A system and method for correcting bias in simulation data of a regional climate model using meteorological observation data from the Korean Peninsula.

Abstract

This invention provides a bias correction system and method for simulation data of regional climate models using meteorological observation data. [Solution] The bias correction system includes a preprocessing unit 100 that converts the format and grid of simulation data from a regional climate model that simulates the climate of the Korean Peninsula region with that of meteorological observation data of the Korean Peninsula and interpolates the simulation data to the positions of observation points; a bias correction unit 200 that corrects temperature data and precipitation data in the simulation data using observation data; and a calculation verification unit 300 that verifies the correction results by comparing the temporal and spatial distributions of the temperature data and precipitation data corrected by the bias correction unit with the temporal and spatial distributions of the observation data.

Description

Technical Field

[0001] The present invention relates to a bias correction system and method, and more specifically, to a bias correction system and method for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula.

Background Art

[0002] According to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (hereinafter referred to as "IPCC") published in 2021, compared with before the Industrial Revolution (1850 - 1900), the global average temperature in the first two decades of the 21st century (2001 - 2020) has been reported to have risen by 0.99°C. Under such an environment of global warming, not only the whole world but also in South Korea, abnormal meteorological phenomena have occurred. For example, in August 2018, in Seoul, the capital of South Korea, the temperature was observed at 39.6°C, and in Hongcheon in Gangwon-do, South Korea, it was observed at 41.0°C, etc. While rewriting the observation record of the daily maximum temperature, many heatstroke patients occurred. Also, in August 2020, there was a heavy rain in the Somjin (Yeongsan) River basin in Jeollanam-do, South Korea, which rewrote the observation record of the daily precipitation of 361.3 mm per day and caused flood damage. The emission of greenhouse gases by humans still continues, and it is clear that future warming will be more serious than it is now. For this reason, at present, in order to effectively respond in the social and economic fields in the future warming environment, an accurate future prediction regarding the climate of the Korean Peninsula is required.

[0003] Many conventional studies have used simulation data from global climate models (GCMs) to predict future climate. However, the horizontal resolution of GCMs is over 100 km, making them unsuitable for simulating the climate of detailed regions such as the Korean Peninsula. To address this problem, it is conceivable to use regional climate models (RCMs) that apply regional elements to boundary conditions generated from GCM simulation data to produce high-resolution simulation data. For the systematic production of RCM simulation data and efficient data management, the World Climate Research Program (WCRP) launched the Coordinated Regional Climate Downscaling Experiment (CORDEX). This project is being implemented on various continents, and is also being carried out for the East Asian region, including the Korean Peninsula (Coordinated Regional Climate Downscaling Experiment in East Asia (CORDEX-EA)). In 2014, CORDEX-EA first produced simulation data for the East Asian region with a resolution of 50 km, and more recently, in 2020, data with an improved resolution of 25 km (CORDEX-EA Phase 2) was produced.

[0004] However, despite improvements in resolution, RCM simulation data still exhibits bias compared to observational data. This bias arises from uncertainties in the RCM's physical parameterization process, boundary conditions calculated from GCMs, and topographic data. In particular, the Korean Peninsula has a complex topography, which tends to result in significant bias in RCMs for temperature and precipitation. Such biases can lead to underestimations or overestimations of climate indices based on specific thresholds, and can distort future climate predictions due to erroneous forecasts of seasonal variations. Distorted future climate predictions may make it difficult to establish appropriate socioeconomic measures in the Korean Peninsula region. [Overview of the project] [Problems that the invention aims to solve]

[0005] The present invention was created to solve the aforementioned problems and aims to provide a bias correction system and method for simulation data of a regional climate model using meteorological observation data, which corrects the bias of a regional climate model using meteorological observation data from the Korean Peninsula. This system includes a temperature correction unit that corrects the bias of temperature data by linear scaling (LS) and variance scaling (VS), and a precipitation correction unit that corrects the bias of precipitation data by power transform (PT) and linear scaling. By including these components, it is possible to calculate simulation data of a regional climate model that reflects the climatic characteristics of the Korean Peninsula, thereby contributing to accurate climate predictions for the Korean Peninsula. Furthermore, by providing accurate climate prediction information, it is possible to contribute to the establishment of appropriate social and economic countermeasures.

[0006] Another objective of the present invention is to provide a bias correction system and method for simulation data of a regional climate model using meteorological observation data from the Korean Peninsula, which can enhance the reliability of the correction process by including a verification process for the results of the correction process.

