimilation.
Figure PCTKR2019002464-appb-M000001
Figure PCTKR2019002464-appb-M000001
수학식 1에서 J는 3차원 변분 자료 동화에서 제어 변수가 동화될 때 발생하는 비용함수(패널티)를 의미하고, y는 관측 정보, x는 모델 분석장 정보, x0는 모델 배경장 정보, B-1은 모델 오차 공분산, R-1은 관측 오차 공분산 및 H(x)는 관측 자료를 모델 변수로 변환하여 주는 관측 연산자를 의미한다.In Equation 1, J means a cost function (penalty) that occurs when a control variable is assimilated in 3D variable data assimilation, y is observation information, x is model analysis field information, x 0 is model background field information, and B -1 is the model error covariance, R -1 is the observation error covariance, and H (x) is the observation operator that converts the observation data into model variables.
초단기예보모델의 3차원 변분 자료 동화(VAR) 단계에서는 3차원 변분 자료 동화 방식의 장점을 극대화 하고, 다른 자료 동화 방식에 비해 성능이 낮은 단점을 보완하기 위해 정해진 시간마다 관측 자료를 동화하는 순환 자료 동화 방식(cycling)을 적용할 수 있다.In the 3D variable data assimilation (VAR) stage of the early short-term forecast model, cyclic data that assimilates observation data at a fixed time to maximize the advantages of the 3D variable data assimilation method and to compensate for the disadvantages of lower performance than other data assimilation methods Cycling can be applied.
도 4는 3차원 변분 자료 동화 방식을 설명하기 위한 도면이다.4 is a diagram for explaining a 3D variable data assimilation method.
도 4에서 Obs는 관측 자료의 변수, Cx는 모델 배경장 변수, Cw는 잠재적으로 사용 가능한 선형장(LS: Linearisation State) 변수 및 CxPlus는 관측 자료와 모델 배경장 간의 차이를 의미하며, 각각 3차원 변분 자료 동화 방식에서 사용되는 변수이다.In FIG. 4, Obs is a variable of observation data, Cx is a model background field variable, Cw is a potentially usable linearization state (LS) variable, and CxPlus represents a difference between observation data and a model background field, each of which is 3D. It is a variable used in the variable data assimilation method.
또한, v_hat은 관측 자료의 결과, CxPlus_hat은 관측 자료와 모델 배경장 간의 차이 결과, Cw_hat은 기울기 및 ModelOb_hat은 관측 자료에 대한 모델 배경장의 결과를 의미하며, 각각 3차원 변분 자료 동화 방식에 따른 최종 산출 결과물이다.In addition, v_hat is the result of observation data, CxPlus_hat is the result of the difference between the observation data and the model background field, Cw_hat is the slope and ModelOb_hat is the result of the model background field for the observation data. It is the result.
도 4를 참조하면, 3차원 변분 자료 동화 방식의 첫 번째 단계는 자료 동화 수행 시 필요한 설정 정보들을 읽어들일 수 있다. 관측 자료의 네임리스트(Observation namelist)에서 관측 자료에 대한 설정 정보를 읽어들이고, 자료 동화 수행 시 최소화 변수(Minimization namelist), 자료 동화 제어 변수(Transform namelist) 및 진단 변수(Diagnose namelist) 등 자료 동화 수행 시 필요한 설정 정보들을 읽어들여 시스템 구동 준비를 할 수 있다.Referring to FIG. 4, the first step of the 3D variable data assimilation method may read setting information necessary when performing data assimilation. Read configuration information for observation data from the Observation namelist and perform data assimilation such as Minimization namelist, Data assimilation control variable (Transform namelist) and Diagnostic variable (Diagnose namelist) You can prepare for system operation by reading the necessary configuration information.
3차원 변분 자료 동화 방식의 두 번째 및 세 번째 단계는 선형장 정보를 불러오기 위한 준비 단계 및 선형장 정보를 읽어들이는 단계에 해당한다.The second and third steps of the 3D variable data assimilation method correspond to a preparation step for loading linear field information and a step for reading linear field information.
3차원 변분 자료 동화 방식의 네 번째 및 다섯 번째 단계는 편차장(PF: Perterbation Field) 분석을 위해 메모리 할당, 배열 등을 설정하고, 모델의 초기장(First guess)과 모델의 배경장의 차이를 통해 편차장을 산출하는 단계에 해당한다. The 4th and 5th steps of the 3D variable data assimilation method set memory allocation, arrangement, etc. for the analysis of the deviation field (PF) through the difference between the model's first guess and the model's background field. This corresponds to the step of calculating the deviation field.
3차원 변분 자료 동화 방식의 여섯 번째 단계는 관측 자료를 수치예보모델에 적용하기 위해 네 가지 제어 변수(AP, mu, chi, psi)의 배열을 구축할 수 있다.In the sixth step of the 3D variable data assimilation method, an array of four control variables (AP, mu, chi, and psi) can be constructed to apply the observed data to the numerical prediction model.
3차원 변분 자료 동화 방식의 일곱 번째 단계는 네임리스트에서 설정한 관측종인 관측 자료에 대한 설정 정보에 대해 자료 동화에 필요한 변수의 배열을 구축할 수 있다.In the seventh step of the 3D variable data assimilation method, it is possible to construct an array of variables necessary for data assimilation for setting information for observation data, which is an observation species set in a name list.
3차원 변분 자료 동화 방식의 여덟 번째 단계는 관측 자료에 대한 최소화(minimization) 과정에 해당한다. 수학식 1과 같이 관측 자료 및 모델 배경장의 비용함수(Jo, Jb)가 산출되고, 전체 비용함수(J)가 산출될 수 있다. 그리고, 관측 자료가 모델 배경장에 적용되는 과정에서 관측 자료와 모델 배경장 간 1에 가까운 기울기(gradient)를 찾기 위한 외부반복순환(Outer loop)과 내부반복순환(Inner loop)이 이루어질 수 있다. The eighth step of the 3D variable data assimilation method corresponds to the process of minimization of the observed data. As shown in Equation 1, the cost functions (J o , J b ) of the observation data and the model background field are calculated, and the total cost function (J) can be calculated. Then, in the process of observation data being applied to the model background field, an outer loop and an inner loop may be performed to find a gradient close to 1 between the observed data and the model background field.
Figure PCTKR2019002464-appb-M000002
Figure PCTKR2019002464-appb-M000002
Figure PCTKR2019002464-appb-M000003
Figure PCTKR2019002464-appb-M000003
Figure PCTKR2019002464-appb-M000004
Figure PCTKR2019002464-appb-M000004
수학식 2는 관측 자료의 패널티를 산출하는 데에 사용되고, 수학식 3 및 수학식 4는 각각 Cw 및 ModelOb 변수의 기울기를 산출하는 데에 사용된다. Equation 2 is used to calculate the penalty of observation data, and equations 3 and 4 are used to calculate the slope of the Cw and ModelOb variables, respectively.
즉, 3차원 변분 자료 동화 방식의 여덟 번째 단계에 따르면, 관측 자료들에 대한 관측 패널티가 계산되고, 이를 이용하여 관측 자료와 모델 배경장 간의 기울기가 계산될 수 있으며, 1에 가까운 최적의 기울기를 찾기 위한 최소화 과정을 수행하여 관측 자료가 적용되기 전 수치예보모델의 초기장과 관측 자료가 적용되었을 때의 초기장 간의 분석 증분(Analysis increment)이 최종 산출될 수 있다. That is, according to the eighth step of the 3D variable data assimilation method, the observation penalty for the observed data is calculated, and the slope between the observed data and the model background field can be calculated using this, and an optimal slope close to 1 is obtained. The analysis increment between the initial field of the numerical prediction model and the initial field when the observation data is applied can be finally calculated by performing the minimization process to find it.
