JP2006064609A - Weather prediction method - Google Patents

Weather prediction method Download PDF

Info

Publication number
JP2006064609A
JP2006064609A JP2004249539A JP2004249539A JP2006064609A JP 2006064609 A JP2006064609 A JP 2006064609A JP 2004249539 A JP2004249539 A JP 2004249539A JP 2004249539 A JP2004249539 A JP 2004249539A JP 2006064609 A JP2006064609 A JP 2006064609A
Authority
JP
Japan
Prior art keywords
time
data
area
weather
wide
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
JP2004249539A
Other languages
Japanese (ja)
Other versions
JP4209370B2 (en
Inventor
Tomohiro Hara
智宏 原
Ryoji Oba
良二 大場
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Mitsubishi Heavy Industries Ltd
Original Assignee
Mitsubishi Heavy Industries Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mitsubishi Heavy Industries Ltd filed Critical Mitsubishi Heavy Industries Ltd
Priority to JP2004249539A priority Critical patent/JP4209370B2/en
Publication of JP2006064609A publication Critical patent/JP2006064609A/en
Application granted granted Critical
Publication of JP4209370B2 publication Critical patent/JP4209370B2/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02ATECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Abstract

<P>PROBLEM TO BE SOLVED: To provide a weather prediction method capable of improving weather prediction accuracy, using observed values. <P>SOLUTION: The observed value is observed between a time (9 o'clock) when an initial value of GPV data is taken and a time (15 o'clock) when reception of the GPV data is finished. A weather predicting calculation, using a weather model (RAMS), is started not from 15 o'clock, but rather from 9 o'clock, and identification processing is performed by the observed value to a predicted operation value from 9 o'clock to 15 o'clock. Since identification processing is performed by using the observed value in this way, the prediction accuracy at 15 o'clock is improved, and moreover, the weather prediction accuracy after 15 o'clock including 15 o'clock is improved. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

本発明は気象予測方法に関するものである。更に詳述すると、広域気象データを基に狭域気象データを予測する際に、広域気象データを受信完了した時刻よりも時間的に前に観測された観測値を用いて同化処理することにより、狭域気象データの予測精度を向上させたものである。   The present invention relates to a weather prediction method. More specifically, when predicting narrow-area meteorological data based on wide-area meteorological data, by assimilating using observation values observed before the time when the wide-area meteorological data was received, The prediction accuracy of the narrow-area meteorological data is improved.

広域気象データを基に、狭域気象データを予測(狭域気象予測)するには、メソスケール(methoscale)気象解析モデルを用いた数値流体解析演算(CFD:Computational Fluid Dynamics)を行っている。
なお、CFDとは、計算領域を格子状に分割して、各格子点の気象要素(風速,温度,気圧,湿度等)について、気象要素の微分方程式を時間積分演算することにより気象データを得る演算手法である。
In order to predict narrow-area meteorological data (narrow-area meteorological prediction) based on wide-area meteorological data, computational fluid dynamics computation (CFD) using a mesoscale meteorological analysis model is performed.
In CFD, the calculation area is divided into a grid, and weather data is obtained by time-integrating the differential equations of the weather elements for the weather elements (wind speed, temperature, pressure, humidity, etc.) at each grid point. It is an arithmetic technique.

広域気象データとしては、例えば、気象庁から1日に2回配信されている気象GPV(Grid Point Value)データがある。このGPVデータは、日本全国を格子状(格子間隔は2km)に分割したときの格子点上の各気象要素(風速,温度,気圧,湿度等)の3時間毎の予測値が51時間にわたり示されている。なおこのGPVデータは、約6時間の配信遅れがある。例えば、ある日の午前9時を初期値とした51時間分のGPVデータ(予測データ)は、その日の午後3時(15時)に配信が完了する。   As the wide area meteorological data, for example, there is meteorological GPV (Grid Point Value) data distributed twice a day from the Japan Meteorological Agency. This GPV data shows the predicted values every 3 hours for each meteorological element (wind speed, temperature, pressure, humidity, etc.) on the grid points for 51 hours when the whole country of Japan is divided into a grid pattern (lattice interval is 2 km). Has been. This GPV data has a delivery delay of about 6 hours. For example, the distribution of 51 hours of GPV data (predicted data) with an initial value of 9:00 am on one day is completed at 3:00 pm (15:00) on that day.

またメソスケール気象モデルとしては、例えば、RAMS(Regional Atmospheric Modeling System:米国のコロラド州立大学で開発された気象解析モデル)がある。このRAMSで示されている気流場解析の基本方程式は、運動方程式、熱エネルギー方程式、水分の拡散方程式、連続の式からなり、次のような式(1)〜(6)で表される。   An example of a mesoscale weather model is RAMS (Regional Atmospheric Modeling System: a weather analysis model developed at Colorado State University, USA). The basic equations of the airflow field analysis shown in this RAMS are composed of equations of motion, thermal energy equations, moisture diffusion equations, and continuity equations, and are expressed by the following equations (1) to (6).

Figure 2006064609
Figure 2006064609

GPVデータを基に、RAMSを用いて気象要素の微分方程式を時間積分演算して、狭域気象予測するには、時間積分演算開始時刻でのGPVデータを微分方程式の初期値として取り込み、一定時毎(例えば1時間毎)のGPVデータを微分方程式の境界条件として取り込んで、RAMSで示されている微分方程式を時間積分演算する。   Based on GPV data, RAMS is used to time-integrate the differential equation of the meteorological element to predict the narrow-area weather, and the GPV data at the time integration start time is taken as the initial value of the differential equation. Each time (for example, every hour) of GPV data is taken in as a boundary condition of the differential equation, and the differential equation shown in RAMS is time-integrated.