[0007] The technical problems of the present invention are not limited in any way to those described above, and other technical problems not mentioned should be clearly understandable to an ordinary person from the following description. [Means for solving the problem]

[0008] To achieve the above objective, the bias correction system for simulation data of a regional climate model using meteorological observation data from the Korean Peninsula, according to the features of the present invention, A bias correction system, A preprocessing unit that converts the format and grid of simulation data from a regional climate model that simulates the climate of the Korean Peninsula region with that of meteorological observation data of the Korean Peninsula, and interpolates the simulation data to the locations of observation points. A bias correction unit that corrects the temperature data and precipitation data in the simulation data using the observation data, A calculation verification unit verifies the correction results by comparing the temporal and spatial distributions of temperature data and precipitation data corrected in the bias correction unit with the temporal and spatial distributions of the observation data. Includes, The bias correction unit is A temperature correction unit sequentially processes linear scaling (LS) to correct the monthly climate values ​​of the simulation data using a correction coefficient to make them the same as the monthly climate values ​​of the observational data, and then variance scaling (VS) to correct the monthly standard deviation of the simulation data using a correction coefficient to make it the same as the monthly standard deviation of the observational data, thereby correcting the bias of the temperature data. A precipitation correction unit performs a power transform (PT) to correct the mean and variance of the distribution using a nonlinear power function, and then sequentially processes the linear scaling to correct the bias of the precipitation data. Its structural features include the inclusion of [a specific element].

[0009] Preferably, the pre-processing unit is The masking unit performs masking to exclude ocean data from the aforementioned simulation data, using the masking data used for the integration of the regional climate model. The masking unit includes a re-gridding unit that converts the grid shape of the simulation data that has been masked from a curved coordinate system to a rectangular coordinate system, To correct bias for each observation point, an interpolation unit is used to interpolate the simulation data, which has been converted to the Cartesian coordinate system, to the observation points using a linear interpolation method (bilinear interpolation). It may include.

[0010] More preferably, the re-lattice portion is The grid shape may be transformed using a first-order conservative remapping method.

[0011] More preferably, the pre-processing unit is The system may further include a data conversion unit that converts the aforementioned observation data into a pre-configured file format that includes a two-dimensional variable in the form of an array of (time, observatory).

[0012] Preferably, the power conversion used in the precipitation correction unit is A constant may be calculated such that the coefficient of variation calculated from the power function of the simulation data is the same as the coefficient of variation calculated from the power function of the observation data, and the bias of the precipitation data may be corrected using the calculated constant.

[0013] Preferably, the calculation verification unit is For temporal distribution, you can compare average values, and for spatial distribution, you can compare Taylor plots.

[0014] A method for correcting the bias of simulation data of a regional climate model using meteorological observation data from the Korean Peninsula, according to the features of the present invention for achieving the above objective, is: A bias correction method in which each step is performed in a computer, (1) A preprocessing step of converting the format and grid of simulation data from a regional climate model that simulates the climate of the Korean Peninsula region and meteorological observation data of the Korean Peninsula, and interpolating the simulation data to the locations of observation points, (2) A bias correction step in which temperature data and precipitation data in the simulation data are corrected using the observation data, (3) A calculation result verification step of comparing the temporal and spatial distributions of the temperature data and precipitation data corrected in the bias correction step with the temporal and spatial distributions of the observation data to verify the correction result; including the bias correction step includes (2-1) After performing linear scaling (LS) for correction using a correction coefficient that makes the monthly climate value of the simulation data the same as the monthly climate value of the observation data, variance scaling (VS) is sequentially processed using a correction coefficient that makes the monthly standard deviation of the simulation data the same as the monthly standard deviation of the observation data to correct the bias of the temperature data, which is a temperature correction step; (2-2) After performing power transform (PT) to correct the mean and variance of the distribution using a non-linear power function, linear scaling is sequentially processed to correct the bias of the precipitation data, which is a precipitation correction step; characterized in its configuration by including the above.

[0015] Preferably, the preprocessing step includes (1-1) A masking step of performing masking to exclude ocean data from the simulation data, but performing masking using the masking data used for the integration of the regional climate model; (1-2) A remeshing step of converting the grid shape of the simulation data masked in the masking step from a curvilinear coordinate system to a rectangular coordinate system; (1-3) An interpolation step of interpolating the simulation data converted to the rectangular coordinate system to the observation points using bilinear interpolation for bias correction for each observation point; and may include the above.

[0016] The power transform used in the precipitation correction step is Calculate a constant so that the coefficient of variation calculated from the power function of the simulation data is the same as the coefficient of variation calculated from the power function of the observed data, and bias correction of the precipitation data may be performed using the calculated constant.

[0017] Further, the present invention may provide a computer program stored in a computer-readable recording medium for causing a computer to execute a method for correcting bias of simulation data of a regional climate model using meteorological observation data of the Korean Peninsula.

