KR101826713B1 - Apparutus, method for forecasting weather and computer readable medium for performing the method - Google Patents

Abstract

Disclosed are a rainfall prediction apparatus, a method thereof, and a computer-readable recording medium having a computer program recorded therein to provide the same. The rainfall prediction apparatus generates a rainfall prediction field by inputting, to a numerical weather prediction model, observation data obtained from an observation device including at least one weather radar, and comprises: an observation data collection unit for collecting observation data measured from the observation device; a background field generation unit for generating a background field by predicting a rainfall after a predetermined time on the basis of the numerical weather prediction model; and an observation data assimilation unit for generating a rainfall prediction field by moving the observation data to the background field, wherein the observation data assimilation unit assimilates a mixing ratio of an aqueous body corresponding to a reflectivity value of a predetermined first threshold value to the background field when a reflectivity value of the observation data is less than the first threshold value and a reflectivity value of the background field is greater than the first threshold value. Accordingly, the present invention can improve accuracy of the rainfall prediction field by assimilating the observation data to the background field of the numerical weather prediction model and by adding a separate non-rainfall data assimilation step to a region in which the numerical weather prediction model wrongly or excessively predicts a rainfall from regions observed as a non-rainfall region in the observation data.

Description

TECHNICAL FIELD [0001] The present invention relates to a precipitation prediction apparatus, a precipitation prediction method, and a computer-readable recording medium on which a computer program for providing the same is recorded.

The present invention relates to a precipitation prediction apparatus, a method, and a computer-readable recording medium on which a computer program for providing the same is recorded. More particularly, the present invention relates to a precipitation prediction apparatus, A prediction device, a method, and a computer program for providing the same.

The weather radar can be used not only to understand the mesoscale phenomenon by observing the intensity and movement of the atmospheric precipitation in three dimensions but also to input into the precipitation data assimilation system of the numerical forecast model to generate the precipitation prediction field.

The rainfall data assimilation system is based on the Weather Research and Forecasting (WRF) model and the data assimilation system of the fifth generation Pennsylvania State University-National Center for Atmospheric Research Mesoscale Model (MM5).

The water assimilation method using the above-described precipitation data assimilation system is based on the warm-water process in order to separate the total water mixing ratio obtained by the variational method into the humidity and the increment of each water body with the total water mixing ratio as the control variable There is an observation operator and an adjoint model that includes microphysical processes. There is also a tangent-linear model and an adjoint model of the Kessler warm-up microphysical parametrization process that can be used to assimilate radar data through four-dimensional variational methods.

However, this conventional precipitation data assimilation system is not only applied to some data assimilation systems but also requires a large amount of database based on the background data of observation data and numerical forecasting models, and does not consider the humidity of the liquid phase and the solid phase There is a problem.

Korean Patent Registration No. 10-1642226

One aspect of the present invention is to assimilate the observed data into the background field from the numerical prediction model so that when the reflectivity value of the observed data is smaller than the minimum value of the reflectivity that can be observed by the observer and the reflectivity value of the background field is larger than the minimum value, And a rainfall prediction device for generating a precipitation prediction field by moving to a background field at a mixture ratio of the opponent corresponding to the precipitation prediction field.

A precipitation prediction apparatus according to an aspect of the present invention is a precipitation prediction apparatus for generating a precipitation prediction field by inputting observation data obtained from an observation instrument including at least one weather radar into a numerical prediction model, A background data generation unit for generating a background data by predicting the precipitation after a predetermined time based on the numerical prediction model, and a data generating unit for generating the data by assimilating the observation data into the background data, Wherein the observation data moving unit is adapted to correspond to the reflectivity value of the first threshold value when the reflectivity value of the observation data is smaller than a predetermined first threshold value and the reflectivity value of the background field is larger than the first threshold value, And the observation data visualization unit that assimilates the mixing ratio of the water bodies to the background field.

Meanwhile, if the observation area is spaced apart from the precipitation area and the maximum observation distance of the observation instrument by a predetermined distance and the reflection value of the background field is larger than a predetermined second threshold, the reflection data of the first threshold value A relative humidity operator may be used to assimilate the mixing ratio of the opponent corresponding to the background area to the background field and reduce the relative humidity of the observation area by a predetermined ratio.

The first threshold value may be a minimum value of the reflectivity that can be observed by the observation device, and the second threshold value may be larger than the first threshold value.

In addition, the observation data moving unit may define the relative humidity operator by the following equation according to the temperature of the background field.

이다. Here, q ^o is the relative humidity operator, q _{s, water} is the mixing ratio of water bodies in the saturated state, q _{s, ice} is the mixing ratio of the water bodies in the saturated state, The temperature of the field,

to be.

,

이고, 여기서

는 수증기에 대한 기체상수에 대한 건조공기의 기체상수의 비로 0.622 이고,

는 융해열로

이고,

는 수증기의 기체상수로

이고, T와 p 는 각각 상기 배경장의 기온과 기압이다.Also,

,

, Where

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

Is a heat of fusion

ego,

Is the gas constant of water vapor

T and p are the temperature and the atmospheric pressure of the background field, respectively.

,

이고, 여기서

는 수증기에 대한 기체상수에 대한 건조공기의 기체상수의 비로 0.622 이고,

는 승화열로

이고,

는 수증기의 기체상수로

Also,

,

, Where

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

As a sublimation heat

ego,

Is the gas constant of water vapor

T and p are the temperature and the atmospheric pressure of the background field, respectively.

Also, the observation data moving unit converts the reflection value of the observation data into a unit value according to Equation (1) below, and converts the converted reflection value into rain, wet eyes, Hail, and dry eye, and the reflectance by the mixing ratio of each of the viscous bodies is converted into the mixing ratio of each viscous body by using the following expression (3), and the converted number The mixing ratio of the upper body can be assimilated to the background.

[Equation 1]

Here, Z is the _dB reflectivity in dBZ unit value before conversion of observations, Z _e is mm ⁶ m ^- is the reflectance value of the ^third unit of the transformed measurement data.