변분 자료 동화부(150)는 이러한 3차원 변분 자료 동화 방식을 이용하여 초단기예보모델에 융해층 고도의 온도를 입력함으로써, 초단기예보모델에 따른 초기장의 온도 분포를 개선시킬 수 있다. 즉, 변분 자료 동화부(150)는 관측 자료 변수(varobs) 및 배경장 변수(cx) 간에 3차원 변분 자료 동화를 수행하여 분석 증분을 산출할 수 있다. 여기서, 관측 자료 변수(varobs)는 융해층의 고도 및 온도에 해당한다.The variable data assimilation unit 150 may improve the temperature distribution of the initial field according to the very short-term forecast model by inputting a high temperature of the melting layer into the very short-term forecast model using the three-dimensional variable data assimilation method. That is, the variable data assimilation unit 150 may calculate an analysis increment by performing a 3D variable data assimilation between the observed data variables (varobs) and the background field variable (cx). Here, the observed data variables (varobs) correspond to the altitude and temperature of the melting layer.
예를 들면, 변분 자료 동화부(150)는 아래 표 1과 같이 10 번의 외부반복순환과 내부반복순환을 통한 최소화 과정을 진행할 수 있다. 변분 자료 동화부(150)는 외부반복순환과 내부반복순환을 통해 수학식 1 및 2와 같이 관측 자료 및 모델 배경장에 대한 비용함수(J)를 산출하고, 수학식 3 및 4와 같이 비용함수(J)의 기울기를 산출할 수 있다. For example, the variable data assimilation unit 150 may perform a process of minimizing through 10 external and internal repeat cycles as shown in Table 1 below. The variable data assimilation unit 150 calculates the cost function (J) for the observed data and the model background field as shown in Equations 1 and 2 through the external and internal iterative cycles, and the cost function as shown in Equations 3 and 4 The slope of (J) can be calculated.
Figure PCTKR2019002464-appb-I000001
Figure PCTKR2019002464-appb-I000001
Figure PCTKR2019002464-appb-I000002
Figure PCTKR2019002464-appb-I000002
표 1을 참조하면 비용함수 및 기울기는 최초 119.6825 및 11.1699로 산출되었으며, 총 10 번의 외부반복순환과 내부반복순환을 통해 비용함수 및 기울기는 최종 80.8426 및 0.005로 산출됨으로써 최적화 된 기울기 값이 산출됨을 확인할 수 있다. 여기서, 기울기 값이 1에 가까울수록 최적화 된 기울기 값으로 간주될 수 있다.Referring to Table 1, it was confirmed that the cost function and slope were calculated as the first 119.6825 and 11.1699, and the optimized slope value was calculated by calculating the final cost function and slope to 80.8426 and 0.005 through 10 external and internal repeat cycles. Can be. Here, the closer the slope value is to 1, it can be regarded as an optimized slope value.
이처럼, 변분 자료 동화부(150)는 3차원 변분 자료 동화 방식을 이용하여 산출한 기울기 값을 관측 자료와 모델 배경장 간의 차이를 나타내는 값으로 하여, In this way, the variable data assimilation unit 150 uses the 3D variable data assimilation method to calculate the slope value as a value representing the difference between the observed data and the model background field,
관측 자료가 적용되기 전 수치예보모델의 초기장과 관측 자료가 적용되었을 때의 초기장 간의 분석 증분을 산출할 수 있다.Analysis increments can be calculated between the initial field of the numerical forecast model before the observation data are applied and the initial field when the observation data are applied.
초기장 예측부(170)는 3차원 변분 자료 동화의 최종 산출물인 분석 증분을 이용하여 초단기예보모델의 초기장을 생성할 수 있다. The initial field predicting unit 170 may generate an initial field of the very short-term forecast model using analytical increments, which are final outputs of 3D variable data assimilation.
구체적으로는, 초기장 예측부(170)는 3차원 변분 자료 동화의 최종 산출물인 분석 증분을 초기 역학장의 연직 온도 프로파일에 적용하여 초기 역학장을 개선할 수 있다. Specifically, the initial field prediction unit 170 may improve the initial dynamic field by applying an analysis increment, which is the final output of the 3D variable data assimilation, to the vertical temperature profile of the initial dynamic field.
초기장 예측부(170)는 위와 같이 분석 증분을 적용하여 생성한 초기 역학장에 기반하여 강수 예측을 수행할 수 있다.The initial field prediction unit 170 may perform precipitation prediction based on the initial dynamic field generated by applying an analysis increment as described above.
이와 같이, 본 발명에 따른 장치(100)는 레이더 반사도 자료에서 밝은 띠를 탐지하여 융해층 고도의 온도를 산출할 수 있다. 종관 관측자료의 경우 시공간적으로 해상도가 낮은 반면 레이더 반사도 자료의 경우 시공간적 해상도가 높으므로 보다 정확한 융해층 고도 및 온도의 산출이 가능하다. 또한, 본 발명에 따른 장치(100)는 이러한 융해층 고도의 온도 자료를 3차원 변분 자료 동화 방식을 통해 초기 역학장의 온도 프로파일에 입력함으로써, 수치예보모델에서 모의한 온도장의 상층, 중층 및 하층의 초기 온도 정보를 보정할 수 있다. 또한, 본 발명에 따른 장치(100)는 개선된 초기 역학장을 이용한 강수 예측을 수행하여 강수 예측 성능을 향상시킬 수 있다. In this way, the apparatus 100 according to the present invention can calculate the temperature of the melting layer altitude by detecting a bright band in the radar reflectivity data. In the case of synoptic observation data, the spatial and temporal resolution is low, while in the case of radar reflectivity data, the spatial and temporal resolution is high, so it is possible to calculate the altitude and temperature of the melting layer more accurately. In addition, the apparatus 100 according to the present invention inputs the temperature data of the molten layer altitude into the temperature profile of the initial dynamic field through a three-dimensional variable data assimilation method, so that the upper, middle and lower layers of the temperature field simulated in the numerical prediction model are obtained. Initial temperature information can be corrected. In addition, the apparatus 100 according to the present invention can improve precipitation prediction performance by performing precipitation prediction using an improved initial dynamic field.
도 5는 본 발명의 일 실시예에 따른 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법의 순서도이다.5 is a flowchart of a method for assimilation of an advanced data of a radar fusion layer based on an ultra-short-term forecasting model according to an embodiment of the present invention.
본 발명의 일 실시예에 따른 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법은 도 1에 도시된 장치(100)와 실질적으로 동일한 구성에서 진행될 수 있다. 따라서, 도 1의 장치(100)와 동일한 구성요소는 동일한 도면부호를 부여하고, 반복되는 설명은 생략하기로 한다.The method for assimilation of the elevation data of the radar melting layer based on the ultra-short-term forecasting model according to an embodiment of the present invention may be performed in substantially the same configuration as the apparatus 100 illustrated in FIG. 1. Therefore, the same components as the device 100 of FIG. 1 are given the same reference numerals, and repeated descriptions will be omitted.
도 5를 참조하면, 관측 자료 수집부(110)는 레이더 관측 자료를 수집하여 융해층 고도 및 온도를 산출할 수 있다(S10).Referring to FIG. 5, the observation data collection unit 110 may collect radar observation data to calculate the altitude and temperature of the melting layer (S10).
관측 자료 수집부(110)는 레이더 반사도 자료에서 밝은 띠를 탐지하고, 밝은 띠를 이용하여 융해층의 고도 및 온도를 산출할 수 있다. The observation data collection unit 110 may detect a bright band from the radar reflectivity data, and calculate the altitude and temperature of the melting layer using the bright band.
관측 자료 수집부(110)는 레이더 반사도 자료를 이용하여 평균 반사도 연직 프로파일을 생성할 수 있다. The observation data collection unit 110 may generate an average reflectivity vertical profile using radar reflectivity data.
관측 자료 수집부(110)는 밝은 띠 탐지에 필요한 5 개의 변수인 로그 취한 평균 반사도(log(ZH)), 일차 미분(f'(h)), 이차 미분(f''(h)), 곡률(C(h)), 곡률의 일차 미분(C'(h)) 연직 프로파일을 생성할 수 있다.Observation data collection unit 110 is a logarithmic mean reflectivity (log (Z H )), primary differential (f '(h)), secondary differential (f''(h)) Curvature (C (h)), primary derivative of curvature (C '(h)), can generate a vertical profile.