Effects of the Invention

[0018] According to the present invention as described above, the bias of the regional climate model is corrected using meteorological observation data of the Korean Peninsula, and a temperature correction unit that corrects the bias of temperature data by linear scaling (LS) and variance scaling (VS), and a precipitation correction unit that corrects the bias of precipitation data by power transform (PT) and linear scaling. By including these, simulation data of a regional climate model reflecting the climate characteristics of the Korean Peninsula can be calculated, which can contribute to accurate prediction of the climate of the Korean Peninsula, and by providing accurate climate prediction information, it can be used to establish appropriate social and economic countermeasures.

[0019] Furthermore, according to the present invention, by including a verification process for the calculation results of the correction process, the reliability of the correction process can be enhanced.

Brief Description of the Drawings

[0020] [Figure 1] It is a diagram showing the configuration of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to an embodiment of the present invention. [Figure 2]This is an example of a NetCDF file after the observation data has been converted in the data conversion unit of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention. [Figure 3] These are examples of (a) the product before masking and (b) the result of the masking process performed by the masking unit of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention (unit: K). [Figure 4] These are examples of (a) pre-relattice and (b) resulting products of the relattice process processed by the relattice section of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention (unit: K). [Figure 5] These are examples of (a) the pre-interpolation and (b) the resulting product of the interpolation process performed by the interpolation unit of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention (unit: °C). [Figure 6] This figure shows the detailed bias correction process of the bias correction unit in a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention. [Figure 7] This is an example of the VS process processed by the temperature correction unit of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention, and shows (a) ASOS observation data of the Korean Peninsula, (b) before the VS process, and (c) an example of the spatial temperature distribution of the result of the VS process (unit: °C). [Figure 8] This is an example of the PT process processed by the precipitation correction unit of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention, and shows (a) observation data of ASOS on the Korean Peninsula, (b) before the PT process, and (c) an example of the spatial distribution of precipitation as a result of the PT process (unit: mm). [Figure 9]This is an example showing the time distribution verification result of the calculation verification unit of the bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention. [Figure 10] This is an example showing the verification results of the spatial distribution of the calculated data verification unit of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention. [Figure 11] This diagram shows the flow of a method for correcting the bias of simulation data for a regional climate model using meteorological observation data from the Korean Peninsula, according to one embodiment of the present invention. [Figure 12] This figure shows the details of the preprocessing step (S100) in a bias correction method for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention. [Modes for carrying out the invention]

[0021] Embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. The advantages and features of the present invention, as well as the methods for achieving them, should become even clearer with reference to the embodiments described in detail below in conjunction with the accompanying drawings. However, the technical idea of ​​the present invention is not limited in any way to the embodiments disclosed below and can be embodied in a variety of different forms. The following embodiments are merely provided to complete the technical idea of ​​the present invention and to fully inform those who are ordinaryly skilled in the art to which the present invention pertains, and the technical idea of ​​the present invention is defined only by the scope of the claims. Throughout the specification, the same reference numerals refer to the same components.

[0022] Furthermore, unless otherwise specifically noted or clearly inconsistent with the context, all terms used in this disclosure, including technical and scientific terms, have the same meaning as those commonly understood by a person of ordinary skill in the art to which the present invention pertains. Generally used, dictionary-defined terms shall not be interpreted in their ideal or overly formal sense unless explicitly defined in this application. Terms used herein are used solely to describe embodiments and are not intended to limit the present invention. In this specification, singular expressions include plural expressions unless clearly otherwise stated in the context.

[0023] The phrases "comprises" and / or "comprising" used in this specification do not preclude the existence or addition of one or more other components, steps, operations and / or elements from the components, steps, operations and / or elements referred to.

[0024] Figure 1 shows the configuration of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention. As shown in Figure 1, the bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention, may include a preprocessing unit 100, a bias correction unit 200, and a calculated data verification unit 300. The bias correction system may be a computer device or server system including one or more processors and memory and storage devices that are communicatively connected to the processors. Each component shown in Figure 1 may represent a functional module that is realized when a computer program stored in the memory or storage device is executed by the processor.

[0025] The preprocessing unit 100 is a component realized by the processor of the bias correction system executing a preprocessing program stored in memory. To solve the problem that direct comparison and correction are not possible because the format and grid of the RCM simulation data and the meteorological observation data are different, and that the observation points and the grid points of the simulation data do not match, the processor can perform calculations to convert the format and grid of the simulation data of a regional climate model that simulates the climate of the Korean Peninsula region and the meteorological observation data of the Korean Peninsula, and can perform calculations to interpolate the simulation data to the positions of the observation points. This has the effect of generating consistent base data for subsequent bias correction calculations. As shown in Figure 1, the preprocessing unit 100 may include a masking unit 120, a re-grid unit 130 and an interpolation unit 140, and may further include a data conversion unit 110. The examples disclosed herein utilize daily average temperature and precipitation data from the Automated Synoptic Observing System (ASOS) at 86 locations on the Korean Peninsula and from the HadGEM3-RA RCM calculated under the CORDEX-EA Phase 2 project.