なお配信されたGPVデータは、時間的にも空間的にも粗い気象データ(時間間隔が3時間で、格子間隔が2km)であるので、実際には、この粗いGPVデータに対して、時間内挿補間演算と空間内挿補間演算をして、時間的にも空間的にも密な(例えば、時間間隔が1時間で、格子間隔が200m)の密なGPVデータを求め、この密なGPVデータを、上述した「初期値」や「境界条件」として用いている。   The distributed GPV data is rough weather data (time interval is 3 hours and lattice interval is 2 km) in terms of time and space. Interpolation calculation and spatial interpolation calculation are performed to obtain dense GPV data that is dense in time and space (for example, the time interval is 1 hour and the lattice interval is 200 m). Data is used as the above-mentioned “initial value” and “boundary condition”.

ここで、GPVデータを基に狭域気象データを予測する、従来手法を、図6を参照して説明する。   Here, a conventional method for predicting narrow-area weather data based on GPV data will be described with reference to FIG.

図6に示すように、このGPVデータは9時を初期値とした51時間分の予測データであり、15時に配信が完了する。気象モデルとしてRAMSを用いた予測計算は、GPVデータの配信完了時(受信完了時)である15時から開始する。RAMSを用いた予測計算では、15時におけるGPVデータ(時間内挿補間演算と空間内挿補間演算をして時間的にも空間的にも密なGPVデータ)を初期値として取り込み、その後の1時間毎のGPVデータ(時間内挿補間演算と空間内挿補間演算をして時間的にも空間的にも密なGPVデータ)を境界条件として取り込み、15時から以降1時間毎の予測計算を行う。   As shown in FIG. 6, this GPV data is 51 hours of predicted data with an initial value of 9 o'clock, and distribution is completed at 15:00. Prediction calculation using RAMS as a weather model starts at 15:00 when GPV data distribution is completed (when reception is completed). In the prediction calculation using the RAMS, GPV data at 15:00 (temporal and spatial interpolation data obtained by temporal interpolation and spatial interpolation) is taken as an initial value, and then 1 GPV data for every hour (GPV data that is dense both temporally and spatially by performing temporal interpolation and spatial interpolation) is taken as a boundary condition, and prediction calculation every hour from 15:00 Do.

なお、対象としている狭域に存在する観測点の観測値は、1時間毎に観測されている。従って、9時から配信完了時(受信完了時)である15時までの間の、9時、10時、11時、12時、13時、14時の各時刻においても観測値が観測されている。   In addition, the observed value of the observation point which exists in the target narrow area is observed every hour. Therefore, observed values are also observed at 9 o'clock, 9 o'clock, 11 o'clock, 11 o'clock, 12 o'clock, 13 o'clock, and 14 o'clock from 9 o'clock to 15 o'clock at the time of delivery completion (reception completion). Yes.

特開2003−90888JP 2003-90888 A

ところで図6に示す従来手法では、RAMSを用いた予測計算に、GPVデータの初期値を取った時刻(9時)からRAMSによる予測計算を開始した時刻(15時)までの6時間における観測値(9時から配信完了時である15時までの間の、9時、10時、11時、12時、13時、14時の各時刻においても観測された観測値)が反映されていない。このため、RAMSを用いた予測計算の精度に限界があった。   By the way, in the conventional method shown in FIG. 6, in the prediction calculation using RAMS, the observed value in 6 hours from the time when the initial value of GPV data was taken (9 o'clock) to the time when prediction calculation using RAMS was started (15 o'clock). (Observed values observed at 9 o'clock, 9 o'clock, 11 o'clock, 11 o'clock, 12 o'clock, 13 o'clock, and 14 o'clock between 9 o'clock and 15 o'clock when delivery is completed) is not reflected. For this reason, there is a limit to the accuracy of prediction calculation using RAMS.

本発明は、上記従来記述に鑑み、気象解析モデルを用いた予測計算の開始時刻よりも時間的に前に観測された観測値を同化処理することにより、予測計算の精度を向上させることができる気象予測方法を提供することを目的とする。   In view of the above-described conventional description, the present invention can improve the accuracy of the prediction calculation by assimilating the observed values temporally before the start time of the prediction calculation using the weather analysis model. The purpose is to provide a weather forecasting method.