관측 자료 수집부(110)는 위와 같은 밝은 띠 탐지에 필요한 5 개의 프로파일을 이용하여 밝은 띠의 상단부, 최정점 및 하단부를 탐지할 수 있다.The observation data collection unit 110 may detect the top, top, and bottom of the bright band using the five profiles required for the detection of the bright band.
관측 자료 수집부(110)는 반사도 자료에서 이차 미분 값이 "+"이고, 일차 미분 값이 "+"에서 "-"로 변하는 지점의 고도를 밝은 띠의 최상점(BBTOP)으로 탐지할 수 있다.The observation data collection unit 110 can detect the altitude of the point where the secondary derivative value is “+” and the primary derivative value is changed from “+” to “-” in the reflectivity data as the highest point of the bright band (BB TOP ). have.
관측 자료 수집부(110)는 반사도 자료에서 이차 미분 값이 "-" 이고, 일차 미분 값이 "-"에서 "+"로 변하는 지점의 고도를 밝은 띠의 최하점(BBBOTTOM)으로 탐지할 수 있다.The observation data collection unit 110 may detect the altitude of the point where the secondary derivative value is "-" in the reflectivity data and the primary derivative value is changed from "-" to "+" as the lowest point of the bright band (BB BOTTOM ). .
이처럼, 관측 자료 수집부(110)는 레이더 반사도 자료에서 밝은 띠 영역에 해당하는 고도를 탐지할 수 있으며, 이를 융해층의 고도로 산출할 수 있다. 또한, 관측 자료 수집부(110)는 융해층의 고도와 레윈죤데 자료에서의 해당 고도를 비교하여 융해층 고도의 온도를 산출할 수 있다. As such, the observation data collection unit 110 may detect the altitude corresponding to the bright band region in the radar reflectivity data, and calculate it as the altitude of the melting layer. In addition, the observation data collection unit 110 may calculate the temperature of the elevation of the melting layer by comparing the elevation of the melting layer with the corresponding elevation in the Lewinsonde data.
관측 자료 수집부(110)는 위와 같이 레이더 반사도 자료로부터 산출한 융해층의 고도 및 온도를 버퍼 형태로 처리하여 데이터베이스에 저장할 수 있다.The observation data collection unit 110 may process the altitude and temperature of the melting layer calculated from the radar reflectivity data in the form of a buffer and store it in a database.
관측 자료 전처리부(130)는 레이더 관측 자료를 수집하여 산출한 융해층 고도 및 온도를 가공할 수 있다(S20).The observation data pre-processing unit 130 may process the altitude and temperature of the melting layer calculated by collecting radar observation data (S20).
관측 자료 전처리부(130)는 데이터베이스에 저장된 융해층의 고도를 불러들여 수치예보모델의 격자에 내삽하고, 융해층의 고도에 해당하는 공간정보 대한 수치예보모델의 배경장 변수를 추출할 수 있다.The observation data pre-processing unit 130 may retrieve the elevation of the melting layer stored in the database, interpolate it into the grid of the numerical prediction model, and extract the background field variable of the numerical prediction model for spatial information corresponding to the elevation of the melting layer.
이를 위해, 관측 자료 전처리부(130)는 연직 표준 고도 배열을 생성할 수 있다. 예를 들면, 관측 자료 전처리부(130)는 이러한 1000hPa 내지 1hPa 구간의 20 개의 연직 표준 고도 배열에서 750hPa 내지 550hPa 구간에 연직으로 4 개의 배열을 추가하여 총 24 개의 연직 표준 고도 배열을 생성할 수 있다. 이는 밝은 띠의 최상점-최정점-최하점을 각각 다른 연직 표준 고도 배열에 적용하기 위함이다.To this end, the observation data pre-processing unit 130 may generate a vertical standard altitude array. For example, the observation data pre-processing unit 130 may generate a total of 24 vertical standard altitude arrays by adding 4 arrays vertically in the 750 hPa to 550 hPa section from 20 vertical standard altitude arrays of 1000 hPa to 1 hPa sections. . This is to apply the bright band top-to-bottom-bottom to different vertical standard altitude arrangements.
관측 자료 전처리부(130)는 관측 자료 변수(varobs), 배경장 변수(cx) 및 배경 오차 변수(cxbgerr)의 네임리스트 또한 24 개의 연직 표준 고도 배열로 수정할 수 있다.The observation data pre-processing unit 130 may also modify the name list of observation data variables (varobs), background field variables (cx), and background error variables (cxbgerr) to 24 vertical standard altitude arrays.
관측 자료 전처리부(130)는 융해층의 고도 및 온도의 관측 자료를 이러한 연직 표준 고도 배열에 내삽하여 관측 자료 변수(varobs)를 추출하고, 초단기예보모델의 배경장에서 관측 자료 변수(varobs)와 동일한 변수를 배경장 변수(cx)로 추출할 수 있다.The observation data pre-processing unit 130 extracts observation data variables (varobs) by interpolating observation data of altitude and temperature of the melting layer into these vertical standard altitude arrays, and observes the observation data variables (varobs) in the background of the very short-term forecast model. The same variable can be extracted as a background field variable (cx).
변분 자료 동화부(150)는 융해층 고도 및 온도와 초단기예보모델의 배경장 간에 3차원 변분 자료 동화 방식을 적용하여 분석 증분을 산출할 수 있다(S30).The variable data assimilation unit 150 may calculate an analysis increment by applying a 3D variable data assimilation method between the altitude and temperature of the melting layer and the background field of the ultra-short-term forecast model (S30).
변분 자료 동화부(150)는 외부반복순환과 내부반복순환을 통해 수학식 1 및 2와 같이 관측 자료 및 모델 배경장에 대한 비용함수(J)를 산출하고, 수학식 3 및 4와 같이 비용함수(J)의 기울기를 산출할 수 있다. The variable data assimilation unit 150 calculates the cost function (J) for the observed data and the model background field as shown in Equations 1 and 2 through the external and internal iterative cycles, and the cost function as shown in Equations 3 and 4 The slope of (J) can be calculated.
변분 자료 동화부(150)는 3차원 변분 자료 동화 방식을 이용하여 산출한 기울기 값을 관측 자료와 모델 배경장 간의 차이를 나타내는 값으로 하여, 관측 자료가 적용되기 전 수치예보모델의 초기장과 관측 자료가 적용되었을 때의 초기장 간의 분석 증분을 산출할 수 있다.The variable data assimilation unit 150 uses the 3D variable data assimilation method as a value representing the difference between the observation data and the model background field, and the initial field and observation of the numerical prediction model before the observation data are applied. Analysis increments between initial fields when data are applied can be calculated.
초기장 예측부(170)는 분석 증분을 이용하여 초단기예보모델의 초기장을 생성하고 이를 이용하여 강수 예측을 수행할 수 있다(S40).The initial field prediction unit 170 may generate an initial field of the very short-term forecast model using the incremental analysis and use it to perform precipitation prediction (S40).
초기장 예측부(170)는 3차원 변분 자료 동화의 최종 산출물인 분석 증분을 초단기예보모델 초기 역학장의 연직 온도 프로파일에 적용하여 초기 역학장을 개선할 수 있다. 초기장 예측부(170)는 위와 같이 분석 증분을 적용하여 생성한 초기 역학장에 기반하여 강수 예측을 수행할 수 있다. The initial field prediction unit 170 may improve the initial dynamic field by applying an analysis increment, which is the final output of the 3D variable data assimilation, to the vertical temperature profile of the initial dynamic field of the very short-term forecast model. The initial field prediction unit 170 may perform precipitation prediction based on the initial dynamic field generated by applying an analysis increment as described above.