[0026] The data conversion unit 110 is a component that functions when a data conversion module incorporated into the preprocessing program is executed by the processor, and the processor can perform calculations to convert the observation data into a pre-configured file format that includes a two-dimensional variable in the array format of (time, observation station). In other words, the data conversion unit 110 performs calculations to convert meteorological observation data for the convenience of the bias correction process by the processor. Meteorological observation data from automated synoptic weather observation systems, automated weather stations (AWS), etc. can be used as observation data. Meteorological observation data from ASOS and AWS can be downloaded from the Japan Meteorological Agency's weather data public portal, and the data conversion unit 110 can download variables for daily average temperature, daily maximum temperature, daily minimum temperature, and daily precipitation. For the downloaded observation data, the data conversion unit 110 can perform calculations to convert it into a NetCDF file format that includes a two-dimensional variable in the array format of (time, observation station) by the processor. The NetCDF file format can store information about time and observation station in the form of coordinate variables. Therefore, by using NetCDF format data, tasks such as extracting time and observation stations during the bias correction process can be performed efficiently. Figure 2 shows an example of a NetCDF file after the observation data has been converted in the data conversion unit 110 of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention.

[0027] The masking unit 120 is a component that functions when a masking module incorporated into the preprocessing program is executed by the processor. The processor performs a masking operation to exclude ocean data from the simulation data, and the masking operation can be performed using the masking data used for the integration of the regional climate model. Such masking is intended to minimize the influence of the ocean grid on land grids adjacent to coastal areas. It is most preferable to use the masking data used for the integration of each RCM in the masking process. Assuming that masking data is used in which the value for land areas is 1 and the value for ocean areas is 0, the masking operation can be performed by dividing the original RCM data by the masking data. That is, the original values ​​of the RCM simulation data should be input to grids where the value after division by the masking data is 1, and missing value processing should be performed on grids where the value after division by the masking data is 0. Figure 3 shows examples of (a) the data before masking and (b) the result of the masking process performed by the masking unit 120 of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention (unit: K).

[0028] The re-grid unit 130 is a component that functions when the re-grid module incorporated into the pre-processing program is executed by the processor. The processor can perform calculations to convert the grid configuration of the simulation data masked in the masking unit 120 from a curvilinear coordinate system to a rectangular coordinate system. Converting the curvilinear coordinate system, which is the grid configuration of RCM simulation data, to a rectangular coordinate system is convenient because the calculation and visualization work is complicated. The re-grid unit 130 can use files with the coordinate format of the rectangular coordinate system directly input, or reanalysis data such as global atmospheric reanalysis data (ERA5) (European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis data version 5).

[0029] More specifically, the re-grid unit 130 can perform calculations to convert the data into a grid shape using a first-order conservative remapping method by the processor. The first-order conservative remapping method is used when performing a re-grid process on data that has undergone a masking process, and by using this method, it is possible to minimize the loss of data that may occur during the re-grid process around the ocean grid that has been treated as missing values. Figure 4 shows examples of (a) data before re-gridging and (b) the result of the re-grid process processed by the re-grid unit 130 of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention (unit: K).

[0030] The interpolation unit 140 is a component that functions when an interpolation module incorporated into the preprocessing program is executed by the processor, and can perform calculations to interpolate the simulation data converted to a Cartesian coordinate system to the observation points using a linear interpolation method (bilinear interpolation) in order to perform bias correction for each observation point by the processor. This is an operation performed in order to perform the bias correction process for each observation point. Figure 5 shows examples of (a) before interpolation and (b) the result of the interpolation process processed by the interpolation unit 140 of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention (unit: °C).

[0031] The bias correction unit 200 is a component realized by the processor of the bias correction system executing a bias correction program stored in memory. In order to solve the problem that RCM simulation data contains systematic biases compared to observational data, which reduces the reliability of future climate predictions, the processor can perform calculations to correct temperature data and precipitation data in the simulation data using observational data. As a result, highly reliable corrected climate data that reflects the climate characteristics of the Korean Peninsula can be physically generated. As shown in Figure 1, the bias correction unit 200 may include a temperature correction unit 210 and a precipitation correction unit 220. Figure 6 is a diagram showing the detailed bias correction process of the bias correction unit 200 in a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention. Table 1 summarizes the abbreviations used in the description of the bias correction unit 200.

[0032] [Table 1]

[0033] Below, the temperature correction unit 210 and the precipitation correction unit 220 will be described in detail based on Figure 6 and Table 1.