上記課題を解決する本発明の構成は、
広域気象データを基に、気象解析モデルを用いて気象要素の微分方程式を時間積分演算することにより狭域気象データを予測する気象予測方法において、
前記広域気象データを受信完了してこの広域気象データを取得すると共に、前記広域気象データの初期値を取った時刻から前記広域気象データを受信完了した時刻までの間に観測された観測値を取得し、
前記広域気象データの初期値を、広域気象データの初期値を取った時刻の観測値により同化し、この同化した値を、前記微分方程式の初期値として時間積分演算を開始し、
前記広域気象データを受信完了した時刻よりも時間的に前となっている前記広域気象データの初期値を取った時刻から、前記広域気象データを受信完了した時刻までの間の予め決めた一定時間間隔離れた各時刻では、当該時刻における広域気象データを境界値として前記微分方程式を時間積分演算して狭域気象データを求めると共に、このようにして求めた狭域気象データを当該時刻における観測値により同化し、この同化した狭域気象データを基にして当該時刻以降の時間積分演算を継続し、
前記広域気象データを受信完了した時刻、及び、前記広域気象データを受信完了した時刻以降の予め決めた一定時間間隔離れた各時刻では、当該時刻における広域気象データを境界値として前記微分方程式を時間積分演算して狭域気象データを求めることを特徴とする。
The configuration of the present invention for solving the above problems is as follows.
In a weather forecasting method that forecasts narrow-area weather data by calculating the time integral of the differential equation of weather elements using a weather analysis model based on the wide-area weather data,
Obtaining the wide-area meteorological data by acquiring the wide-area meteorological data, and obtaining observation values observed between the time when the initial value of the wide-area meteorological data is taken and the time when the reception of the wide-area meteorological data is completed And
Assimilating the initial value of the regional meteorological data with the observed value at the time when the initial value of the regional meteorological data was taken, the assimilated value is used as the initial value of the differential equation, and the time integration operation is started.
A predetermined fixed time from the time when the initial value of the wide-area weather data that is temporally prior to the time when reception of the wide-area weather data is completed to the time when reception of the wide-area weather data is completed At each time separated by the interval, the differential weather equation is time-integrated with the wide-area weather data at the time as the boundary value to obtain the narrow-area weather data, and the narrow-area weather data thus obtained is the observed value at the time. Assimilation, and based on this assimilated narrow-area meteorological data, continue the time integration calculation after that time,
At the time when the reception of the wide area weather data is completed, and at each time separated by a predetermined time interval after the time when the reception of the wide area weather data is completed, the differential equation is expressed as a time value with the wide area weather data at the time as a boundary value. It is characterized by obtaining the narrow-area weather data by integrating calculation.

また本発明の構成は、上記気象予測方法において、
観測値により同化する気象要素は、少なくとも風速の東西成分及び風速の南北成分を含むこと、
観測値により同化する際に用いる重みファクターは、観測地点から離れるにしたがい小さくなっていること、
観測値により同化する際に用いる重みファクターは、演算により求める狭域気象データが高さ方向に沿う複数箇所のデータである場合に、高層の狭域気象データに用いるものほど、小さくなっていること、
同化の処理を複数回行うこと、
同化の処理は、観測値が観測された時刻を含みこの時刻の前後を含む時間帯の狭域気象データに対して行うことを特徴とする。
Further, the configuration of the present invention is the above-described weather prediction method,
Meteorological elements assimilated by observations include at least the east-west component of wind speed and the north-south component of wind speed,
The weighting factor used for assimilation by observation value decreases with distance from the observation point,
The weighting factor used for assimilation by observation values is smaller when the narrow-area meteorological data obtained by calculation is data at multiple locations along the height direction, the one used for the high-rise narrow-area meteorological data. ,
Performing assimilation multiple times,
The assimilation process is performed on narrow-area meteorological data in a time zone that includes the time at which the observed value was observed and before and after this time.

本発明では、広域気象データの初期値を取った時刻から、この広域気象データを受信完了した時刻までの間に観測された観測値を用いて同化をしているため、受信完了時刻から以降に予測演算して得た気象データの精度を向上させることができる。   In the present invention, assimilation is performed using observation values observed between the time when the initial value of the wide-area weather data is taken and the time when the reception of the wide-area weather data is completed. It is possible to improve the accuracy of weather data obtained by predictive calculation.

本発明を実施するための最良の形態を、以下に詳細に説明する。   The best mode for carrying out the present invention will be described in detail below.

[気象予測の演算手法]
図1及び図2は本発明の実施例を示す説明図である。本実施例では、メソスケール気象解析モデルとしてRAMSを用い、広域気象データとしてGPVデータを用いたものである。RAMSやGPVデータは、[背景技術]で説明したものと同じであるので、その詳細説明は省略する。
また、実施例の気象予測演算は、図2に示す気象解析モデル演算部10により行う。この気象解析モデル演算部10により行う演算手法を、以下に示す。
[Calculation method of weather forecast]
1 and 2 are explanatory views showing an embodiment of the present invention. In this embodiment, RAMS is used as a mesoscale weather analysis model, and GPV data is used as wide-area weather data. Since the RAMS and GPV data are the same as those described in [Background Art], detailed description thereof is omitted.
Moreover, the weather prediction calculation of an Example is performed by the weather analysis model calculating part 10 shown in FIG. A calculation method performed by the weather analysis model calculation unit 10 is shown below.

GPVデータは9時を初期値とした51時間分の予測データであり、15時に配信が完了する。   The GPV data is predicted data for 51 hours with 9 o'clock as an initial value, and distribution is completed at 15:00.

本実施例では、狭域気象予測をするため、15時にGPVデータの受信(GPVデータの取得)を完了すると共に、過去6時間分(9時、10時、11時、12時、13時、14時の各時刻)の観測値をも取得する。ここにおいて「過去6時間分の観測値」とは、GPVデータの初期値を取った時刻からGPVデータを受信完了した時刻までの時間において観測された観測値である。   In this embodiment, in order to make a narrow-area weather forecast, the reception of GPV data (GPV data acquisition) is completed at 15:00, and the past 6 hours (9 o'clock, 10 o'clock, 11 o'clock, 12 o'clock, 13 o'clock, Observed values at each time of 14:00) are also acquired. Here, the “observed values for the past 6 hours” are observed values observed from the time when the initial value of the GPV data is taken to the time when the GPV data is completely received.