&Quot; (2) "

Here, Z (q _rain ), Z (q _wetsnow ), Z (q _drysnow ), and Z (q _graupel ) are the reflectivities of rain, wet snow, dry snow, to be.

 &Quot; (3) "

Here, ρ is the air density of the background, and q _rain , q _wetsnow , q _drysnow , and q _graupel are the mixing ratios of rain, wet snow, dry snow, and hail hail, respectively.

A precipitation prediction method according to another aspect of the present invention is a precipitation prediction method for generating a precipitation prediction field by inputting observation data obtained from an observation instrument including at least one weather radar into a numerical forecast model, And a precipitation prediction field is generated by assimilating the observation data into the background field to generate a precipitation prediction field, wherein the reflectance value of the observation data is calculated by: The mixing ratio of the opaque body corresponding to the reflectivity value of the first threshold value is assimilated to the background field when the reflectance value of the background field is smaller than the predetermined first threshold value and larger than the first threshold value.

On the other hand, in order to assemble the observation data to the background field to generate a precipitation prediction field, it is preferable that the observation area is spaced apart from the precipitation area and the maximum observation distance of the observation instrument by a predetermined distance, Further comprising applying a relative humidity operator to assimilate the mixing ratio of the opaque body corresponding to the reflectivity value of the first threshold value to the background field and to reduce the relative humidity of the observation area by a predetermined ratio .

The first threshold value may be a minimum value of the reflectivity that can be observed by the observation device, and the second threshold value may be larger than the first threshold value.

Further, the relative humidity operator can be defined by the following equation according to the temperature of the background field.

이다. Here, q ^o is the relative humidity operator, q _{s, water} is the mixing ratio of water bodies in the saturated state, q _{s, ice} is the mixing ratio of the water bodies in the saturated state, The temperature of the field,

to be.

,

이고, 여기서

는 수증기에 대한 기체상수에 대한 건조공기의 기체상수의 비로 0.622 이고,

는 융해열로

이고,

는 수증기의 기체상수로

Also,

,

, Where

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

Is a heat of fusion

ego,

Is the gas constant of water vapor

T and p are the temperature and the atmospheric pressure of the background field, respectively.

,

이고, 여기서

는 수증기에 대한 기체상수에 대한 건조공기의 기체상수의 비로 0.622 이고,

는 승화열로

이고,

는 수증기의 기체상수로

이고, T와 p 는 각각 상기 배경장의 기온과 기압이다.Also,

,

, Where

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

As a sublimation heat

ego,

Is the gas constant of water vapor

T and p are the temperature and the atmospheric pressure of the background field, respectively.

In order to assimilate the mixture ratio of the opaque material to the background field, the reflectance value of the observation data is converted into a unit value according to Equation (1) below, and the converted reflectance value is converted into the reflectance value of the background field The reflectivity of each of the viscous bodies is defined as the sum of the reflectivities by the mixing ratio of the viscous material, which is one of rain, wet eyes, hail hail, and dry eyes depending on the temperature. And the mixing ratio of the converted water bodies can be assimilated into the background field

[Equation 1]

Here, Z is the _dB reflectivity in dBZ unit value before conversion of observations, Z _e is mm ⁶ m ^- is the reflectance value of the ^third unit of the transformed measurement data.

&Quot; (2) "

Here, Z (q _rain ), Z (q _wetsnow ), Z (q _drysnow ), and Z (q _graupel ) are the reflectivities of rain, wet snow, dry snow, to be.

 &Quot; (3) "

Here, ρ is the air density of the background, and q _rain , q _wetsnow , q _drysnow , and q _graupel are the mixing ratios of rain, wet snow, dry snow, and hail hail, respectively.

Yet another aspect of the present invention includes a computer-readable recording medium on which a computer program is recorded, for providing the above-described precipitation prediction method.

According to the present invention described above, the observation data are assimilated into the background field of the numerical prediction model, and the numerical prediction model among the regions observed in the nasal water region in the observation data is a nasal water The accuracy of the precipitation prediction field can be improved by adding the data acquisition process.

1 is a block diagram of a precipitation prediction apparatus according to an embodiment of the present invention.
2 is a view showing an example of a point of an observation instrument and an observation area of a radar for obtaining observation data.
3 is a view showing an example of preprocessed observation data.
4 is a diagram showing an example of a domain and a topographic altitude of a numerical prediction model to be used for background field generation.
FIG. 5 is a diagram showing the ratio of the mixing ratio of the obtained precipitation prediction field to the mixing ratio of the water vapor when the precipitation data are assimilated into the background field.
6 is a flowchart of a precipitation prediction method according to another embodiment of the present invention.
FIG. 7 is a flowchart showing details of the process of assimilating the observation data shown in FIG. 6 into the background field.
8 is a view showing an actual case for confirming the effect of the nasal water data assimilation.
FIGS. 9 to 17 are views showing the effect of the nasal water data assimilation. FIG.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention, and the manner of achieving them, will be apparent from and elucidated with reference to the embodiments described hereinafter in conjunction with the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Is provided to fully convey the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout the specification.

The terminology used herein is for the purpose of illustrating embodiments and is not intended to be limiting of the present invention. In the present specification, the singular form includes plural forms unless otherwise specified in the specification. It is noted that the terms " comprises "and / or" comprising ", as used herein, do not exclude the presence or addition of one or more other elements, steps and operations.

1 is a block diagram of a precipitation prediction apparatus according to an embodiment of the present invention.

Referring to FIG. 1, the precipitation prediction apparatus 100 according to the present embodiment is a device for generating a precipitation prediction field by moving observational data obtained from an observation instrument 10 to a background field from a numerical prediction model. For this purpose, the precipitation prediction apparatus 100 includes an observation data collection unit 110, a background genetic part 120, and an observation data animation unit 130.

The observation data collection unit 110 collects observation data measured from the observation instrument 10.