[0034] The temperature correction unit 210 is a component that functions when the temperature correction module incorporated into the bias correction program is executed by the processor. The processor performs a linear scaling (LS) operation to correct the monthly climate values ​​of the simulation data using a correction coefficient to make them the same as the monthly climate values ​​of the observational data. Then, it sequentially processes a variance scaling (VS) operation to correct the monthly standard deviation of the simulation data using a correction coefficient to make it the same as the monthly standard deviation of the observational data, thereby correcting the bias of the temperature data. That is, as shown in Figure 6, in the bias correction process for temperature data, the processor can perform an LS operation followed by a VS operation. The LS method can make the monthly climate values ​​of the RCM simulation data the same as the monthly climate values ​​of the observational data. In order to calculate the monthly climate values, the daily data for each year during the set correction reference period can be integrated into a single daily data set. At this time, since the correction reference period is the current period, past experimental data of the RCM can be used. After searching for the monthly time point (index) for the integrated daily data, the monthly climate values ​​are calculated for the daily data for each month. Next, using the calculated lunar climate values, the following correction coefficients are used to make the lunar climate values ​​in the RCM simulation data the same as the observed data values. The file TIFF2026082799000003.tif844 can be calculated. The correction factor is calculated monthly, and the LS calculation can be performed using the calculated correction factor as follows. Assuming that the correction factor does not change over time, the same correction factor can be used for simulation data of future scenarios. By using the LS calculation, the bias that causes the monthly average value of the simulation data to appear different from the monthly average value of the observed data can be reduced.

[0035]

number

[0036]

number

[0037] The LS-calculated data can be stored by the processor in NetCDF file format so that it can be used for VS calculations. Next, in order to perform a VS calculation to correct for bias in variance, the lunar climate values ​​of the data corrected by the aforementioned LS calculation are set to 0. That is, the following equations 3 and 4 can be executed.

[0038]

number

[0039]

number

[0040] Subsequently, a calculation is performed to correct for variance. The variance correction is calculated by dividing the standard deviation of the observed data by the standard deviation of the RCM simulation data, using the following correction coefficients. Using JPEG2026082799000008.jpg1629, the following can be performed as shown in equations 5 and 6.

[0041]

number

[0042]

number

[0043] Finally, the coefficients used in Equations 3 and 4 JPEG2026082799000011.jpg722, By adding JPEG2026082799000012.jpg720 to the variance-corrected data, the lunar climate values ​​are restored (Equations 7 and 8). Using the VS operation, the problem that the lunar standard deviation still does not match the observed data can be resolved.

[0044]

number

[0045]

number

[0046] Figure 7 shows an example of the VS process processed by the temperature correction unit 210 in a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention, and (a) observation data of ASOS of the Korean Peninsula, (b) before the VS process, and (c) an example of the spatial temperature distribution of the result of the VS process (unit: °C).

[0047] The precipitation correction unit 220 is a component that functions when the precipitation correction module incorporated into the bias correction program is executed by the processor. The processor performs a power transform (PT) operation to correct the mean and variance of the distribution using a nonlinear power function, and then sequentially performs linear scaling operations to correct the bias of the precipitation data. The power transform used in the precipitation correction unit 220 calculates a constant such that the coefficient of variation calculated from the power function of the simulation data is the same as the coefficient of variation calculated from the power function of the observation data, and the bias of the precipitation data can be corrected using the calculated constant.

[0048] More specifically, for precipitation data, a nonlinear approach is used. The mean and bias can be corrected using the PT method, which employs the power function of JPEG2026082799000015.jpg712. This process can be constructed independently for the current period and future periods, and either can be performed first. The PT method uses a constant b such that the coefficient of variation calculated from the power functions of the observed data and the RCM simulation data are identical. m This can be done by determining the coefficient of variation. The coefficient of variation is the value obtained by dividing the monthly standard deviation by the monthly climate value, and it exists for each month.

[0049]

number

[0050] In this case, the calculation of monthly climate values ​​and monthly standard deviations can be performed using daily data from three months centered around the target month, rather than using only the daily data for the target month. For example, if it is necessary to calculate the monthly climate values ​​and monthly standard deviation for January, daily data from December, January, and February can be used.

[0051] Along with this, the constant b m This can be calculated by the processor repeatedly executing numerical analysis algorithms such as bisection. For continuous functions, If the image is JPEG2026082799000017.jpg721, then the solution c of the function exists between intervals a and b. This method utilizes the intermediate value theorem, which states that there must always be at least one instance of JPEG2026082799000018.jpg714. In other words, by repeatedly performing a bisection method that sets intervals in a direction that reduces the difference in function values ​​(the difference in the coefficient of variation between the observed data and the RCM simulation data) corresponding to the interval, an accurate function solution b can be obtained. mWe are trying to calculate the exact function b by independently performing the bisection method for all months and grids until the difference in coefficients of variation becomes sufficiently small (less than 0.0001). m It is possible to calculate this.