이와 같은, 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법은 애플리케이션으로 구현되거나 다양한 컴퓨터 구성요소를 통하여 수행될 수 있는 프로그램 명령어의 형태로 구현되어 컴퓨터 판독 가능한 기록 매체에 기록될 수 있다. 상기 컴퓨터 판독 가능한 기록 매체는 프로그램 명령어, 데이터 파일, 데이터 구조 등을 단독으로 또는 조합하여 포함할 수 있다.As described above, the method for assimilation of the advanced data of the radar fusion layer based on the ultra-short-term forecasting model of the present invention may be implemented as an application or implemented in the form of program instructions that can be executed through various computer components to be recorded in a computer-readable recording medium. . The computer-readable recording medium may include program instructions, data files, data structures, or the like alone or in combination.
상기 컴퓨터 판독 가능한 기록 매체에 기록되는 프로그램 명령어는 본 발명을 위하여 특별히 설계되고 구성된 것들이거니와 컴퓨터 소프트웨어 분야의 당업자에게 공지되어 사용 가능한 것일 수도 있다.The program instructions recorded on the computer-readable recording medium are specially designed and configured for the present invention, and may be known and available to those skilled in the computer software field.
컴퓨터 판독 가능한 기록 매체의 예에는, 하드 디스크, 플로피 디스크 및 자기 테이프와 같은 자기 매체, CD-ROM, DVD 와 같은 광기록 매체, 플롭티컬 디스크(floptical disk)와 같은 자기-광 매체(magneto-optical media), 및 ROM, RAM, 플래시 메모리 등과 같은 프로그램 명령어를 저장하고 수행하도록 특별히 구성된 하드웨어 장치가 포함된다.Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs, DVDs, and magneto-optical media such as floptical disks. media), and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like.
프로그램 명령어의 예에는, 컴파일러에 의해 만들어지는 것과 같은 기계어 코드뿐만 아니라 인터프리터 등을 사용해서 컴퓨터에 의해서 실행될 수 있는 고급 언어 코드도 포함된다. 상기 하드웨어 장치는 본 발명에 따른 처리를 수행하기 위해 하나 이상의 소프트웨어 모듈로서 작동하도록 구성될 수 있으며, 그 역도 마찬가지이다.Examples of program instructions include not only machine language codes produced by a compiler, but also high-level language codes executable by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules to perform processing according to the present invention, and vice versa.
이하, 도 6a 내지 도 12d를 참조하여 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 유리한 효과에 대해 설명한다.Hereinafter, with reference to FIGS. 6A to 12D, an advantageous effect according to the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention will be described.
먼저, 도 6a 내지 도 7d은 층운형 사례에서 레이더 관측 자료로부터 산출한 밝은 띠 고도를 나타내는 그래프이다.First, FIGS. 6A to 7D are graphs showing bright band altitudes calculated from radar observation data in a stratospheric case.
먼저, 2016년 5월 2일 층운형 사례를 대상으로 한 광덕산(GDK), 오성산(KSN), 진도(JNI) 및 고산(GSN)의 네 개 지점의 S-band 기상 레이더 관측 자료를 수집하였으며, 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따라 각 레이더 관측 자료를 이용하여 산출한 밝은 띠 고도는 도 6a 내지 도 6d과 같다.First, on May 2, 2016, S-band weather radar observation data were collected for four locations: Gwangdeoksan (GDK), Oseongsan (KSN), Jindo (JNI), and Gosan (GSN), for stratified cloud cases. The bright band altitude calculated using the radar observation data according to the radar melting layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention is shown in FIGS. 6A to 6D.
도 6a를 참조하면, 광덕산(GDK) 레이더 관측 자료로부터 밝은 띠의 최고점, 최정점 및 최저점의 평균 고도는 각각 4.22km, 3.65km 및 3.14km 로 산출되었고, 밝은 띠의 최고점, 최정점 및 최저점 고도의 온도는 각각 -2.1℃, 0.8℃ 및 4.4℃로 산출되었다. Referring to FIG. 6A, the average altitudes of the peak, peak, and bottom of the bright band from the Gwangdeoksan (GDK) radar observation data were calculated to be 4.22 km, 3.65 km, and 3.14 km, respectively, and the temperature of the peak, peak, and lowest altitude of the bright band. Was calculated as -2.1 ° C, 0.8 ° C and 4.4 ° C, respectively.
도 6b를 참조하면, 오성산(KSN) 레이더 관측 자료로부터 밝은 띠의 최고점, 최정점 및 최저점의 평균 고도는 각각 4.35km, 3.70km 및 3.19km 로 산출되었고, 밝은 띠의 최고점, 최정점 및 최저점 고도의 온도는 각각 -1.15℃, 2.6℃ 및 5.4℃로 산출되었다.Referring to FIG. 6B, the average altitudes of the peaks, peaks, and troughs of the bright band from the Oseong Mountain (KSN) radar observations were calculated to be 4.35 km, 3.70 km, and 3.19 km, respectively, and the temperatures of the peak, peak, and trough altitudes of the bright bands, respectively. Was calculated as -1.15 ° C, 2.6 ° C and 5.4 ° C, respectively.
도 6c를 참조하면, 진도(JNI) 레이더 관측 자료로부터 밝은 띠의 최고점, 최정점 및 최저점의 평균 고도는 각각 4.46km, 3.82km 및 3.33km 로 산출되었고, 밝은 띠의 최고점, 최정점 및 최저점 고도의 온도는 각각 -2.0℃, 2.4℃ 및 5.1℃로 산출되었다. Referring to FIG. 6C, the average altitudes of the peak, peak, and bottom of the bright band from the intensity (JNI) radar observations were calculated to be 4.46 km, 3.82 km, and 3.33 km, respectively, and the temperature of the peak, peak, and lowest altitude of the bright band. Was calculated as -2.0 ° C, 2.4 ° C and 5.1 ° C, respectively.
도 6d를 참조하면, 고산(GSN) 레이더 관측 자료로부터 밝은 띠의 최고점, 최정점 및 최저점의 평균 고도는 각각 4.64km, 4.08km 및 3.53km 로 산출되었고, 밝은 띠의 최고점, 최정점 및 최저점 고도의 온도는 각각 -2.7℃, 1.0℃ 및 3.5℃로 산출되었다. Referring to FIG. 6D, from the alpine (GSN) radar observation data, the average altitudes of the peaks, peaks, and troughs of the bright bands were calculated to be 4.64 km, 4.08 km, and 3.53 km, respectively, and the temperatures of the peaks, peaks, and trough altitudes of the bright bands, respectively. Was calculated as -2.7 ° C, 1.0 ° C and 3.5 ° C, respectively.
이때, 광덕산(GDK) 및 오성산(KSN) 레이더 관측지점과 비교를 위한 오산의 레윈존데 관측지점의 융해층 평균 고도는 4.18km로 관측되었다.At this time, the average altitude of the melting layer of Osan's Lewinzonde observation point for comparison with the observation points of Gwangdeok Mountain (GDK) and Oseong Mountain (KSN) radar was 4.18 km.
진도(JNI) 레이더 관측지점과 비교를 위한 광주 레윈존데 관측지점의 융해층 평균 고도는 4.09km 로 관측되었다.For comparison with the Jindo (JNI) radar observation point, the average elevation of the melting layer at the Lewinzonde Gwangju observation point was 4.09 km.
고산(GSN) 레이더 관측지점과 비교를 위한 제주 레윈존데 관측지점의 융해층 평균 고도는 3.84km 로 관측되었다.The average elevation of the melting layer of Jeju Lewinzonde observation point for comparison with the observation point of the radar observation point (GSN) was 3.84 km.
또한, 2016년 7월 1일 층운형 사례를 대상으로 한 광덕산(GDK), 오성산(KSN), 진도(JNI) 및 고산(GSN)의 네 개 지점의 S-band 기상 레이더 관측 자료를 수집하였으며, 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따라 각 레이더 관측 자료를 이용하여 산출한 밝은 띠 고도는 도 7a 내지 도 7d와 같다.Also, on July 1, 2016, S-band weather radar observation data were collected for four locations: Gwangdeoksan (GDK), Oseongsan (KSN), Jindo (JNI), and Gosan (GSN), for stratified cloud cases. The bright band altitude calculated using the radar observation data according to the radar melting layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention is shown in FIGS. 7A to 7D.
도 7a를 참조하면, 광덕산(GDK) 레이더 관측 자료로부터 밝은 띠의 최고점, 최정점 및 최저점의 평균 고도는 각각 4.22km, 3.65km 및 3.14km 로 산출되었고, 밝은 띠의 최고점, 최정점 및 최저점 고도의 온도는 각각 -2.1℃, 0.8℃ 및 4.4℃로 산출되었다. Referring to FIG. 7A, the average altitudes of the peaks, peaks, and troughs of the bright bands from the Gwangdeoksan (GDK) radar observations were calculated to be 4.22 km, 3.65 km, and 3.14 km, respectively, and the temperature of the peaks, peaks, and troughs altitudes of the bright bands, respectively. Was calculated as -2.1 ° C, 0.8 ° C and 4.4 ° C, respectively.