Meanwhile, the observation instrument 10 may be at least one of an automated weather station (AWS), a wind profiler, and weather radar data, and may perform other weather analysis to provide observation data It can be included in the observation instrument 10 of this embodiment. For example, observation data can be collected from various kinds of observation instruments 10 positioned as shown in FIG. Referring to FIG. 2, the purple circle indicates the positions of about 690 AWS among the 900 AWS operated by the Korea Meteorological Administration, and the blue rhomb is the position of the vertical wind gauge equipment, (Chulwon, Munsan, Gangneung, Wonju, Ulsan, Chunwolryeong, Gunsan, Changwon, and West Sea maritime comprehensive base). The red circle indicates the position of the weather radar, and the maximum observation area is indicated by a circle centering on the weather radar. The weather radar may be an S-band Doppler radar operated by the Korea Meteorological Administration or an S-band dual-polar radar operated by the Ministry of Land, Transport and Maritime Affairs. In FIG. 2, KWK, GDK, , Jindo (JNI), Guddoksan (PSN), Gosan (GSN), Seongsan (SSP), Biseulsan (BSL) and Sobaeksan (SBS) weather radar are displayed.

The observation data collection unit 110 performs a process of preprocessing the observation data to be input to the observation data animation unit 130.

The observation data collecting unit 110 extracts only the observation data necessary for data assimilation from the observation data, and inputs the observation data to the observation data assimilation unit 130. For example, the observation data collection unit 110 may extract only the ground pressure, the sea surface pressure, the 10-minute average wind speed, the wind direction, the air temperature, and the dew point temperature among the observation variables in the AWS data and input them to the observation data moving unit 130 have.

The observation data collection unit 110 can check the quality of the weather radar data. The observation data collection unit 110 can remove the non-vapor echo such as the terrestrial echo and the echo echo using the FUZZY algorithm, and can negate the error by using the error value of each radar reflectivity value for the real case. The observation data collecting unit 110 can treat the same region as a region having no precipitation with respect to a minimum reflectivity value that is lower than -15 dBZ that can be observed by the weather radar. In addition, the observation data collection unit 110 excludes the value from being used for data assimilation when the difference between the gaze speed and the observation value in the background field from the numerical prediction model is equal to or greater than the threshold value, Only the visual velocity of the data with a reflection value of 0 dBZ or more can be used.

The observation data collection unit 110 performs pre-processing on the quality-checked weather radar data. In order for the observation data to be assimilated to the background field from the numerical forecasting model in the observation data acquisition unit 130, the observation data input to the observation data acquisition unit 130 should be uniformly distributed as in the background field. For this purpose, the observation data collecting unit 110 grids the observation data of the weather radar of the spherical coordinate system, and the horizontal grid of the numerical prediction model can be used. For example, if the WRF model is used as the numerical forecasting model, the observation data collecting unit 110 is configured to have a vertical coordinate of 100 m from the ground up to 2 km above the ground, and the vertical layer spacing of the WRF model in the free atmosphere at 2 to 15 km It is possible to construct a vertical coordinate of 250 m intervals.

The preprocessing process in the observation data collection unit 110 will be described in detail as follows. First, in order to consider the radar beam width, the observation data collection unit 110 assumes that the width of the radar beam is a normal distribution shape, weighting the distance between the center of the grid and the reflectivity of the Plane Position Indicator (PPI) Interpolate into the lattice of the numerical forecast model. In addition, the PPI plane of the radar and the elevation angle is vertically interpolated with an inverse distance weighting. Since the same phenomenon is considered as a scalar value for each radar, unlike the gaze speed, the spectral data collection unit 110 needs to reduce the reflectivity of each grid radar to a single value, To make a composite sheet. This synthesis field is used as an input data of the observation data moving unit 130, and an example of the pre-processed weather radar data through the above-described process is as shown in FIG. Referring to FIG. 3, each circle is an observation data to be inputted into the data acquisition process, and the color of each circle is the intensity of reflection by precipitation, and the intensity of the precipitation can be known by referring to the color bar on the right side. On the other hand, the ivory and gray circles represent the data of the nasal passages, while the passages of the nectar passages represent the nasal passages adjacent to the precipitation areas. Because the area of the ivory circle is adjacent to the precipitation area even in the nasal water area, the control of the relative humidity of the negative may affect the precipitation intensity in the other precipitation area. Therefore, unlike the nasal water area of gray circle, only the mixing ratio of the water bodies in the area is adjusted to perform the assimilation of nasal water data.

로 설정되고, 지면 자료의 수평해상도는

로 설정될 수 있다. 또한, 상기 수치예보모델은 한국에 영향을 미치는 강수사례를 모의하기 위해 3 개의 도메인의 등지격자계로 구성될 수 있다. 상기 도메인 정보와 각 도메인마다 사용한 미세물리 과정은 표 1 과 같다.The background generation unit 120 generates a background field by predicting the precipitation after a predetermined time based on the numerical prediction model. At this time, the numerical forecasting model can be a weather research and forecasting (WRF) model, which is a medium-scale high-resolution weather model. For example, the WRF model 3.7.1 version as shown in FIG. 4 can be used. Referring to FIG. 4, the numerical prediction model is based on the NCEP Climate Forecast System Reanalysis (CFSR) version 2 with 38 vertical layers, and its initial and boundary conditions are as follows.

And the horizontal resolution of the ground data is set to

Lt; / RTI > In addition, the numerical forecasting model can be composed of three domain back-ground systems to simulate the precipitation cases affecting Korea. Table 1 shows the domain information and the microphysical processes used for each domain.

[Table 1]

The observation data moving unit 130 assimilates the observation data from the observation data collection unit 110 to the background field generated by the background genetic unit 120 to generate a precipitation prediction field.