気象モデルとしてRAMSを用いた予測計算は、GPVデータの配信完了時刻(受信完了時刻)である15時を起点として6時間前(つまり9時)のGPVデータ(時間内挿補間演算と空間内挿補間演算をして時間的にも空間的にも密なGPVデータ,以降同様)、つまりGPVデータの初期値を、RAMSを用いた予測計算の初期値として取り込む。また15時を起点として6時間前(つまり9時)の観測値を取り込む。そして、この観測値により、予測計算の初期値となる9時のGPVデータを同化する。なお同化の具体的手法については後述する。   Prediction calculation using RAMS as a weather model is based on GPV data (temporal interpolation and spatial interpolation) 6 hours before (that is, 9 o'clock) starting from 15:00, which is the GPV data distribution completion time (reception completion time). GPV data that is dense in terms of time and space by interpolation and the same applies hereinafter), that is, the initial value of the GPV data is taken as the initial value of the prediction calculation using RAMS. Moreover, the observation value of 6 hours before (that is, 9:00) is taken in from 15:00. Then, the 9 o'clock GPV data, which is the initial value of the prediction calculation, is assimilated by this observation value. A specific method of assimilation will be described later.

このようにして同化した初期値を基に、RAMSによる気象要素の微分方程式を時間積分演算を開始する。   Based on the initial values assimilated in this way, the time integration operation of the differential equation of the weather element by RAMS is started.

次に、GPVデータの配信完了時刻(受信完了時刻)である15時を起点として5時間前(つまり10時)のGPVデータを境界値として取り込む。また15時を起点として5時間前(つまり10時)の観測値を取り込む。
そして、5時間前(つまり10時)のGPVデータを境界値として取り込むことにより、5時間前の狭域気象データを予測演算する。また、予測演算した狭域気象データを、5時間前の観測値により同化する。そして、同化した狭域気象データを基にして、この時刻以降のRAMSによる気象要素の微分方程式の時間積分演算を継続する。
Next, GPV data 5 hours before (that is, 10:00) starting from 15:00, which is the distribution completion time (reception completion time) of GPV data, is fetched as a boundary value. Moreover, the observation value of 5 hours before (that is, 10:00) is taken in from 15:00.
Then, by fetching GPV data five hours ago (that is, 10 o'clock) as a boundary value, the narrow-area weather data five hours ago is predicted and calculated. Moreover, the forecasted and calculated narrow-area meteorological data is assimilated with the observed values of 5 hours ago. Then, based on the assimilated narrow-area meteorological data, the time integration calculation of the differential equation of the meteorological element by the RAMS after this time is continued.

次に、GPVデータの配信完了時刻(受信完了時刻)である15時を起点として4時間前(つまり11時)のGPVデータを境界値として取り込む。また15時を起点として4時間前(つまり11時)の観測値を取り込む。
そして、4時間前(つまり11時)のGPVデータを境界値として取り込むことにより、4時間前の狭域気象データを予測演算する。また、予測演算した狭域気象データを、4時間前の観測値により同化する。そして、同化した狭域気象データを基にして、この時刻以降のRAMSによる気象要素の微分方程式の時間積分演算を継続する。
Next, GPV data 4 hours before (ie, 11:00) starting from 15:00, which is the distribution completion time (reception completion time) of GPV data, is fetched as a boundary value. In addition, the observation value 4 hours before (that is, 11:00) from 15:00 is taken in.
Then, the GPV data 4 hours before (that is, 11:00) is taken in as a boundary value to predict and calculate the narrow area weather data 4 hours ago. In addition, the forecasted and calculated narrow-area meteorological data is assimilated with the observation values obtained 4 hours ago. Then, based on the assimilated narrow-area meteorological data, the time integration calculation of the differential equation of the meteorological element by the RAMS after this time is continued.

以降同様に、GPVデータの配信完了時刻(受信完了時刻)である15時を起点として3時間前(つまり12時)、2時間前(つまり13時)、1時間前(つまり14時)においても、その時刻のGPVデータを境界値として取り込んでその時刻の狭域気象データを予測演算し、この予測演算した狭域気象データをその時刻の観測値により同化し、同化した狭域気象データを基にして、その時刻以降のRAMSによる気象要素の微分方程式の時間積分演算を継続していく。   In the same manner, the GPV data distribution completion time (reception completion time) starts at 15:00, 3 hours before (that is, 12:00), 2 hours before (that is, 13:00), and 1 hour before (that is, 14:00) Then, the GPV data at that time is taken as a boundary value, and the narrow-area meteorological data at that time is predicted and calculated, and the predicted and calculated narrow-area meteorological data is assimilated with the observed values at that time. Then, the time integration calculation of the differential equation of the weather element by the RAMS after that time is continued.

現在時刻(つまり15時)及び、現在時刻(つまり15時)以降は、1時間毎(15時,16時,17時・・・)のGPVデータ(時間内挿補間演算と空間内挿補間演算をして時間的にも空間的にも密なGPVデータ)を境界条件として取り込み、15時から以降1時間毎の狭域気象データの予測計算を行う。   The current time (ie, 15:00) and after the current time (ie, 15:00), the GPV data (time interpolation interpolation and spatial interpolation interpolation) every hour (15:00, 16:00, 17:00 ...) The GPV data (which is dense in terms of time and space) is taken in as a boundary condition, and prediction calculation of narrow-area weather data is performed every hour from 15:00.

このように本実施例では、GPVデータの受信完了時刻の6時間前のGPVデータ(つまりGPVデータの初期値)からRAMSによる計算をスタートさせ、6時間分の観測値を同化させながら計算を進めるようにしたので、GPVデータの受信完了時刻(15時)とその後の1時間毎の時刻における狭域気象データの予測精度を向上させることができる。   As described above, in this embodiment, the calculation by the RAMS is started from the GPV data 6 hours before the GPV data reception completion time (that is, the initial value of the GPV data), and the calculation proceeds while assimilating the observation values for 6 hours. Since it did in this way, the prediction precision of the narrow area | region weather data in the reception completion time (15 o'clock) of GPV data and the time for every hour after that can be improved.