도 7b를 참조하면, 오성산(KSN) 레이더 관측 자료로부터 밝은 띠의 최고점, 최정점 및 최저점의 평균 고도는 각각 4.35km, 3.70km 및 3.19km 로 산출되었고, 밝은 띠의 최고점, 최정점 및 최저점 고도의 온도는 각각 -1.15℃, 2.6℃ 및 5.4℃로 산출되었다.Referring to FIG. 7B, the average altitudes of the peaks, peaks, and troughs of the bright bands from the Oseong Mountain (KSN) radar observations were calculated to be 4.35 km, 3.70 km, and 3.19 km, respectively, and the temperatures of the peaks, peaks, and troughs altitudes of the bright bands, respectively. Was calculated as -1.15 ° C, 2.6 ° C and 5.4 ° C, respectively.
도 7c를 참조하면, 진도(JNI) 레이더 관측 자료로부터 밝은 띠의 최고점, 최정점 및 최저점의 평균 고도는 각각 4.46km, 3.82km 및 3.33km 로 산출되었고, 밝은 띠의 최고점, 최정점 및 최저점 고도의 온도는 각각 -2.0℃, 2.4℃ 및 5.1℃로 산출되었다. Referring to FIG. 7C, the average altitudes of the peak, peak, and bottom of the bright band from the intensity (JNI) radar observations were calculated to be 4.46 km, 3.82 km, and 3.33 km, respectively, and the temperature of the peak, peak, and lowest altitude of the bright band. Was calculated as -2.0 ° C, 2.4 ° C and 5.1 ° C, respectively.
도 7d를 참조하면, 고산(GSN) 레이더 관측 자료로부터 밝은 띠의 최고점, 최정점 및 최저점의 평균 고도는 각각 4.64km, 4.08km 및 3.53km 로 산출되었고, 밝은 띠의 최고점, 최정점 및 최저점 고도의 온도는 각각 -2.7℃, 1.0℃ 및 3.5℃로 산출되었다. Referring to FIG. 7D, from the alpine (GSN) radar observation data, the average altitudes of the peaks, peaks, and troughs of the bright bands were calculated to be 4.64 km, 4.08 km, and 3.53 km, respectively, and the temperatures of the peaks, peaks, and troughs altitudes of the bright bands, respectively. Was calculated as -2.7 ° C, 1.0 ° C and 3.5 ° C, respectively.
이때, 광덕산(GDK) 및 오성산(KSN) 레이더 관측지점과 비교를 위한 오산의 레윈존데 관측지점의 융해층 평균 고도는 4.18km로 관측되었다.At this time, the average altitude of the melting layer of Osan's Lewinzonde observation point for comparison with the observation points of Gwangdeok Mountain (GDK) and Oseong Mountain (KSN) radar was 4.18 km.
진도(JNI) 레이더 관측지점과 비교를 위한 광주 레윈존데 관측지점의 융해층 평균 고도는 4.09km 로 관측되었다.For comparison with the Jindo (JNI) radar observation point, the average elevation of the melting layer at the Lewinzonde Gwangju observation point was 4.09 km.
고산(GSN) 레이더 관측지점과 비교를 위한 제주 레윈존데 관측지점의 융해층 평균 고도는 3.84km 로 관측되었다.The average elevation of the melting layer of Jeju Lewinzonde observation point for comparison with the observation point of the radar observation point (GSN) was 3.84 km.
이와 같은 레이더 관측 자료로부터 산출한 융해층 고도 및 온도를 살펴보면, 남쪽 지역일수록 저기압이 접근하거나 그 영향권에 있을 때 밝은 띠의 고도 및 온도가 높게 산출됨을 확인할 수 있다. 또한, 레윈존데 관측지점에서 관측되는 융해층 고도 이하에서 밝은 띠의 최정점이 산출됨을 확인할 수 있다. 이로부터 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따라 밝은 띠로부터 산출하는 융해층의 온도 정보를 수치예보모델에 적용하는 경우, 수치예보모델의 개선 효과를 기대할 수 있다.When looking at the altitude and temperature of the melting layer calculated from these radar observations, it can be seen that the higher the altitude and temperature of the bright band, the lower the pressure of the region approaching or the lower the region. In addition, it can be seen that the peak of a bright band is calculated below the altitude of the melting layer observed at the Lewinsonde observation point. From this, when the temperature information of the melting layer calculated from the bright band is applied to the numerical prediction model according to the radar melting layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention, an improvement effect of the numerical prediction model can be expected.
도 8a 및 도 8b는 장마전선 사례에서 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따라 생성한 초기 역학장의 연직 온도 프로파일을 나타내는 그래프이다.8A and 8B are graphs showing the vertical temperature profile of the initial dynamic field generated according to the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention in the rainy season case.
2016년 7월 1일 0000 UTC ~ 1200 UTC 장마전선 사례의 레이더 관측 자료로부터 산출하는 융해층의 고도 및 온도와 모델 배경장 정보 간의 자료 동화를 수행한 결과 생성되는 초기 역학장의 연직 온도 프로파일은 도 8a 및 도 8b와 같다.July 1, 2016 0000 UTC to 1200 UTC The vertical temperature profile of the initial dynamic field generated as a result of data assimilation between the elevation and temperature of the fusion layer and model background field information from radar observation data of the rainy season case is shown in FIG. 8A and It is like FIG. 8B.
도 8a 및 도 8b를 참조하면, 관악산(KWK) 및 광덕산(GDK) 레이더 사이트의 초기 역학장 연직 프로파일을 확인할 수 있는데, 대기 하층에서 음의 변화를 나타내고, 대기 중상층에서 양의 변화를 나타내어 전체적으로 대기의 안정화가 이루어짐을 확인할 수 있다. 8A and 8B, it is possible to confirm the initial dynamic field vertical profile of the Gwanaksan (KWK) and Gwangdeoksan (GDK) radar sites, showing a negative change in the lower atmosphere and a positive change in the upper middle atmosphere, thereby showing the overall atmosphere. It can be confirmed that the stabilization of the.
도 9a 내지 도 12d는 장마전선 사례에서 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 역학장 변화를 나타낸 그래프이다.9A to 12D are graphs showing changes in the dynamic field according to an assimilation method of an elevation data of a radar fusion layer based on the ultra-short-term forecast model of the present invention in the rainy season front-line case.
2016년 7월 1일 0000 UTC ~ 1200 UTC 장마전선 사례에서 대기의 하층(1500m 고도), 대기 중층(4000m 고도) 및 대기 상층(8000m 고도)를 기준으로 하여 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따라 레이더 관측 자료가 적용되어 보정된 역학장과, 자료 동화 방식이 적용되지 않은 규준실험을 비교한 관측 증분 분석 결과는 각각 도 9a 내지 도 10f와 같다.July 1, 2016 0000 UTC ~ 1200 UTC In the rainy season case, based on the lower layer of the atmosphere (1500m altitude), the upper atmosphere (4000m altitude) and the upper atmosphere (8000m altitude), the radar fusion layer based on the ultra-short-term forecast model of the present invention The results of the observational incremental analysis comparing the dynamic field corrected by applying the radar observation data according to the altitude data assimilation method and the normative experiment without the data assimilation method are shown in FIGS. 9A to 10F, respectively.
도 9a 내지 도 9d를 참조하면, 대기 하층에서는 내륙지방을 중심으로 기압 및 온도가 각각 0.08hPa 및 0.1℃ 이상 감소하였고, qvapor은 서해안 지방을 중심으로 증가하였으며, qrain은 백령도 인근 지역에서 주로 감소한 결과를 보인다.9A to 9D, in the lower atmosphere, atmospheric pressure and temperature decreased by 0.08 hPa and 0.1 ° C or higher, respectively, in the inland region, q vapor increased mainly in the west coast region, and q rain was mainly in the region near Baeknyeongdo. It shows reduced results.
도 10a 내지 도 10f를 참조하면, 대기 중층에서는 서해안 지방을 중심으로 기압이 0.06hPa 이상 감소하였고, 온도는 수도권을 포함한 중서부지방에서 0.1℃ 이상 증가, 광주 인근지역, 남해상 및 제주도는 0.2℃ 이상 감소하였고, qvapor은 수도권 및 경기도 지방에서 감소, 전남 및 제주 인근에서 증가하였으며, qrain은 남서쪽 해상에서 증가한 결과를 보인다. 또한, qgraupel은 백령도 인근 지역에서 증가하였고, qcf는 감소한 결과를 보인다.10A to 10F, in the atmospheric middle layer, the air pressure in the western coastal region decreased by more than 0.06 hPa, and the temperature increased by more than 0.1 ° C in the midwestern region including the metropolitan area, the area near Gwangju, the southern sea and Jeju Island decreased by more than 0.2 ° C. Q vapor decreased in the metropolitan area and Gyeonggi-do, increased near Jeonnam and Jeju, and q rain increased in the southwestern sea. In addition, q graupel increased in the vicinity of Baeknyeongdo Island, and q cf decreased.