), 비 혼합비(

), 얼음 혼합비(

), 눈 혼합비(

), 싸락우박 혼합비(

), 연직 속도를 통제 변수로 사용하는 cloud control variable option을 추가로 사용할 수 있다. 또한, 상기 WRFDA 는 서로 다른 두 시간의 기상장으로부터 동일한 시간까지 예보하여 두 예보장의 차이를 수치모델의 오차로 두는 NMC (Parrish and Derber, 1992)을 이용하여 배경 오차 공분산 행렬 (background error covariacne maxtrix, )을 계산한다. The observation data assimilation unit 130 of this embodiment inputs the observation data using the data assimilation system of the numerical prediction model and assimilates the input observation data into the background field from the numerical prediction model. For example, three-dimensional reanalysis data assimilation by WRFDA, a data assimilation system of WRF, can be used. Specifically, the WRFDA is a control variable associated with the atmosphere, including a stream function, an unbalanced velocity potential, an unbalanced temperature, a pseudo-relative humidity, an unbalanced air pressure surface pressure) can be used, and the cloud water mixing ratio (

), The mixing ratio (

), Ice mixing ratio (

), Snow mixture ratio

), Mixture ratio of hail hail

), And a cloud control variable option that uses the vertical speed as a control variable. In addition, the WRFDA estimates the background error covariacne maxtrix using the NMC (Parrish and Derber, 1992), which predicts the difference between the two forecasts for the same time from two different time periods, ).

In the following, we will explain in detail the observation data assimilation using this data assimilation system.

The observation data acquisition unit (130) assimilates the observation data into the background field using the water body mixture ratio assimilation method and the relative humidity assimilation method.

The water body mixing ratio assimilation method is a method of calculating the mixing ratio of the water bodies using the relational expression of the reflectance and the water body mixture ratio, and assimilating the water body mixture into the background field instead of the reflectance.

To this end, the observation data moving unit 130 first converts the reflectance in units of dBZ to the reflectance in units of mm ⁶ m ^-3 according to the following equation (1).

[Equation 1]

Here, Z is the _dB reflectivity in dBZ unit value before conversion of observations, Z _e is mm ⁶ m ^- is the reflectance value of the ^third unit of the transformed measurement data.

The converted reflectance value can be expressed as a sum of the reflectivities by mixing ratio of each of the visors such as rain, dry snow, wet snow, and hail hail as shown in Equation 2 below.

&Quot; (2) "

Where Z _e is the converted reflectivity value and Z (q _rain ), Z (q _wetsnow ), Z (q _drysnow ), and Z (q _graupel ) are respectively the ratio of rain, wet snow, dry snow, Reflectivity.

Considering that the higher the temperature of the background field is, the higher the liquid phase in the vortex and the lower the temperature, the higher the solid phase, the observation data moving part 130 calculates the conversion The reflectivity value can be defined as the sum of the reflectivities by the mixture ratio of at least one of the rain, the dry snow, the wet snow, and the loose hail.

&Quot; (3) "

That is, in the observation data moving part 130, there is an obesity of a liquid phase in a region where the background temperature (T) is higher than 5 ° C, and rain, It is assumed that rain, dry snow, and hail hail exist in the area of 5 ° C <T <0 ° C, and that dry snow and hail hail exist in the area of T ≤-5 ° C. In addition, the observation data moving part 130 linearly interpolates the reflectance by each of the water bodies according to the temperature of the background field, and distributes the reflectance so that the mixing ratio of the water body and the hull of the eyes is half.

The observation data moving unit 130 converts the reflection value defined by the sum of the reflectivities by the mixing ratio of the visceral bodies into the mixing ratio value of each visceral body by using the following expression (4).

&Quot; (4) &quot;

,

,

,

,

,

,

Here, ρ is the air density of the background field, and q _rain , q _wetsnow , q _drysnow , and q _graupel are the mixing ratios of rain, wet snow, dry snow, and hail hail, respectively.

The observation data moving unit 130 finally inputs the conversion ratio values of the converted water bodies into the data moving system as observation data instead of the reflection value.

The observation data assimilation unit 130 assimilates the observation data into the background field using the relative humidity assimilation method, which is a method of assimilating the cloud humidity when the measured reflectance value is above a certain threshold value. When it is determined that the relative humidity of the area is saturated by 100% based on the measured reflectance value, the moving part of the observation data 130 adds the relative humidity assimilation method to the above- Assimilate the data. At this time, the threshold value of the reflectivity value may be 30 dBZ.

The observation data moving unit 130 executes relative humidity assimilation by applying a relative humidity operator. Specifically, the observed data moving unit 130 defines the relative humidity operator as a saturated mixing ratio for water when the observed reflectivity value is equal to or greater than 30 dBZ as shown in Equation (5) below, and the saturated mixing ratio value for the water is Clausius -Clapeyron &lt; / RTI &gt; equation (6).

&Quot; (5) &quot;

Where q ^o _vapor is the relative humidity operator and q _{s, water} is the mixing ratio of the water body, which is the _water at saturation.

&Quot; (6) &quot;

,

,

는 수증기에 대한 기체상수에 대한 건조공기의 기체상수의 비로 0.622 이고,

는 융해열로

이고,

는 수증기의 기체상수로

이고, T와 p 는 각각 배경장의 기온과 기압이다.here

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

Is a heat of fusion

ego,

Is the gas constant of water vapor

T and p are the temperature and pressure of the background field, respectively.

That is, by inputting the relative humidity operator calculated according to the above-described process into the data assimilation system, the observed data moving part 130 can increase the humidity and the temperature of the corresponding area, thereby making the environment convective and maintainable.

5 shows the ratio of the mixing ratio of the precipitation prediction field to the mixing ratio of the water vapor obtained by the observation data acquiring unit 130 as described above. FIG. 5 (a) is a weather radar observation data at an altitude of 3 km observed on July 23, 2013, and FIG. 5 (b) It is the appearance of background from the data. 5 (a) and 5 (b), no observed precipitation was observed in the inland and eastern waters of the observation data. However, in the background section, it is known that there is a discrepancy between the observed data and the background data have. 5 (c) and 5 (d) show the comparison between the precipitation prediction field and the background field obtained after assimilating the observation data into the background field, respectively. The increment ratio of the mixing ratio and the water vapor mixture ratio Lt; / RTI &gt; In this case, when the increment is positive, it means that the value of the variable increases after data assimilation, and when it is negative, it means that the value of the variable decreases. As shown in (c) of Fig. 5, the rainfall ratio of the rainfall area which was not predicted in the background field in the west of Gyeonggi-do increased, It can be seen that the accuracy of the prediction field is improved. 5 (d), it can be seen that the positive increment is better in the region where the observed reflectivity value is 30 or more. However, all of the areas (c) and (d) of Fig. 5 did not show a sufficient increment of the sound for the area remote from the eco-area or the area where some excessive precipitation was mistakenly predicted. Therefore, it is necessary to separate precipitation data assimilation process for the region where the excessive rainfall is predicted in the nasal water area.