[0052] b calculated in this way m Using this, bias correction can be performed as shown in equations 10 and 11 below.

[0053]

number

[0054]

number

[0055] Subsequently, using the LS method, the lunar climate values ​​are made identical to the observed lunar climate values, as shown in equations 12 and 13 below.

[0056]

number

[0057]

number

[0058] Figure 8 shows an example of the PT process processed by the precipitation correction unit 220 of the bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention, and (a) observation data of ASOS of the Korean Peninsula, (b) before the PT process, and (c) an example of the spatial distribution of precipitation as a result of the PT process (unit: mm).

[0059] The calculation verification unit 300 is a component realized by the processor of the bias correction system executing a calculation verification program stored in memory. Since it is necessary to quantitatively verify the reliability of the correction data calculated by the bias correction unit 200 and to technically compare and verify the degree of improvement before and after correction, the processor can perform calculations to verify the correction results by comparing the temporal and spatial distributions of the temperature and precipitation data corrected by the bias correction unit 200 with the temporal and spatial distributions of the observed data. More specifically, the calculation verification unit 300 can perform calculations by the processor to compare average values ​​for the temporal distribution and Taylor plots for the spatial distribution. Furthermore, the verification of the correction results can be performed on a large number of RCM simulation data (ensembles). In this invention, simulation data of six types of RCMs (HadGEM3-RA, CCLM, WRF, GRIMs, RegCM, SNU-MM5) integrated under the CORDEX-EA Phase 2 project are disclosed as examples. In this case, the RCM was dynamically refined from three GCMs (UKESM, HadGEM2-AO, MPI-ESM-LR, and GFDL-ESM2M). Meanwhile, in this example, daily average temperature and precipitation data from 86 ASOS observation sites, including those in South and North Korea, were used.

[0060] The calculation verification unit 300 can perform calculations using the processor to verify the time distribution by comparing the diurnal climate values ​​of the ensemble with the observed data. After calculating the diurnal climate values ​​for each ensemble, it can generate a time-series time distribution map. At this time, the time series can display the observed data and the diurnal climate values ​​of each ensemble.

[0061] Figure 9 shows an example of the verification results of the time distribution of the calculated data verification unit 300 of the bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention. (a) shows a comparative time series of daily climate values ​​for temperature data, and (b) shows a comparative time series of daily climate values ​​for precipitation data. In Figure 9, the observation data is shown by thick black lines, each RCM ensemble is shown grouped by the corresponding GCM, and ensembles that have not been bias corrected (bias-corrected ensembles) are shown by dotted lines (solid lines). Regarding temperature, the temperature of the Korean Peninsula shows seasonal variation with a maximum value in August, which is summer, and a minimum value in January, which is winter (Figure 9(a)). Regarding the time distribution of temperature, it was confirmed that the ensemble before bias correction had a negative bias (-4.6 to -0.8°C). However, it can be confirmed that the bias in the time distribution is sufficiently eliminated after applying the VS method.

[0062] Next, regarding precipitation, precipitation on the Korean Peninsula shows seasonal variation with maximum values ​​in summer and minimum values ​​in winter. At this time, a bimodal temporal distribution is observed, with the first maximum precipitation value appearing in July-August and the second maximum value in August-September (Figure 9(b)). Regarding the temporal distribution of precipitation, it was confirmed that the ensemble before bias correction had a dry bias (-3.12~1.04 mm / day) from July to October and a wet bias (0.07~1.73 mm / day) in other months. Furthermore, it was confirmed that the bimodal temporal distribution was not simulated in most ensembles. However, it can be confirmed that the PT method corrects not only the dry bias and wet bias, but also the bimodal temporal distribution.

[0063] On the other hand, spatial distribution can be verified by calculating statistical values ​​(spatial correlation coefficient and standard deviation) between the observed data and the ensemble. The closer the spatial correlation coefficient and standard deviation are to 1, the more similar the spatial distribution of the ensemble is to the observed data. Taylor diagrams are used to display the statistical values, and Taylor diagrams can be created for each ensemble.

[0064] Figure 10 shows an example of the verification results of the spatial distribution of the calculation verification unit 300 of the bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention. (a) shows a Taylor plot for temperature data, and (b) shows a Taylor plot for precipitation data. In Figure 9, each RCM ensemble is shown grouped by the corresponding GCM, and ensembles that have not been bias corrected (bias-corrected ensembles) are indicated by the symbol × (○).