[同化の第1の手法]
ここで、気象データを観測値により同化する同化の第1の手法について説明する。
RAMSで使用されている同化手法は式(7)で表される。
[First method of assimilation]
Here, a first assimilation method for assimilating weather data with observed values will be described.
The assimilation technique used in RAMS is expressed by equation (7).

Figure 2006064609
Figure 2006064609

気象解析モデル演算部10では、観測値が存在する時刻において、気象解析モデル内の一連の方程式(運動方程式、熱力学方程式等)に対して時間積分演算を終えた後、風速のU(東西)成分、V(南北)成分に対して、式(7)に示す同化式により、地表面付近の(最下層格子点の)風速のU成分,V成分を観測値に近づける(ナッジングさせる)。   The meteorological analysis model calculation unit 10 finishes the time integration calculation for a series of equations (motion equation, thermodynamic equation, etc.) in the meteorological analysis model at the time when the observation value exists, and then the wind speed U (east-west) For the component and V (north-south) component, the U component and V component of the wind velocity near the ground surface (at the lowest lattice point) are brought closer to the observed value (nudged) by the assimilation equation shown in equation (7).

即ち、気象解析モデル演算部10は、図3(a)に示す同化の影響範囲の半径rを適宜設定し、観測データを観測した観測地点を中心とした影響範囲内(半径r内)に存在する複数の気象データ(計算格子の格子点上の計算値)のU成分,V成分に対して、式(7)で求めた値を加えることにより同化を行う。
なお、この例では、重みのファクターεは図3(b)に示すように、水平方向に均一にしている。
That is, the meteorological analysis model calculation unit 10 appropriately sets the radius r of the assimilation influence range shown in FIG. 3A and exists within the influence range (within the radius r) centered on the observation point where the observation data is observed. Assimilation is performed by adding the value obtained by the equation (7) to the U and V components of a plurality of meteorological data (calculated values on the grid points of the calculation grid).
In this example, the weighting factor ε is uniform in the horizontal direction as shown in FIG.

[同化の第2の手法]
同化の第2の手法では、同化の第1の手法に追加して、更に気象場を実現象に近づけるため、気温に対しても式(7)を適用する。即ち、気象要素φとして気温を追加する。
[Second method of assimilation]
In the second assimilation method, in addition to the first assimilation method, the equation (7) is also applied to the temperature in order to bring the weather field closer to the actual phenomenon. That is, the temperature is added as the weather element φ.

[同化の第3の手法]
同化の第3の手法では、同化の第1,第2の手法に追加して、更に気象場を実現象に近づけるため、気圧に対しても式(7)を適用する。即ち、気象要素φとして気圧を追加する。
[Third method of assimilation]
In the third method of assimilation, in addition to the first and second methods of assimilation, the equation (7) is also applied to the atmospheric pressure in order to bring the weather field closer to the actual phenomenon. That is, the atmospheric pressure is added as the weather element φ.

[同化の第4の手法]
同化の第4の手法では、同化の第1,第2,第3の手法に追加して、更に気象場を実現象に近づけるため、湿度に対しても式(7)を適用する。即ち、気象要素φとして湿度を追加する。
[Fourth method of assimilation]
In the fourth method of assimilation, in addition to the first, second, and third methods of assimilation, Equation (7) is also applied to the humidity in order to bring the weather field closer to the actual phenomenon. That is, humidity is added as the weather element φ.

[同化の第5の手法]
RAMSによる気象要素の微分方程式の時間積分演算をして求めた気象データ(格子点上のデータ)としては、例えば、地上(最下層格子点)でのデータ、地上から例えば100mの高さ(第2層格子点)でのデータ、地上から例えば200mの高さ(第3層格子点)でのデータがある。なお、高さ方向に関しては、第10層格子点程度(地上から1000m程度)まで気象データを設定することができる。
[Fifth method of assimilation]
As the meteorological data (data on the grid points) obtained by the time integration calculation of the differential equation of the meteorological element by RAMS, for example, data on the ground (the lowest grid point), for example, a height of 100 m from the ground (first There are data at two-layer grid points) and data at a height of 200 m from the ground (third-layer grid points). In addition, regarding the height direction, weather data can be set up to about the 10th layer lattice point (about 1000 m from the ground).

そこで、上述した同化の第1〜第4の手法に追加して、同化の第5の手法では、地上のデータだけでなく、上層の格子点のデータ(例えば第2層格子点や第3層格子点)に対しても、式(7)の適用をする。   Therefore, in addition to the first to fourth methods of assimilation described above, in the fifth method of assimilation, not only the ground data but also the data of the upper layer lattice points (for example, the second layer lattice points and the third layer) Equation (7) is also applied to (grid points).

これは、地表付近の影響が及んで発達する大気境界層の厚さは1000m程度であり、地表付近の影響は上空にまで及ぶと考えられるからである。   This is because the thickness of the atmospheric boundary layer that develops due to the influence of the vicinity of the ground surface is about 1000 m, and the influence of the vicinity of the ground surface is considered to extend to the sky.

[同化の第6の手法]
観測値を同化する場合には、観測地点から離れるほど同化の影響は小さく、観測地点周辺で同化の影響が大きいことが現実的である。このため、上述した同化の第1〜第5の手法に追加して、同化の第6の手法では、観測地点からの距離に応じて重みファクターεの重みを決定する。例えば図4(a)に示すように、X,Y方向(平面方向)に対しては、重みファクターεの大きさは、観測地点が最も大きく観測地点から離れるに従い小さくなるような、一種のガウス分布のような重みをつける。
[Sixth method of assimilation]
When assimilating observed values, the further away from the observation point, the smaller the influence of assimilation, and the greater the influence of assimilation around the observation point is realistic. For this reason, in addition to the first to fifth methods of assimilation described above, the sixth method of assimilation determines the weight of the weight factor ε according to the distance from the observation point. For example, as shown in FIG. 4 (a), for the X and Y directions (plane direction), the weight factor ε is a kind of Gaussian in which the observation point is the largest and decreases as the distance from the observation point increases. Give weight like distribution.