In view of this, the observation data moving unit 130 according to the present embodiment further includes the nasal water data moving unit 131 to separate the nasal water data according to the reflectance value of the background field Of the nasal passages are assimilated to effectively suppress the overpredicted precipitation in the nasal passages.

More specifically, in order to adjust the mixing ratio of the water body in the nasal water region, the nasal water data assimilation unit 131 determines whether the reflection value of the observation data is smaller than the predetermined first threshold value and the reflectance value of the background field is larger than the first threshold value The mixing ratio of the opaque body corresponding to the reflectivity value of the first threshold value is assimilated in the background field. Here, the first threshold value may be a minimum value of reflectivity that the observing instrument can observe, for example, -15 dBZ which is the minimum observable minimum reflectivity of the S-band weather radar. That is, if the reflectivity value of the observed data is smaller than the minimum observed reflectivity of the weather radar, the nasal water data acquisition unit 131 recognizes the corresponding region as a nasal water region, and if the reflectance value of the predicted background field in the corresponding region is smaller than the minimum observed reflectivity The mixing ratio of the water bodies corresponding to the minimum observation reflectance is input to the data acquisition system as an observation value to suppress the precipitation in the corresponding region. At this time, the mixture ratio of the viscous body inputted to the data assimilation system is calculated according to the following Equation (7), and the mixing ratio of each of the viscous bodies is calculated according to the above-described body composition ratio mathematical expression (Equations 1 to 4).

&Quot; (7) &quot;

Here, q ^o _rain , q ^o _snow , and q ^o _graupel are the mixing ratios of the rain, wet snow, dry snow, and hibernate hail input to the data assimilation system, respectively, to be.

Further, in addition to assimilating the mixing ratio of the water bodies corresponding to the reflectivity value of the first threshold value to the background field when the reflectance value of the background field is larger than the predetermined second threshold value, Apply a relative humidity operator to the observed data assimilation to reduce the relative humidity of the observed area by a certain percentage. Here, the second threshold may be greater than the first threshold, and the second threshold may be set to 20 dBZ. In other words, when the nasal water data assimilation unit 131 predicts over-precipitation in which the reflectance value predicted in the background field exceeds a certain value even though it is the nasal water area in the observation data, in addition to the adjustment of the above- Relative humidity is adjusted. This can solve the problem that precipitation can be predicted again due to strong mechanical forcing even if the relative humidity is saturated among the regions where strong precipitation is predicted in the background field.

For this, the nasal water data acquisition unit 131 first defines a relative humidity operator according to the following equation (8).

&Quot; (8) &quot;

이다. 또한,

,

이고, 여기서

는 수증기에 대한 기체상수에 대한 건조공기의 기체상수의 비로 0.622 이고,

는 융해열로

이고,

는 수증기의 기체상수로

이고, T와 p 는 각각 상기 배경장의 기온과 기압이다. 또한,

,

이고, 여기서

는 수증기에 대한 기체상수에 대한 건조공기의 기체상수의 비로 0.622 이고,

는 승화열로

이고,

는 수증기의 기체상수로

이고, T와 p 는 각각 상기 배경장의 기온과 기압이다.Here, q ^o is the relative humidity operator, q _{s, water} is the mixing ratio of water bodies in the saturated state, q _{s, ice} is the mixing ratio of the water bodies in the saturated state, The temperature of the field,

to be. Also,

,

, Where

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

Is a heat of fusion

ego,

Is the gas constant of water vapor

T and p are the temperature and the atmospheric pressure of the background field, respectively. Also,

,

, Where

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

As a sublimation heat

ego,

Is the gas constant of water vapor

T and p are the temperature and the atmospheric pressure of the background field, respectively.

That is, assuming that only the liquid state cloud exists when the background temperature of the background field is higher than 5 ° C, the nasal water data assimilation section 131 reduces the relative humidity of the corresponding area to 95% of the relative humidity of water, Assuming that only solid phase clouds are present at temperatures below -5 ° C, the relative humidity in the area is reduced to 100% of the relative humidity for ice, ie by removing only the supersaturated water vapor mixing ratio for ice. In addition, the nasal water data assimilation section 131 adjusts the relative humidity of the corresponding area in a manner that linearly increases in accordance with the background temperature when the background temperature is higher than -5 ° C and lower than 5 ° C.

On the other hand, relative humidity adjustments for these nasal passages may have an effect on the surrounding precipitation area, and may affect the outside of the observation area if the nasal passages are at the edge of the observation area. Therefore, the nasal water data animating unit 131 according to the present embodiment performs relative humidity adjustment using the relative humidity operator only when the observation region is spaced apart from the maximum observation distance of the precipitation region and the weather radar by a predetermined distance or more . For example, the nasal water data moving unit 131 can set the predetermined distance to 30 km in consideration of the fact that the distance at which the relative humidity is substantially changed by the observation data is 30 km.

6 is a flowchart of a precipitation prediction method according to another embodiment of the present invention.

6, the observation data collection unit 210 collects (210) observation data measured from an observation instrument including at least one weather radar, and calculates a precipitation after a predetermined time based on a numerical forecast model in a background field generation unit The background is generated by predicting (220). The collected observation data is inputted to a data assimilation system 230 through a quality inspection and a preprocessing process 230 and is assimilated into a background field 240. Finally, a precipitation prediction field generated 250 by assimilation of observation data into a background field is generated 250, .

Meanwhile, the detailed process in which the observation data are assimilated into the background field is shown in FIG.