[0065] For temperature, the spatial correlation coefficients of the ensembles before bias correction were found to be 0.9–0.95, and the standard deviations were 1.06–1.23 (Figure 10(a)). This means that while most ensembles successfully simulated the spatial distribution of temperature as shown by observations, they had a slight bias. However, the VS method was found to have eliminated the spatial simulation bias for temperature for all ensembles. For precipitation, the spatial correlation coefficients of the ensembles before bias correction were found to be 0.23–0.72, and the spatial standard deviations were 0.34–1.63 (Figure 10(b)). This means that most ensembles did not adequately simulate the spatial distribution of precipitation as shown by observations. However, the PT method was found to have eliminated the spatial simulation bias for precipitation for all ensembles.

[0066] This document discloses a description and example of the verification process for the results of a bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention. As a result, it was confirmed that the bias correction process corrected the bias so that the climate characteristics of the Korean Peninsula were sufficiently reflected in the RCM simulation data. Therefore, it is expected that the reliability of future climate forecast information for the Korean Peninsula can be improved by using RCM scenario data that has undergone the bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to one embodiment of the present invention. Ultimately, it is expected that the provision of accurate climate forecast information will be useful in establishing appropriate social and economic countermeasures.

[0067] Figure 11 is a diagram showing the flow of a bias correction method for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention. As shown in Figure 11, the bias correction method for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention is a bias correction method in which each step is executed on a computer, and may include: a preprocessing step (S100) in which the format and grid of the simulation data of a regional climate model that simulates the climate of the Korean Peninsula region are converted and the meteorological observation data of the Korean Peninsula are interpolated into the positions of observation points; a bias correction step (S200) in which temperature data and precipitation data in the simulation data are corrected using observation data; and a calculation verification step (S300) in which the temporal distribution and spatial distribution of the temperature data and precipitation data corrected in the bias correction step are compared with the temporal distribution and spatial distribution of the observation data to verify the correction result.

[0068] Figure 12 is a diagram showing the details of the preprocessing step (S100) in a bias correction method for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention. As shown in Figure 12, the preprocessing step (S100) of the bias correction method for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to one embodiment of the present invention may include: a masking step (S110) in which ocean data is masked from the simulation data, but the masking is performed using masking data used in the integration of the regional climate model; a re-grid step (S120) in which the grid shape of the simulation data that has been masked in the masking step is converted from a curvilinear coordinate system to a rectangular coordinate system; and an interpolation step (S130) in which the simulation data converted to a rectangular coordinate system is interpolated to the observation points using a bilinear interpolation method in order to correct the bias for each observation point.

[0069] Detailed information regarding each step has been sufficiently explained in conjunction with the bias correction system for simulation data of the regional climate model using meteorological observation data of the Korean Peninsula, as described above, so a detailed explanation will be omitted.

[0070] Embodiments of the present invention can also be implemented in the form of a computer program stored on a computer-readable recording medium for causing a computer to execute on a computer a bias correction method for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula according to the above embodiment of the present invention. Here, the computer-readable medium is any available medium accessible by a computer and can include all volatile and non-volatile media, isolated and non-isolated media. The computer-readable medium can also include both computer storage media and communication media. Computer storage media include all volatile and non-volatile, isolated and non-isolated media implemented by any method or technique for storing information such as computer-readable instructions, data structures, program modules, or other data. Communication media typically include computer-readable instructions, data structures, program modules, or other data such as modulated data signals like carrier waves, or other transmission mechanisms and include any information transmission medium.

[0071] As described above, according to the present invention, by correcting the bias of the regional climate model using meteorological observation data from the Korean Peninsula, and including a temperature correction unit 210 that corrects the bias of temperature data by linear scaling (LS) and variance scaling (VS), and a precipitation correction unit 220 that corrects the bias of precipitation data by power transform (PT) and linear scaling, it is possible to calculate simulation data for the regional climate model that reflects the climate characteristics of the Korean Peninsula. This contributes to accurate climate prediction for the Korean Peninsula and helps in formulating appropriate social and economic countermeasures by providing accurate climate prediction information. Furthermore, by including a verification process for the calculations of the correction process, the reliability of the correction process can be increased.

[0072] While embodiments of the present invention have been described above based on the attached drawings, those with ordinary skill in the art to which the present invention belongs should understand that the present invention can be implemented in other specific forms without changing the technical idea or essential features of the present invention. Therefore, the embodiments described above should be understood to be illustrative in all respects and not limiting. [Explanation of symbols]

[0073] 100 Pre-processing 110 Data conversion unit 120 Masking section 130 Regrid section 140 Interpolation part 200 Bias Correction Section 210 Temperature Correction Unit 220 Precipitation correction section 300 Calculation Verification Department S100 Pretreatment step S110 Masking Step S120 Re-lattice step S130 Insertion step S200 Bias Correction Step S300 Calculation Verification Step