更に、RAMSによる気象要素の微分方程式の時間積分演算をして求めた気象データ(格子点上のデータ)としては、地上(最下層格子点)でのデータ、地上から例えば100mの高さ(第2層格子点)でのデータ、地上から例えば200mの高さ(第3層格子点)でのデータがある。
そこで、図4(b)(c)(d)に示すように、重みファクターεはガウス分布の重みとなっているが、上空に行くほど重みが小さくするようにした。つまり、上空に行くほど、同化の影響を小さくしている。
Furthermore, as the meteorological data (data on the grid points) obtained by time integration calculation of the differential equation of the meteorological element by RAMS, the data on the ground (the lowest grid point), for example, a height of 100 m from the ground (first There are data at two-layer grid points) and data at a height of 200 m from the ground (third-layer grid points).
Therefore, as shown in FIGS. 4B, 4C, and 4D, the weight factor ε is a weight of the Gaussian distribution, but the weight is made smaller as it goes up. In other words, the effect of assimilation decreases as the distance increases.

[同化の第7の手法]
上述した同化の第1〜第6の手法に追加して、同化の効果をより強化したい場合には、観測時刻において式(7)を複数回使用する。即ち、同化の第7の手法では、式(7)で求められた計算値に対して、式(8)に示すように、更に観測値を同化させる。これを繰り返すほど、計算値は観測値に近づくことになる。
[Seventh method of assimilation]
In addition to the first to sixth methods of assimilation described above, when it is desired to further enhance the assimilation effect, Expression (7) is used a plurality of times at the observation time. That is, in the seventh method of assimilation, the observed value is further assimilated as shown in the equation (8) with respect to the calculated value obtained by the equation (7). As this is repeated, the calculated value approaches the observed value.

Figure 2006064609
Figure 2006064609

[同化の第8の手法]
同化の第1〜第7の手法に追加して、同化の効果をより強化したい場合には、観測時刻に対してのみ式(1)を適用させずに、図5に示すように、観測時刻前後に対しても式(1)を適用する。つまり同化の第8の手法では、観測時刻1点での同化ではなく、同化させる時間帯を決定する。この手法は、地形の影響が小さく(観測地点の周辺に地形の変動が少なく)、風向等の時間的な変動が小さい場合に有効である。
[Eighth method of assimilation]
In addition to the first to seventh methods of assimilation, when it is desired to further enhance the effect of assimilation, the equation (1) is not applied only to the observation time, as shown in FIG. Formula (1) is applied also to front and back. That is, in the eighth method of assimilation, the time zone to be assimilated is determined instead of assimilation at one observation time point. This method is effective when the influence of the topography is small (the topography has little fluctuation in the vicinity of the observation point) and the temporal fluctuation such as the wind direction is small.

[同化の第9の手法]
同化の第9の手法では、同化の第1〜第8の手法に対して、式(7)の代わりに、式(9)を使用する。式(9)は式(7)の右辺に、φについての通常の方程式(運動方程式、熱エネルギー方程式等)F(φ)を加えており、式(7)に比べて、物理的な現象を考慮しながら同化を行う。なおF(φ)は、気象解析モデル内の各気象要素φについての通常の式であり、具体的には式(1)〜式(6)が相当する。
[Ninth method of assimilation]
In the ninth method of assimilation, equation (9) is used instead of equation (7) with respect to the first to eighth methods of assimilation. Formula (9) adds a normal equation (kinetic equation, thermal energy equation, etc.) F (φ) for φ to the right side of Formula (7), and the physical phenomenon is compared with Formula (7). Assimilate with consideration. Note that F (φ) is a normal expression for each weather element φ in the weather analysis model, and specifically corresponds to expressions (1) to (6).

Figure 2006064609
Figure 2006064609

ここで、同化について説明する。
同化とは、気象解析モデルに観測値を取り込んで、気象解析モデルの計算結果を観測値に近づける手法である。
本件で使用する同化式は式(7)で表される。式(7)を差分化した形で表すと、
Here, assimilation will be described.
Assimilation is a method of taking observed values into a weather analysis model and bringing the calculation result of the weather analysis model closer to the observed value.
The assimilation formula used in this case is expressed by formula (7). When Expression (7) is expressed in a differentiated form,

Figure 2006064609
Figure 2006064609

この式では、右辺第二項の分だけ計算値(φbefore)を補正(同化)している。
つまり、計算値(φbefore)が観測値(φo)より小さければ、右辺第二項は計算値(φbefore)を大きくするように作用し、逆に計算値が観測値より大きければ、右辺第二項は計算値(φbefore)を小さくするように作用して、新しい計算値(φafter)を得る。
In this equation, the calculated value (φ before ) is corrected (analyzed) by the amount of the second term on the right side.
In other words, if the calculated value (φ before ) is smaller than the observed value (φ o ), the second term on the right side acts to increase the calculated value (φ before ). Conversely, if the calculated value is larger than the observed value, the right side The second term acts to reduce the calculated value (φ before ) to obtain a new calculated value (φ after ).

本発明は、広域気象データから狭域気象データを求める気象予測方法に適用することができる。   The present invention can be applied to a weather prediction method for obtaining narrow area weather data from wide area weather data.