Referring to FIG. 7, first, in the area 310 where the reflectance value of the observation data is predicted to be larger than the first threshold value, the precipitation data is assimilated into the background field by using the water body mixture ratio method and the relative humidity assimilation method. . Specifically, when the reflection value of the observed data is smaller than the third threshold value (350), the reflectance value of the observed data is changed to the mixture ratio of the voxel data only by using the water body mixture ratio assimilation method, (361). The concrete aerobic mixing ratio assimilation method is replaced with the description of FIG. 1 described above. In addition, if the reflectivity value of the observed data is greater than the third threshold (350), the moving part of the observation data is applied to the observation area using the relative humidity assimilation method using the relative humidity operator (362,363). The application process of the specific relative humidity assimilation method is replaced with the description of FIG. On the other hand, the first threshold value for determining whether or not the nasal water region exists may be a minimum reflectance value observable by the weather radar, for example, -15 dBZ which is the minimum observable minimum reflectivity of the S-band weather radar. In addition, the third threshold value for determining whether the relative humidity assimilation is further applied may be 30 dBZ.

The observation data assimilation part further includes the nasal water data assimilation part and performs the additional data assimilation to the nasal water area by the nasal water data assimilation part. Specifically, if the reflectance of the observation data is smaller than the first threshold value and the threshold value of the background field is larger than the first threshold value, the mixing ratio of the opponent corresponding to the first threshold value may be determined (310, 320, 331, 332). In other words, if the reflectance value of the background field predicted in the numerical prediction model is overestimated, the nasal water data assimilation section may be configured to reflect the reflectance of the first threshold value Assimilate to a water body mixing ratio. At this time, the process of converting the water body mixture ratio is the same as the above-mentioned water body mixture ratio conversion method. In addition, if the non-nasal water data moving part is spaced apart from the precipitation area and the maximum observation distance of the observation instrument by a predetermined distance or more and the reflection value of the background field is larger than the second threshold value larger than the first threshold value, (310, 320, 341, 342, 344, and 345). The relative humidity assimilation is performed by using a relative humidity operator to assimilate the relative humidity of the observing area to a corresponding water body mixing ratio. Meanwhile, the second threshold value may be 20 dBZ, and the predetermined distance may be 30 km.

According to the present invention described above, the observation data are assimilated into the background field of the numerical prediction model, and the numerical prediction model among the regions observed in the nasal water region in the observation data is a nasal water The accuracy of the precipitation prediction field can be improved by adding the data acquisition process. Hereinafter, the effect of the present invention described above will be confirmed by applying the nasal water data assimilation to the real case through FIGS. 8 to 17. FIG.

8 is a view showing an actual case for confirming the effect of the nasal water data assimilation. 8 (a) is a ground day diagram of 00 UTC on July 23, 2013, (b) is an upper layer analysis field of the corresponding time, (c) is a COMS water vapor image, 12 hours of AWS cumulative precipitation from 21 UTC on the 22nd to 09 UTC on the 23rd. According to Fig. 8, the medium-scale convective system developed in the slow stream affects the Korean peninsula and the lower jet developed at the edge of the high pressure of the North Pacific at that time. As a result of this, warm and humid air is moved to Korea, and in the north, cold and dry air in Mongolia is moved to Manchuria, stagnation wires are formed in the northern part of the Korean peninsula, It can be seen that a large amount of water vapor coagulates in a region where the wind direction changes rapidly due to the influence of the wind, and a precipitation zone is formed and the approach to the Korean peninsula can be confirmed. By this precipitation zone, .

FIG. 9 shows the result of a single observation experiment performed when only the nasal water data assimilation is performed in the background field. According to FIG. 9, the non-mixing ratio was reduced from 0.5286 g / kg to 0.0954 g / kg (FIG. 9 (a)) as a result of applying the above-described nasal water data assimilation process to the background data of the nasal water area , And the water vapor mixing ratio decreased from 12.087 g / kg to 11.931 g / kg (FIG. 9 (b)). The horizontal increment and vertical increment of the water vapor mixing ratio were 25 km and 3 km, respectively )) Can be known. 9 (c)), the radii of the horizontal increment and the vertical increment of the warm are 30 km and 3 km, respectively (Fig. 9 (f)), It can be seen that the weak cyclone rotation appears around the lattice input nasal pass information (FIG. 9 (d)).

FIG. 10 is a diagram showing a change in the precipitation prediction field when the observation data assimilation process according to the present invention is applied to the actual case of FIG. 8; FIG. 10 (a) is a composite diagram of the radar reflectivity of 00 UTC on July 23, 2013, (b) is a background field derived from a numerical forecast model, and (c) (E) and (f) show the horizontal increment of the mixing ratio and the water vapor mixing ratio of the precipitation prediction field obtained from the experiment (hereinafter, referred to as CRVRF PART) And the horizontal increment of the mixing ratio of water vapor in the precipitation prediction field obtained from the experiment in which the data are further assimilated (hereinafter referred to as CRVRF ALL). According to FIG. 10, it can be seen that the precipitation prediction field obtained from the CRVRF PART did not eliminate the mistakenly predicted precipitation on the offshore Mokpo and the East Sea, whereas the precipitation prediction field obtained on the CRVRF ALL shows the non- It can be seen that the water vapor mixing ratio is reduced.

11 to 13 are diagrams showing 1 hour cumulative precipitation and AWS cumulative precipitation for each experiment after data assimilation and 9-hour prediction. 11 to 13 (b)). In the CONVENTION experiment, radar precipitation data were further added to the AWS (Figs. 11 to 13 (a) One CRVRF PART experiment (FIGS. 11-13C) and a CRVRF ALL experiment (FIGS. 11-13D) in which the nasal passive data assimilation is additionally performed in the CRVRF PART experiment. According to FIGS. 11 to 13, CONVENTION shows that the precipitation of the Yeongseo region was not simulated at all, but also simulated precipitation up to Gyeongbuk where strong precipitation was not observed. The CRVRF PART and CRVRF ALL show relatively low precipitation It can be seen that it was well predicted. In addition, the intensity and location of the CRVRF ALL scattered in Chungbuk compared to the CRVRF PART are more in agreement with the observed data.

14 to 17 are graphs showing the results of verifying the accuracy quantitatively using the forecasting index for each experiment.