Claims

1. A bias correction system, A preprocessing unit (100) converts the format and grid of simulation data from a regional climate model that simulates the climate of the Korean Peninsula region and meteorological observation data of the Korean Peninsula, and interpolates the simulation data to the locations of observation points. A bias correction unit (200) corrects the temperature data and precipitation data in the simulation data using the observation data, A calculation verification unit (300) verifies the correction results by comparing the temporal and spatial distributions of temperature data and precipitation data corrected in the bias correction unit (200) with the temporal and spatial distributions of the observation data, Includes, The bias correction unit (200) is A temperature correction unit (210) sequentially processes linear scaling (LS) to correct the monthly climate values ​​of the simulation data using a correction coefficient to make them the same as the monthly climate values ​​of the observation data, and then variance scaling (VS) to correct the monthly standard deviation of the simulation data using a correction coefficient to make it the same as the monthly standard deviation of the observation data, thereby correcting the bias of the temperature data. A precipitation correction unit (220) performs a power transform (PT) to correct the mean and variance of the distribution using a nonlinear power function, and then sequentially processes the linear scaling to correct the bias of the precipitation data. A bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, characterized by including the following:

2. The preprocessing unit (100) is A masking unit (120) performs masking to exclude ocean data from the simulation data, using the masking data used for the integration of the regional climate model. The masking unit (120) includes a re-gridging unit (130) that converts the grid shape of the simulation data that has been masked from a curved coordinate system to a rectangular coordinate system, An interpolation unit (140) interpolates the simulation data, which has been converted to the Cartesian coordinate system, to the observation points using a linear interpolation method (biline interpolation) in order to correct the bias for each observation point, A bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, as described in claim 1, characterized by including the following:

3. The aforementioned re-lattice section (130) is A bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, as described in claim 2, characterized by transforming the grid shape using a first-order conservative remapping method.

4. The preprocessing unit (100) is A bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to claim 2 or 3, further comprising a data conversion unit (110) that converts the observation data into a pre-set file format that includes a two-dimensional variable in the form of an array of (time, observation station).

5. The power conversion used in the precipitation correction unit (220) is, A bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to any one of claims 1 to 3, characterized by calculating a constant such that the coefficient of variation calculated from the power function of the simulation data is the same as the coefficient of variation calculated from the power function of the observation data, and correcting the bias of the precipitation data using the calculated constant.

6. The aforementioned calculation verification unit (300) A bias correction system for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, characterized in that the temporal distribution is compared with average values ​​and the spatial distribution is compared with Taylor plots, as described in any one of claims 1 to 3.

7. A bias correction method in which each step is performed in a computer, (1) A preprocessing step (S100) in which the simulation data of a regional climate model that simulates the climate of the Korean Peninsula region is converted in format and grid to the meteorological observation data of the Korean Peninsula, and the simulation data is interpolated to the location of the observation point, (2) A bias correction step (S200) in which temperature data and precipitation data in the simulation data are corrected using the observation data, (3) A calculation verification step (S300) in which the temporal and spatial distributions of temperature data and precipitation data corrected in the bias correction step are compared with the temporal and spatial distributions of the observation data to verify the correction results, Includes, The bias correction step (S200) is performed as follows: (2-1) A temperature correction step in which the bias of the temperature data is corrected by sequentially performing linear scaling (LS) to correct the monthly climate values ​​of the simulation data using a correction coefficient to make them the same as the monthly climate values ​​of the observation data, and then performing variance scaling (VS) to correct the monthly standard deviation of the simulation data using a correction coefficient to make it the same as the monthly standard deviation of the observation data, (2-2) A precipitation correction step in which a power transform (PT) is performed to correct the mean and variance of the distribution using a nonlinear power function, and then the linear scaling is processed sequentially to correct the bias of the precipitation data, A method for correcting the bias of simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, characterized by including the following:

8. The aforementioned pretreatment step (S100) is (1-1) Masking is performed to exclude ocean data from the simulation data, but the masking step (S110) is performed using the masking data used for the integration of the regional climate model, (1-2) A re-grid step (S120) in which the grid shape of the simulation data that has been masked in the masking step is converted from a curved coordinate system to a rectangular coordinate system, (1-3) In order to correct the bias for each observation point, an interpolation step (S130) is performed in which the simulation data converted to the Cartesian coordinate system is interpolated to the observation points using a linear interpolation method (biline interpolation), A method for correcting the bias of simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, as described in claim 7, characterized by including the following:

9. The power transformation used in the precipitation correction step is A method for correcting the bias of simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, according to claim 7, characterized by calculating a constant such that the coefficient of variation calculated from the power function of the simulation data is the same as the coefficient of variation calculated from the power function of the observation data, and correcting the bias of the precipitation data using the calculated constant.

10. A computer program stored on a computer-readable recording medium for causing a computer to execute a bias correction method for simulation data of a regional climate model using meteorological observation data of the Korean Peninsula, as described in any one of claims 7 to 9.

Images