本発明の実施例に係る気象予測方法を示す説明図。Explanatory drawing which shows the weather prediction method which concerns on the Example of this invention. 本発明の実施例に係る気象予測方法を示す説明図。Explanatory drawing which shows the weather prediction method which concerns on the Example of this invention. 同化手法を示す説明図。Explanatory drawing which shows an assimilation method. 同化手法を示す説明図。Explanatory drawing which shows an assimilation method. 同化手法を示す説明図。Explanatory drawing which shows an assimilation method. 従来の気象予測方法を示す説明図。Explanatory drawing which shows the conventional weather prediction method.

符号の説明Explanation of symbols

10 気象解析モデル演算部   10 Meteorological analysis model calculation unit

Claims (6)

広域気象データを基に、気象解析モデルを用いて気象要素の微分方程式を時間積分演算することにより狭域気象データを予測する気象予測方法において、
前記広域気象データを受信完了してこの広域気象データを取得すると共に、前記広域気象データの初期値を取った時刻から前記広域気象データを受信完了した時刻までの間に観測された観測値を取得し、
前記広域気象データの初期値を、広域気象データの初期値を取った時刻の観測値により同化し、この同化した値を、前記微分方程式の初期値として時間積分演算を開始し、
前記広域気象データを受信完了した時刻よりも時間的に前となっている前記広域気象データの初期値を取った時刻から、前記広域気象データを受信完了した時刻までの間の予め決めた一定時間間隔離れた各時刻では、当該時刻における広域気象データを境界値として前記微分方程式を時間積分演算して狭域気象データを求めると共に、このようにして求めた狭域気象データを当該時刻における観測値により同化し、この同化した狭域気象データを基にして当該時刻以降の時間積分演算を継続し、
前記広域気象データを受信完了した時刻、及び、前記広域気象データを受信完了した時刻以降の予め決めた一定時間間隔離れた各時刻では、当該時刻における広域気象データを境界値として前記微分方程式を時間積分演算して狭域気象データを求めることを特徴とする気象予測方法。
In a weather forecasting method that forecasts narrow-area weather data by calculating the time integral of the differential equation of the weather element using the weather analysis model based on the wide-area weather data,
Obtaining the wide-area meteorological data by acquiring the wide-area meteorological data, and obtaining observation values observed between the time when the initial value of the wide-area meteorological data is taken and the time when the reception of the wide-area meteorological data is completed And
Assimilating the initial value of the wide-area weather data with the observed value at the time when the initial value of the wide-area weather data was taken, this assimilated value is used as the initial value of the differential equation to start the time integration operation,
A predetermined fixed time from the time when the initial value of the wide-area weather data that is temporally prior to the time when reception of the wide-area weather data is completed to the time when reception of the wide-area weather data is completed At each time separated by the interval, the differential weather equation is time-integrated with the wide-area weather data at the time as the boundary value to obtain the narrow-area weather data, and the narrow-area weather data thus obtained is the observed value at the time. Assimilation, and based on this assimilated narrow-area meteorological data, continue the time integration calculation after that time,
At the time when the reception of the wide area weather data is completed, and at each time separated by a predetermined time interval after the time when the reception of the wide area weather data is completed, the differential equation is expressed as a time value with the wide area weather data at the time as a boundary value. A meteorological forecasting method characterized by obtaining narrow area weather data by integrating calculation.
請求項1において、観測値により同化する気象要素は、少なくとも風速の東西成分及び風速の南北成分を含むことを特徴とする気象予測方法。   2. The weather prediction method according to claim 1, wherein the meteorological element assimilated by the observed value includes at least a east-west component of wind speed and a north-south component of wind speed. 請求項1または請求項2において、観測値により同化する際に用いる重みファクターは、観測地点から離れるにしたがい小さくなっていることを特徴とする気象予測方法。   3. The weather prediction method according to claim 1, wherein the weighting factor used when assimilating according to the observation value decreases as the distance from the observation point increases. 請求項1乃至請求項3のいずれか一項において、観測値により同化する際に用いる重みファクターは、演算により求める狭域気象データが高さ方向に沿う複数箇所のデータである場合に、高層の狭域気象データに用いるものほど、小さくなっていることを特徴とする気象予測方法。   In any one of Claims 1 to 3, the weighting factor used when assimilating with the observation value is a high-rise layer when the narrow-area meteorological data obtained by calculation is data at a plurality of locations along the height direction. A weather prediction method characterized in that the smaller the data used for narrow-area weather data, the smaller. 請求項1乃至請求項4の何れか一項において、同化の処理を複数回行うことを特徴とする気象予測方法。   5. The weather prediction method according to claim 1, wherein the assimilation process is performed a plurality of times. 請求項1乃至請求項5の何れか一項において、同化の処理は、観測値が観測された時刻を含みこの時刻の前後を含む時間帯の狭域気象データに対して行うことを特徴とする気象予測方法。   The assimilation process according to any one of claims 1 to 5, wherein the assimilation process is performed on narrow-area meteorological data in a time zone that includes a time when an observed value is observed and includes the time before and after the time. Weather forecast method.
JP2004249539A 2004-08-30 2004-08-30 Weather forecasting method Active JP4209370B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2004249539A JP4209370B2 (en) 2004-08-30 2004-08-30 Weather forecasting method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2004249539A JP4209370B2 (en) 2004-08-30 2004-08-30 Weather forecasting method

Publications (2)

Publication Number Publication Date
JP2006064609A true JP2006064609A (en) 2006-03-09
JP4209370B2 JP4209370B2 (en) 2009-01-14