First, FIG. 14 is a graph showing a result of calculating a fractional skill score (FSS) on the results of CONVENTION, CRVRF PART, and CRVRF ALL. Here, FSS is a degree of proficiency in the rate at which a target phenomenon occurs at an adjacent point, and is calculated according to the following equation (9).

&Quot; (9) &quot;

와

는 각각 격자점 주위로 일정 반경 내에서 대상 현상이 일어나거나 및/또는 예보되는 비율의 확률값이고, N 은 검증에 사용된 총 격자점 수를 의미한다. 따라서, 이러한 FSS 에 따르면, 예보가 완벽할 경우

와

가 같아져 그 값이 1 이 되고, 반대일 경우는 0 이 된다. 도 14 에서는 검증하고자 하는 격자점 주위로 반경 6㎞ 내의 격자점 자료를 이용하여 FSS 를 계산하였다. 상술한 방식으로 각 실험결과에 FSS 를 계산한 결과, 10 dBZ 를 기준으로 한 도 14 의 (a)에 따르면, 레이더 자료를 동화한 CRVRF PART 와 CRVRF ALL 이 그렇지 않은 CONVENTION 보다 6시간 반동안 높은 점수를 나타냄을 알 수 있다. 또한, 도 14 의 (b) 및 (c) 에 따르면 25 dBZ 와 35 dBZ 에서는 레이더 자료를 동화한 CRVRF PART 와 CRVRF ALL 이 그렇지 않은 CONVENTION 보다 9 시간 예보 내내 높은 점수를 나타냄을 알 수 있다. 또한, 전체적으로 비강수자료를 동화한 CRVRF ALL 이 그렇지 않은 CRVRF PART 에 비해 높은 점수가 산출되었으며, 특히 그 특징이 비교적 높은 강수(25 dBZ ,35 dBZ)를 기준으로 하였을 때 두드러지게 나타남을 알 수 있다.here,

Wow

Is a probability value of the rate at which a phenomenon occurs and / or predicted within a certain radius around a grid point, and N is the total number of grid points used for verification. Thus, according to these FSS, if the forecast is perfect

Wow

Is equal to 1, and when it is the opposite, it becomes 0. In Fig. 14, the FSS is calculated using the lattice point data within a radius of 6 km around the lattice point to be verified. As a result of calculating the FSS for each test result in the above-described manner, according to FIG. 14 (a) based on 10 dBZ, CRVRF PART and CRVRF ALL assimilating radar data are higher than CONVENTION . &Lt; / RTI &gt; 14B and 14C, it can be seen that CRVRF PART and CRVRF ALL assimilating radar data at 25 dBZ and 35 dBZ show higher scores over the 9-hour forecast than CONVENTION without the CRVRF PART and CRVRF ALL. In addition, CRVRF ALL, which assimilates nasal water as a whole, has a higher score than CRVRF PART which does not, and it can be seen that its characteristics are prominent when compared with relatively high precipitation (25 dBZ, 35 dBZ) .

FIG. 15 is a view showing a result of calculating a false alarm ratio (FAR) in the CONVENTION, CRVRF PART, and CRVRF ALL test results. Here, the FAR can be calculated as a ratio of non-occurrence to non-occurrence observations as shown in Equation (10) below, and the closer the value is to '0', the more predictable the non-occurrence of the phenomenon.

&Quot; (10) &quot;

Here, R and F are as shown in Table 2 below.

[Table 2]

15, which is a result of calculating the FAR for each test result in the above-described manner, the CRVRF PART and the CRVRF ALL that assimilate the radar data during the 9-hour prediction period as a whole are compared with each other CONVENTION and CRVRF ALL assimilating non-nasal water data showed lower false-rate than non-CRVRF PART.

FIG. 16 is a graph showing the FSS and the 9-hour average value of the false rate in each experiment according to the reflectivity. Referring to FIG. 16, it can be seen that CRVRF PART and CRVRF ALL are both better than CONVENTION in both FSS and false rate. In addition, the FSS value of CRVRF ALL increased by 0.05 in precipitation above 15 dBZ, and the false rate decreased by 0.02 in precipitation below 35 dBZ, indicating that CRVRF ALL was better than CRVRF PART.

FIG. 17A is a graph showing the variation of the 60 minutes accumulated precipitation amount (BIAS) for each experiment, FIG. 17B is a graph showing the root mean square error (RMSE) In the graph of FIG.

First, the deviation of the 60-minute cumulative precipitation and the calculation of the mean square root error value are given by the following equations (11) and (12).

&Quot; (11) &quot;

&Quot; (12) &quot;

와

는 각각 관측된 변수와 예측된 변수이고, N 은 사용된 총 관측값의 수이며,

는 격자로 된 모델자료를 관측지점으로 내삽하여 산출된다.here,

Wow

Are the observed and predicted variables respectively, N is the number of total observations used,

Is calculated by interpolating the grid data from the model data to the observation point.

According to FIG. 17 (a), CONVENTION shows a negative deviation for a forecast time of 6 hours and underestimates the precipitation, and CRVRF PART and CRVRF ALL show a positive deviation for a forecasted time of 7 hours and overexerted precipitation. In addition, the CRVRF ALL assimilating the nasal water information shows a slight decrease in the amount of precipitation during the first 3 hours than the non CRVRF PART.

On the other hand, according to FIG. 17 (b) showing the square root deviation of the amount of precipitation, all of the experiments show similar trend during the forecasting time.

Such precipitation prediction apparatuses and methods can be implemented in the form of program instructions that can be executed through various computer components and recorded on a computer-readable recording medium. The computer-readable recording medium may include program commands, data files, data structures, and the like, alone or in combination.

The program instructions recorded on the computer-readable recording medium may be ones that are specially designed and configured for the present invention and are known and available to those skilled in the art of computer software.

Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape, optical recording media such as CD-ROMs and DVDs, 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 machine language code such as those generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware device may be configured to operate as one or more software modules for performing the processing according to the present invention, and vice versa.

While the present invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, You will understand. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive.