도 11a 내지 도 11e를 참조하면, 대기 상층에서는 대기 하층에서의 결과와 반대로 내륙지방을 중심으로 기압이 0.06hPa 이상 증가, 남해상에는 0.08hPa 이상 감소하였고, 온도는 내륙지방을 중심으로 0.06℃ 이상 증가하였고, qvapor은 백령도 인근지역에서 증가, 수도권 및 남부내륙 일부 지방에는 약하게 감소하였으며, qcf는 백령도 인근지역에서 감소한 결과를 보인다.Referring to FIGS. 11A to 11E, in the upper atmosphere, the air pressure increased by 0.06 hPa or more, and decreased by 0.08 hPa or more in the south sea, and the temperature increased by 0.06 ° C or more in the inland region, as opposed to the results in the lower atmosphere. Q vapor increased in the vicinity of Baeknyeongdo Island, weakly decreased in some areas of the metropolitan area and the southern inland, and q cf decreased in the vicinity of Baeknyeongdo Island.
이와 같이, 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따라 융해층 고도의 온도를 자료 동화한 결과, 장마전선에 의해 내륙지방에 발생한 강수 모의를 조정해주기 위해 대기의 하층, 중층 및 상층의 역학장이 변화하였음을 확인할 수 있다. 하층의 경우, 온도가 낮아져 상대적으로 차가워졌으며, 남부지방에 위치한 장마전선에 의해 모의된 강수를 줄여주기 위해 중부지방의 기압은 감소하고 남부지방의 기압은 증가하였다. 또한, 중서부지방의 qvapor이 증가하여 해당 지역이 습해졌다. 중층의 경우, 기압변화 패턴은 하층과 유사하고, qvapor은 중부지방에서 감소, 남부지방에서 증가하였고, 온도는 중부지방에서 증가, 남부지방에서 감소하였다. 이로부터, 중부지방은 대기 중층이 건조해지고 남부지방은 대기 중층이 습해졌음을 확인할 수 있다. 상층의 경우, 내륙지방을 중심으로 온도가 증가하여 따뜻해졌으며, 대기 중, 하층과 반대되는 패턴을 보였다. 전체적으로 내륙지방을 중심으로 대기 하층이 차가워지고 대기 상층이 따뜻해짐에 따라 대기 안정화를 모의하도록 역학장이 변하였음을 확인할 수 있다.As such, as a result of data assimilation of the temperature of the fusion layer altitude according to the radar melting layer altitude data assimilation method based on the ultra-short-term forecasting model of the present invention, the lower layer, the middle layer of the atmosphere and It can be confirmed that the upper dynamic field has changed. In the case of the lower layer, the temperature was lowered and it became relatively cold, and in order to reduce the precipitation simulated by the rainy season electric line located in the southern region, the air pressure in the central region decreased and the air pressure in the southern region increased. In addition, the q vapor in the Midwest increased, making the area wet. In the middle layer, the pressure change pattern is similar to the lower layer, and q vapor decreases in the central region and increases in the southern region, and temperature increases in the central region and decreases in the southern region. From this, it can be confirmed that the middle layer of the atmosphere became dry and the southern layer became humid. In the case of the upper layer, the temperature increased with the center of the inland area warming up, and showed the opposite pattern to the atmosphere and the lower layer. As a whole, it can be confirmed that the dynamic field changed to simulate the stabilization of the atmosphere as the lower layer of the atmosphere became colder and the upper layer warmed around the inland region.
도 12a 내지 도 12d는 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 강수 모의 결과와 규준 실험 간 강수 모의 결과를 비교한 그래프이다.12A to 12D are graphs comparing the precipitation simulation result between the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention and the precipitation simulation result between the normative experiments.
2016년 7월 1일 0000 UTC ~ 1200 UTC의 12 시간 동안 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 강수 모의 결과(EXP)와, 자료 동화 방식이 적용되지 않은 규준실험에 다른 강수 모의 결과(CTRL)와, 지상관측자료(AWS) 및 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 강수 모의 결과(EXP)와 자료 동화 방식이 적용되지 않은 규준실험에 다른 강수 모의 결과(CTRL)의 차이(EXP-CTRL)는 각각 도 12a 내지 도 12d에 도시되어 있다.On July 1, 2016, the simulation result (EXP) according to the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention for 12 hours from 0000 UTC to 1200 UTC was applied to the normative experiment to which the data assimilation method was not applied. Precipitation simulation results (EXP) and data assimilation methods based on different precipitation simulation results (CTRL), ground observation data (AWS) and radar fusion layer elevation data based on the very short-term forecast model of the present invention The differences (EXP-CTRL) of the different precipitation simulation results CTRL are shown in FIGS. 12A to 12D, respectively.
도 12a 내지 도 12d를 참조하면, 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 강수 모의 결과(EXP)와 자료 동화 방식이 적용되지 않은 규준실험에 다른 강수 모의 결과(CTRL)를 비교하였을 때에 관측 증분 결과와 같이 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 강수 모의 결과(EXP)의 경우, 남부내륙 및 남해상을 중심으로 강수량이 감소하고, 중부내륙을 중심으로 강수량이 증가한 경향을 보인다.12A to 12D, precipitation simulation results (EXP) according to the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention and other precipitation simulation results (CTRL) in the normative experiment to which the data assimilation method is not applied When comparing the results of precipitation simulation (EXP) according to the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention as in the observation incremental results, precipitation decreases centered on the southern and southern seas, and the central inland. In the center, precipitation tends to increase.
초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 강수 모의 결과(EXP)와 자료 동화 방식이 적용되지 않은 규준실험에 다른 강수 모의 결과(CTRL)를 각각 지상관측자료(AWS)와 비교하면, 본 발명의 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방식에 따른 강수 모의 결과(EXP)의 경우, 규준실험에 다른 강수 모의 결과(CTRL)에서 과대모의 하였던 남부지방의 강수가 감소하여 개선된 결과를 보인다.When comparing the precipitation simulation results (EXP) according to the advanced data assimilation method based on the early short-term forecast model and the other precipitation simulation results (CTRL) in the normative experiment to which the data assimilation method is not applied, compared to the ground observation data (AWS), In the case of the precipitation simulation result (EXP) according to the radar fusion layer elevation data assimilation method based on the ultra-short-term forecast model of the present invention, the precipitation result in the southern region, which was overestimated in the precipitation simulation result (CTRL) in the normative experiment, was improved. Looks like
이상에서는 실시예들을 참조하여 설명하였지만, 해당 기술 분야의 숙련된 당업자는 하기의 특허 청구범위에 기재된 본 발명의 사상 및 영역으로부터 벗어나지 않는 범위 내에서 본 발명을 다양하게 수정 및 변경시킬 수 있음을 이해할 수 있을 것이다.Although described above with reference to embodiments, those skilled in the art understand that various modifications and changes can be made to the present invention without departing from the spirit and scope of the present invention as set forth in the claims below. Will be able to.