Family

ID=36111214

Family Applications (1)

Application Number Title Priority Date Filing Date
JP2004249539A Active JP4209370B2 (en) 2004-08-30 2004-08-30 Weather forecasting method

Country Status (1)

Country Link
JP (1) JP4209370B2 (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008089418A (en) * 2006-10-02 2008-04-17 Mitsubishi Heavy Ind Ltd Diffusion estimation system, method and program
JP2008241433A (en) * 2007-03-27 2008-10-09 Nec Corp Observation data assimilation method
JP2009128180A (en) * 2007-11-22 2009-06-11 Toshiba Corp Fog prediction device and fog prediction method
JP2010060443A (en) * 2008-09-04 2010-03-18 Japan Weather Association Weather forecast device, method, and program
JP2017111074A (en) * 2015-12-18 2017-06-22 三菱重工業株式会社 Weather data assimilation method, weather forecasting method, and weather forecasting system
CN115903088A (en) * 2022-12-20 2023-04-04 中国民用航空局空中交通管理局航空气象中心 Meteorological element nowcasting method and system based on advection diffusion model

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008089418A (en) * 2006-10-02 2008-04-17 Mitsubishi Heavy Ind Ltd Diffusion estimation system, method and program
JP2008241433A (en) * 2007-03-27 2008-10-09 Nec Corp Observation data assimilation method
JP2009128180A (en) * 2007-11-22 2009-06-11 Toshiba Corp Fog prediction device and fog prediction method
JP2010060443A (en) * 2008-09-04 2010-03-18 Japan Weather Association Weather forecast device, method, and program
JP2017111074A (en) * 2015-12-18 2017-06-22 三菱重工業株式会社 Weather data assimilation method, weather forecasting method, and weather forecasting system
CN115903088A (en) * 2022-12-20 2023-04-04 中国民用航空局空中交通管理局航空气象中心 Meteorological element nowcasting method and system based on advection diffusion model

Also Published As

Publication number Publication date
JP4209370B2 (en) 2009-01-14

Similar Documents

Publication Publication Date Title
JP4404220B2 (en) Gas condition prediction apparatus, method, program, and diffusion condition prediction system
Katsman et al. Boundary current eddies and their role in the restratification of the Labrador Sea
Zheng et al. Sea surface temperature biases under the stratus cloud deck in the southeast Pacific Ocean in 19 IPCC AR4 coupled general circulation models
Drushka et al. Vertical structure of Kelvin waves in the Indonesian Throughflow exit passages
Deveson et al. A classification of FASTEX cyclones using a height‐attributable quasi‐geostrophic vertical‐motion diagnostic
Drushka et al. In situ observations of Madden–Julian oscillation mixed layer dynamics in the Indian and western Pacific Oceans
US20130066569A1 (en) Power generation predicting apparatus and method thereof
Mo et al. Estimating the extreme wind speed for regions in China using surface wind observations and reanalysis data
Gao et al. An efficient dual-resolution approach for ensemble data assimilation and tests with simulated Doppler radar data
Hally et al. An ensemble study of HyMeX IOP6 and IOP7a: sensitivity to physical and initial and boundary condition uncertainties
Tsujino et al. Potential vorticity mixing and rapid intensification in the numerically simulated Supertyphoon Haiyan (2013)
JP2007248355A (en) Device and method for estimating weather data
JP4209370B2 (en) Weather forecasting method
Anderson et al. Model dynamics of summertime low‐level jets over northwestern Mexico
JP5738140B2 (en) Substance release amount estimation apparatus, method and program
Diansky et al. Thermohaline structure of waters in the North Atlantic in different phases of the Atlantic multidecadal oscillation
JP3851641B2 (en) Rainfall prediction method and rainfall prediction program
Kipling et al. Dynamic subgrid heterogeneity of convective cloud in a global model: description and evaluation of the Convective Cloud Field Model (CCFM) in ECHAM6–HAM2
Sun et al. Using the inverse of expected error variance to determine weights of individual ensemble members: application to temperature prediction
Davis et al. High-resolution hurricane forecasts
KR102002593B1 (en) Method and apparatus for analyzing harmful gas diffusion in a specific space
Kunii et al. Verification and intercomparison of mesoscale ensemble prediction systems in the Beijing 2008 Olympics Research and Development Project
CN109766611B (en) Wind farm power simplified prediction method considering terrain gradient
Truong et al. A modified Kain–Fritsch scheme and its application for the simulation of an extreme precipitation event in Vietnam
Rémy et al. Ensemble Kalman filter data assimilation in a 1D numerical model used for fog forecasting

Legal Events

Date Code Title Description
A621 Written request for application examination

Free format text: JAPANESE INTERMEDIATE CODE: A621

Effective date: 20060912

A977 Report on retrieval

Free format text: JAPANESE INTERMEDIATE CODE: A971007

Effective date: 20080806

TRDD Decision of grant or rejection written
A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

Effective date: 20081007

A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

A61 First payment of annual fees (during grant procedure)

Free format text: JAPANESE INTERMEDIATE CODE: A61

Effective date: 20081022

R151 Written notification of patent or utility model registration

Ref document number: 4209370

Country of ref document: JP

Free format text: JAPANESE INTERMEDIATE CODE: R151

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20111031

Year of fee payment: 3

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20111031

Year of fee payment: 3

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20121031

Year of fee payment: 4

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20121031

Year of fee payment: 4

FPAY Renewal fee payment (event date is renewal date of database)

Free format text: PAYMENT UNTIL: 20131031

Year of fee payment: 5

R250 Receipt of annual fees

Free format text: JAPANESE INTERMEDIATE CODE: R250