110: Observation data collection unit
120: Background
130: observation data moving part

Claims (11)

A precipitation prediction apparatus for generating a precipitation prediction field by inputting observation data obtained from an observation instrument including at least one weather radar into a numerical prediction model,
An observation data collection unit for collecting observation data measured by the observation equipment;
A background generating unit for generating a background field by predicting the precipitation after a predetermined time based on the numerical prediction model; And
And an observation data moving part for generating the precipitation prediction field by moving the observation data to the background field, wherein the observation data moving part updates the reflection value of the background data by using the reflection value of the observation data, the reflection value of the observation data being smaller than a predetermined first threshold value, And the observation data assimilation unit for assimilating the mixing ratio of the water bodies corresponding to the reflectivity value of the first threshold value to the background field when the water level is greater than the first threshold value. The method according to claim 1,
The observation data moving unit
When the observation area is spaced apart from the precipitation area and the maximum observation distance of the observation instrument by a predetermined distance and the reflection value of the background field is larger than a predetermined second threshold value, A relative humidity operator for assimilating the mixing ratio to the background field and reducing the relative humidity of the observation area by a predetermined ratio. 3. The method of claim 2,
Wherein the first threshold is a minimum value of reflectivity that the observing instrument can observe, and the second threshold is greater than the first threshold. 3. The method of claim 2,
The observation data moving unit
Wherein the relative humidity operator is defined according to the following equation according to the temperature of the background field.

Here, q ^o is the relative humidity operator, q _{s, water} is the mixing ratio of water bodies in the saturated state, q _{s, ice} is the mixing ratio of the water bodies in the saturated state, The temperature of the field,

to be.
Also,

,

, Where

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

Is a heat of fusion

ego,

Is the gas constant of water vapor

T and p are the temperature and the atmospheric pressure of the background field, respectively.
Also,

,

, Where

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

As a sublimation heat

ego,

Is the gas constant of water vapor

T and p are the temperature and the atmospheric pressure of the background field, respectively. The method according to claim 1,
The observation data moving unit
Transforming a reflectivity value of the observation data into unit values according to Equation 1 below,
The converted reflectivity value is defined as a sum of the reflectivities by the mixture ratio of any one of rain, wet snow, hail hail, and dry snow depending on the temperature of the background field as shown in Equation (2)
The reflectance by the mixing ratio of each viscous body is converted into the mixing ratio of each viscous body by using the following expression (3)
And a mixing ratio of the converted water bodies to the background field.
[Equation 1]

Here, Z is the _dB reflectivity in dBZ unit value before conversion of observations, Z _e is mm ⁶ m ^- is the reflectance value of the ^third unit of the transformed measurement data.
&Quot; (2) &quot;

Here, Z (q _rain ), Z (q _wetsnow ), Z (q _drysnow ), and Z (q _graupel ) are the reflectivities of rain, wet snow, dry snow, to be.
&Quot; (3) &quot;

Here, ρ is the air density of the background, and q _rain , q _wetsnow , q _drysnow , and q _graupel are the mixing ratios of rain, wet snow, dry snow, and hail hail, respectively. A precipitation prediction method for generating a precipitation prediction field by inputting observation data obtained from an observation instrument including at least one weather radar into a numerical prediction model,
Collecting observation data measured from the observation instrument,
Generating a background field by predicting precipitation after a predetermined time based on the numerical prediction model,
Wherein when a reflectivity value of the observation data is smaller than a predetermined first threshold value and a reflectivity value of the background field is greater than the first threshold value, And the mixing ratio of the opaque material corresponding to the reflectance value of the value is adapted to the background field. The method according to claim 6,
And generating the precipitation prediction field by moving the observation data to the background field,
When the observation area is spaced apart from the precipitation area and the maximum observation distance of the observation instrument by a predetermined distance and the reflection value of the background field is larger than a predetermined second threshold value, Further comprising applying a relative humidity operator to assimilate the mixing ratio to the background field and to reduce the relative humidity of the observation area by a certain percentage. 8. The method of claim 7,
Wherein the first threshold is a minimum value of reflectivity that the instrument can observe and the second threshold is greater than the first threshold. 8. The method of claim 7,
Wherein the relative humidity operator is defined by the following equation according to the temperature of the background field.

Here, q ^o is the relative humidity operator, q _{s, water} is the mixing ratio of water bodies in the saturated state, q _{s, ice} is the mixing ratio of the water bodies in the saturated state, The temperature of the field,

to be.
Also,

,

, Where

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

Is a heat of fusion

ego,

Is the gas constant of water vapor

T and p are the temperature and the atmospheric pressure of the background field, respectively.
Also,

,

, Where

Is 0.622 as a ratio of the gas constant of the dry air to the gas constant for the water vapor,

As a sublimation heat

ego,

Is the gas constant of water vapor

T and p are the temperature and the atmospheric pressure of the background field, respectively. The method according to claim 6,
To assimilate the mixture ratio of the viscous material to the background field,
Transforming a reflectivity value of the observation data into unit values according to Equation 1 below,
The converted reflectivity value is defined as a sum of the reflectivities by the mixture ratio of any one of rain, wet snow, hail hail, and dry snow depending on the temperature of the background field as shown in Equation (2)
The reflectance by the mixing ratio of each viscous body is converted into the mixing ratio of each viscous body by using the following expression (3)
And a mixing ratio of the converted water bodies is adapted to the background field.
[Equation 1]

Here, Z is the _dB reflectivity in dBZ unit value before conversion of observations, Z _e is mm ⁶ m ^- is the reflectance value of the ^third unit of the transformed measurement data.
&Quot; (2) &quot;

Here, Z (q _rain ), Z (q _wetsnow ), Z (q _drysnow ), and Z (q _graupel ) are the reflectivities of rain, wet snow, dry snow, to be.
&Quot; (3) &quot;

Here, ρ is the air density of the background, and q _rain , q _wetsnow , q _drysnow , and q _graupel are the mixing ratios of rain, wet snow, dry snow, and hail hail, respectively. A computer-readable recording medium having recorded thereon a computer program for providing a precipitation prediction method according to any one of claims 6 to 10.

Images