Claims (15)

  1. 기상 레이더 시스템으로부터 레이더 관측 자료를 수집하여 융해층의 고도 및 온도를 포함하는 관측 자료를 산출하는 단계;Collecting radar observation data from a weather radar system and calculating observation data including altitude and temperature of the melting layer;
    상기 관측 자료를 자료 동화에 적용할 수 있도록 전처리하는 단계;Pre-processing the observation data to be applied to data assimilation;
    상기 관측 자료를 초단기예보모델의 배경장 간에 3차원 변분 자료 동화 방식을 적용하여 분석 증분을 산출하는 단계; 및Calculating an analysis increment by applying a three-dimensional variable data assimilation method between the background fields of the very short-term forecast model of the observed data; And
    상기 분석 증분을 이용하여 초단기예보모델의 초기장을 생성하고 강수를 예측하는 단계를 포함하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법.A method of assembling altitude data of a radar melting layer based on an ultra-short forecast model, comprising generating an initial field of the early short-term forecast model and predicting precipitation using the analysis increment.
  2. 제1항에 있어서,According to claim 1,
    기상 레이더 시스템으로부터 레이더 관측 자료를 수집하여 융해층의 고도 및 온도를 포함하는 관측 자료를 산출하는 단계는,The step of collecting the radar observation data from the weather radar system and calculating the observation data including the altitude and temperature of the melting layer,
    상기 레이더 관측 자료에서 밝은 띠를 탐지하는 단계; Detecting a bright band in the radar observation data;
    상기 밝은 띠를 상기 융해층의 고도로 산출하는 단계; 및Calculating the bright band to a height of the fusion layer; And
    상기 융해층의 고도와 레윈죤데 자료를 비교하여 상기 융해층 온도를 산출하는 단계를 포함하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법.Comprising the step of calculating the melting layer temperature by comparing the elevation of the melting layer and the Lewinsonde data Radar melting layer elevation data assimilation method based on the short-term forecast model.
  3. 제1항에 있어서,According to claim 1,
    기상 레이더 시스템으로부터 레이더 관측 자료를 수집하여 융해층의 고도 및 온도를 포함하는 관측 자료를 산출하는 단계는,The step of collecting the radar observation data from the weather radar system and calculating the observation data including the altitude and temperature of the melting layer,
    상기 관측 자료를 버퍼 형태로 처리하여 데이터베이스에 저장하는 단계를 포함하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법.A method of assembling altitude data of a radar melting layer based on an ultra-short-term forecast model, which includes processing the observation data in a buffer and storing it in a database.
  4. 제1항에 있어서,According to claim 1,
    상기 관측 자료를 자료 동화에 적용할 수 있도록 전처리하는 단계는,Pre-processing so that the observation data can be applied to data assimilation,
    상기 관측 자료에 포함되는 상기 융해층의 고도를 수치예보모델의 격자에 내삽하는 단계; 및Interpolating the altitude of the melting layer included in the observation data into a grid of a numerical prediction model; And
    상기 융해층의 고도에 해당하는 공간정보에 대한 상기 수치예보모델의 배경장 변수를 추출하는 단계를 포함하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법.And extracting background field variables of the numerical prediction model for spatial information corresponding to the altitude of the melting layer.
  5. 제4항에 있어서,The method of claim 4,
    상기 관측 자료를 자료 동화에 적용할 수 있도록 전처리하는 단계는,Pre-processing so that the observation data can be applied to data assimilation,
    1000hPa 내지 1hPa 구간의 20 개의 연직 표준 고도 배열에서 750hPa 내지 550hPa 구간에 연직으로 4 개의 배열을 추가하여 총 24 개의 연직 표준 고도 배열을 생성하는 단계; Generating a total of 24 vertical standard altitude arrays by adding four arrays vertically in a range of 750 hPa to 550 hPa in 20 vertical standard altitude arrays of 1000 hPa to 1 hPa sections;
    상기 연직 표준 고도 배열에 상기 융해층의 고도를 내삽하여 관측 자료 변수를 추출하는 단계; 및Extracting observational data variables by interpolating the elevation of the melt layer in the vertical standard altitude arrangement; And
    상기 초단기예보모델의 배경장에서 상기 관측 자료 변수와 동일한 변수를 상기 배경장 변수로 추출하는 단계를 포함하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법.And extracting a variable identical to the observed data variable from the background field of the very short-term forecast model as the background field variable.
  6. 제1항에 있어서,According to claim 1,
    상기 관측 자료를 초단기예보모델의 배경장 간에 3차원 변분 자료 동화 방식을 적용하여 분석 증분을 산출하는 단계는,The step of calculating an analysis increment by applying a three-dimensional variable data assimilation method between the background fields of the very short-term forecast model and the observed data
    외부반복순환과 내부반복순환을 통해 수학식 상기 관측 자료 및 상기 초단기예보모델의 배경장에 대한 비용함수 및 기울기를 산출하는 단계; 및Calculating a cost function and a slope for the background data of the observation data and the ultra-short-term forecasting model through equations of external iteration and internal iteration; And
    상기 기울기 값을 이용하여 상기 관측 자료가 적용되기 전의 상기 초단기예보모델의 초기장과 상기 관측 자료가 적용되었을 때의 상기 초단기예보모델의 초기장 간의 분석 증분을 산출하는 단계를 포함하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법.Using the gradient value, calculating an analysis increment between the initial field of the early short-term forecast model before the observation data is applied and the initial field of the early short-term forecast model when the observation data is applied based on the early short-term forecast model Radar melting layer elevation data assimilation method.
  7. 제1항에 있어서,According to claim 1,
    상기 분석 증분을 이용하여 초단기예보모델의 초기장을 생성하고 강수를 예측하는 단계는,The step of generating an initial field of the very short-term forecasting model and predicting the precipitation using the analysis increments,
    상기 분석 증분을 상기 초단기예보모델의 초기장의 연직 온도 프로파일에 적용하는 단계를 포함하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법.And applying the analysis increment to a vertical temperature profile of an initial field of the very short-term forecasting model.
  8. 제1항 내지 제7항 중 어느 하나의 항에 따른 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 방법을 수행하기 위한, 컴퓨터 프로그램이 기록된 컴퓨터로 판독 가능한 기록 매체.A computer readable recording medium having a computer program recorded thereon for performing the method for assimilation of the advanced data of the radar fusion layer based on the ultra-short-term forecasting model according to any one of claims 1 to 7.
  9. 기상 레이더 시스템으로부터 레이더 관측 자료를 수집하여 융해층의 고도 및 온도를 포함하는 관측 자료를 산출하는 관측 자료 수집부;An observation data collection unit that collects radar observation data from a weather radar system and calculates observation data including altitude and temperature of the melting layer;
    상기 관측 자료를 자료 동화에 적용할 수 있도록 전처리하는 관측 자료 전처리부;An observation data pre-processing unit pre-processing the observation data to be applied to data assimilation;
    상기 관측 자료를 초단기예보모델의 배경장 간에 3차원 변분 자료 동화 방식을 적용하여 분석 증분을 산출하는 변분 자료 동화부; 및A variable data assimilation unit which calculates an analysis increment by applying a 3D variable data assimilation method between the background fields of the very short-term forecast model; And
    상기 분석 증분을 이용하여 초단기예보모델의 초기장을 생성하고 강수를 예측하는 초기장 예측부를 포함하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치.A radar melting layer elevation data assimilation device based on an ultra-short forecast model including an initial field prediction unit for generating an initial field of an early short-term forecast model and predicting precipitation using the analysis increment.
  10. 제9항에 있어서,The method of claim 9,
    상기 관측 자료 수집부는,The observation data collection unit,
    상기 레이더 관측 자료에서 밝은 띠를 탐지하고, 상기 밝은 띠를 상기 융해층의 고도로 산출하며, 상기 융해층의 고도와 레윈죤데 자료를 비교하여 상기 융해층 온도를 산출하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치.A radar fusion layer based on an ultra-short-term forecast model that detects a bright band from the radar observation data, calculates the bright band as the altitude of the fusion layer, and calculates the fusion layer temperature by comparing the altitude and the Lewinsonde data of the fusion layer. Advanced data assimilation device.
  11. 제9항에 있어서,The method of claim 9,
    상기 관측 자료 수집부는,The observation data collection unit,
    상기 관측 자료를 버퍼 형태로 처리하여 데이터베이스에 저장하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치.An advanced radar fusion layer elevation data assimilation device that processes the observation data in a buffer and stores it in a database.
  12. 제9항에 있어서,The method of claim 9,
    상기 관측 자료 전처리부는,The observation data pre-processing unit,
    상기 관측 자료에 포함되는 상기 융해층의 고도를 수치예보모델의 격자에 내삽하고, 상기 융해층의 고도에 해당하는 공간정보에 대한 상기 수치예보모델의 배경장 변수를 추출하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치.A radar based on the short-term forecast model that interpolates the altitude of the melting layer included in the observation data into a grid of a numerical prediction model and extracts background field variables of the numerical prediction model for spatial information corresponding to the altitude of the melting layer. A device for assimilation of high-level materials in the fusion layer.
  13. 제12항에 있어서,The method of claim 12,
    상기 관측 자료 전처리부는,The observation data pre-processing unit,
    1000hPa 내지 1hPa 구간의 20 개의 연직 표준 고도 배열에서 750hPa 내지 550hPa 구간에 연직으로 4 개의 배열을 추가하여 총 24 개의 연직 표준 고도 배열을 생성하고, 상기 연직 표준 고도 배열에 상기 융해층의 고도를 내삽하여 관측 자료 변수를 추출하며, 상기 초단기예보모델의 배경장에서 상기 관측 자료 변수와 동일한 변수를 상기 배경장 변수로 추출하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치.A total of 24 vertical standard altitude arrays are generated by adding 4 arrays vertically in a range of 750 hPa to 550 hPa in 20 vertical standard altitude arrays of 1000 hPa to 1 hPa, and the altitude of the melt layer is interpolated to the vertical standard altitude array. A radar fusion layer elevation data assimilation device based on a very short-term forecast model that extracts observation data variables and extracts the same variables as the background field variables from the background field of the very short-term forecast model.
  14. 제9항에 있어서,The method of claim 9,
    상기 변분 자료 동화부는,The variable data assimilation unit,
    외부반복순환과 내부반복순환을 통해 수학식 상기 관측 자료 및 상기 초단기예보모델의 배경장에 대한 비용함수 및 기울기를 산출하고, 상기 기울기 값을 이용하여 상기 관측 자료가 적용되기 전의 상기 초단기예보모델의 초기장과 상기 관측 자료가 적용되었을 때의 상기 초단기예보모델의 초기장 간의 분석 증분을 산출하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치.Equation through the external repetition cycle and the internal repetition cycle calculates the cost function and slope of the observation data and the background field of the very short-term forecast model, and uses the slope value to calculate the initial short-term forecast model before the observation data is applied. A radar fusion layer elevation data assimilation device based on an early short-term forecast model that calculates an analysis increment between the initial field and the initial field of the early short-term forecast model when the observation data are applied.
  15. 제9항에 있어서,The method of claim 9,
    상기 초기장 예측부는,The initial field prediction unit,
    상기 분석 증분을 상기 초단기예보모델의 초기장의 연직 온도 프로파일에 적용하는 초단기예보모델 기반의 레이더 융해층 고도 자료 동화 장치.A radar fusion layer elevation data assimilation device based on the ultrashort prediction model that applies the analysis increments to the vertical temperature profile of the initial field of the ultrashort prediction model.
PCT/KR2019/002464 2018-11-05 2019-03-04 Method for assimilating radar melting layer elevation data on basis of ultra-short forecast model, and recording medium and apparatus for performing same WO2020096141A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
KR1020180134527A KR102023613B1 (en) 2018-11-05 2018-11-05 Method of radar melting level data assimilation based on um vdaps, computer readable medium and apparatus for performing the method
KR10-2018-0134527 2018-11-05

Publications (1)

Publication Number Publication Date
WO2020096141A1 true WO2020096141A1 (en) 2020-05-14

Family

ID=68067491

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/KR2019/002464 WO2020096141A1 (en) 2018-11-05 2019-03-04 Method for assimilating radar melting layer elevation data on basis of ultra-short forecast model, and recording medium and apparatus for performing same

Country Status (2)

Country Link
KR (1) KR102023613B1 (en)
WO (1) WO2020096141A1 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111595293A (en) * 2020-05-29 2020-08-28 山东交通学院 Surface deformation monitoring method and system based on multi-source monitoring data fusion
CN115391685A (en) * 2022-10-27 2022-11-25 南京信息工程大学 Quality control method, system, storage medium and equipment for wind field assimilation of scatterometer
CN117214979A (en) * 2023-10-08 2023-12-12 中国气象局广州热带海洋气象研究所(广东省气象科学研究所) Phased array radar data assimilation method, device and equipment

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111505645A (en) * 2020-04-26 2020-08-07 上海眼控科技股份有限公司 Radar cloud chart data display method and device, computer equipment and storage medium
CN113704695B (en) * 2021-08-25 2024-03-01 厦门市气象台(厦门市海洋气象台、海峡气象开放实验室) Initial value small disturbance method suitable for regional numerical mode set simulation prediction
CN115270405B (en) * 2022-06-22 2024-01-16 中国气象局广州热带海洋气象研究所(广东省气象科学研究所) Convection scale set forecasting method and system based on multisource and multisype disturbance combination
CN115617820B (en) * 2022-12-19 2023-04-28 广东省气象公共服务中心(广东气象影视宣传中心) Deep learning data set manufacturing method for quantitative precipitation estimation of position-related radar

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011027546A (en) * 2009-07-24 2011-02-10 Toshiba Corp Weather radar system and rainfall rate calculation method for the same
KR20120024258A (en) * 2010-09-06 2012-03-14 전자부품연구원 Transparent conductive nano particle of core-shell structure and ink including the same
KR101255966B1 (en) * 2012-03-21 2013-04-23 대한민국(기상청장) Method and system for detecting bright band using three dimensional radar reflectivity

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20170069396A (en) * 2015-12-11 2017-06-21 한국외국어대학교 연구산학협력단 Very short range microscale weather forecast model for real-time forecasting
KR101758234B1 (en) * 2016-08-09 2017-07-14 한국과학기술정보연구원 Apparatus for data assimilation and method for data assimilation
KR101826713B1 (en) * 2017-03-16 2018-03-22 경북대학교 산학협력단 Apparutus, method for forecasting weather and computer readable medium for performing the method

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2011027546A (en) * 2009-07-24 2011-02-10 Toshiba Corp Weather radar system and rainfall rate calculation method for the same
KR20120024258A (en) * 2010-09-06 2012-03-14 전자부품연구원 Transparent conductive nano particle of core-shell structure and ink including the same
KR101255966B1 (en) * 2012-03-21 2013-04-23 대한민국(기상청장) Method and system for detecting bright band using three dimensional radar reflectivity

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
KOREA METEOROLOGICAL INSTITUTE: "An Analysis of Research Trend on Rainfall Forecasting Method Using Radar Data Assimilation of High Resolution Numerical Model. National Assembly Library.", July 2017 (2017-07-01) *
YOON. SUNCHANG: "Development of COMBased Quantitative Forecasting System for Long-term Rainfall", WATER RESOURCES MANAGEMENT CENTER,KOREA WATER RESOURCES CORPORATION, 24 December 2012 (2012-12-24), pages 1 - 244 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111595293A (en) * 2020-05-29 2020-08-28 山东交通学院 Surface deformation monitoring method and system based on multi-source monitoring data fusion
CN115391685A (en) * 2022-10-27 2022-11-25 南京信息工程大学 Quality control method, system, storage medium and equipment for wind field assimilation of scatterometer
CN115391685B (en) * 2022-10-27 2023-03-24 南京信息工程大学 Quality control method, system, storage medium and equipment for wind field assimilation of scatterometer
CN117214979A (en) * 2023-10-08 2023-12-12 中国气象局广州热带海洋气象研究所(广东省气象科学研究所) Phased array radar data assimilation method, device and equipment
CN117214979B (en) * 2023-10-08 2024-03-22 中国气象局广州热带海洋气象研究所(广东省气象科学研究所) Phased array radar data assimilation method, device and equipment

Also Published As

Publication number Publication date
KR102023613B1 (en) 2019-09-20

Similar Documents

Publication Publication Date Title
WO2020096141A1 (en) Method for assimilating radar melting layer elevation data on basis of ultra-short forecast model, and recording medium and apparatus for performing same
Stroeve et al. Evaluation of the MODIS (MOD10A1) daily snow albedo product over the Greenland ice sheet
WO2020022624A1 (en) Method for calculating power generation amount reflecting farm environment and detailed technology for predicting solar farm weather, and real-time solar power generation amount prediction system using the method
WO2014073884A1 (en) Numerical simulation system and numerical simulation method for atmospheric flow by computational fluid dynamics
CN109631776B (en) Automatic measurement method for icing thickness of high-voltage transmission line conductor
WO2012087387A2 (en) Thermal powerline rating and clearance analysis using thermal imaging technology
US20200117839A1 (en) Design support apparatus, design support method and design support learned model creation apparatus for photovoltaic power generation facilities
CN112116573A (en) High-precision infrared image anomaly detection method and system
Murphy et al. Cloud-to-ground lightning warnings using electric field mill and lightning observations
Guo et al. Detecting and positioning of wind turbine blade tips for uav-based automatic inspection
Zhang et al. Cloud motion tracking system using low-cost sky imager for PV power ramp-rate control
Montanya et al. Potential use of space-based lightning detection in electric power systems
WO2022265200A1 (en) Method for correcting reflectivity using dual polarization variable-based bright band detection result, and recording medium and device for performing same
JP6951639B2 (en) Image analysis device, image analysis method and image analysis program
CN112542800A (en) Method and system for identifying transmission line fault
JP3229566B2 (en) Lightning prediction device
WO2014081171A1 (en) Apparatus and method for calculating wind load
Yang et al. Analyzing of cloud macroscopic characteristics in the Shigatse area of the Tibetan Plateau using the total-sky images
KR102522685B1 (en) Apparatus and Method for detecting galloping power lines
CN112326039B (en) Photovoltaic power plant patrols and examines auxiliary system
CN114167234A (en) Insulator aging detection method based on hyperspectrum of unmanned aerial vehicle
CN113744196A (en) Real-time engineering construction monitoring method and system
Huang et al. Research on recognition and location method of insulator in infrared image based on deep learning
WO2020013433A1 (en) High-accuracy and high-speed wireless positioning method and apparatus
CN116071368B (en) Insulator pollution multi-angle image detection and fineness analysis method and device

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19882087

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19882087

Country of ref document: EP

Kind code of ref document: A1