JP4204034B2 - Tropical cyclone wind speed prediction method and apparatus - Google Patents

Description

ここで、lat(t)は、時刻ｔにおける緯度を表す。経度についても、前記(1)式と同様の式により、２時間後の経度を算出することができる。
【００１６】
前記(1)式と同様の線形補間法により、３時間後、４時間後、５時間後、７時間後、・・・の熱帯性低気圧中心位置（緯度、経度）が求められる。
(2)熱帯性低気圧中心気圧の予測
熱帯性低気圧中心気圧の予測は、線形補間法およびｋ次移動平均フィルタを用いて、下記の▲１▼、▲２▼の手順のように算出する。
【００１７】
▲１▼線形補間法
実況情報時の時刻ｔ_ｎより１時間後の時刻をｔ_ｎ＋１、ｍ時間後の時刻をｔ_ｎ＋ｍとすると、２時間後の時刻ｔ_ｎ＋２における熱帯性低気圧中心気圧は、下記の(2)式により算出される。
【００１８】
【数２】

ここで、hPa(t)は、時刻ｔにおける熱帯性低気圧中心気圧を表す。
【００１９】
▲２▼ｋ次移動平均フィルタ
２時間後の時刻をｔ_ｎ＋２とし、該時刻ｔ_ｎ＋２より１〜（ｋ−１）／２時間前の時刻をｔ_{ｎ＋２−（ｋ−１）／２}、・・・、ｔ_ｎ＋１ 、１〜（ｋ−１）／２時間後の時刻をｔ_ｎ＋３、・・・、ｔ_{ｎ＋２＋（ｋ−１）／２}とすると、２時間後の熱帯性低気圧中心気圧hPa(t_ｎ＋２)’は、次の(3)により算出することができる。
ここで、ｋは奇数である。
【００２０】
【数３】

なお、実況情報より１、２時間前のデータが存在しない時には、前記(2)式を用いて、該１、２時間前の熱帯性低気圧中心気圧を算出する。前記と同様にして、３時間後、４時間後、５時間後、７時間後、・・・の熱帯性低気圧中心気圧を求める。
(3)熱帯性低気圧進行速度の予測
熱帯性低気圧進行速度の予測は、前記気象表示装置２から取得した熱帯性低気圧特性予測情報から、各予測ポイントの速度を算出し、算出した速度を基に、線形補間法およびＬ次移動平均フィルタを用いて算出する。
【００２１】
▲１▼熱帯性低気圧進行速度の算出
熱帯性低気圧進行速度は、各予測ポイントの熱帯性低気圧中心位置より、次のように算出する。例えば、図４において、熱帯性低気圧特性予測情報としての１時間後の熱帯性低気圧中心位置をＡ、ｍ時間後の熱帯性低気圧中心位置をＢ、Ａを通る緯度線とＢを通る子午線の交点をＣ、北極点をＮとする。この球面三角形ＮＡＢに球面三角法を適用し、ＡＢ間の角距離⌒ＡＢ、および距離Ｒを、下記の(4)式,(5)式により求める。
【００２２】
【数４】

Ｒ＝Ｒ_ｅ ×ＡＢ ・・・(5)
ここで、⌒は角距離、Ｒ_ｅ は地球半径［ｋｍ］を示す。
【００２３】
前記距離Ｒと移動時間ｔにより、速度spを次の(6)式により求める。
【００２４】
速度sp＝Ｒ／ｔ ・・・(6)
▲２▼線形補間法
次に、実況情報時の時刻ｔ_ｎより１時間後の時刻をｔ_ｎ＋１、ｍ時間後の時刻をｔ_ｎ＋ｍとして、２時間後の時刻ｔ_ｎ＋２における熱帯性低気圧進行速度を、次の(7)式により算出する。
【００２５】
【数５】

ここで、sp(t)は時刻ｔにおける熱帯性低気圧進行速度を示す。
【００２６】
なお、実況情報より１、２時間前の実績データを抽出し、データが存在しなければ、(7)式を用いて、１、２時間前の熱帯性低気圧進行速度を算出する。
【００２７】
▲３▼Ｌ次移動平均フィルタ
次に、２時間後の時刻をｔ_ｎ＋２とし、該時刻ｔ_ｎ＋２より１〜（Ｌ−１）／２時間前の時刻をｔ_{ｎ＋２−（Ｌ−１）／２}、・・・、ｔ_ｎ＋１ 、１〜（Ｌ−１）／２時間後の時刻をｔ_ｎ＋３、・・・、ｔ_{ｎ＋２＋（Ｌ−１）／２}とすると、２時間後の熱帯性低気圧進行速度sp(t_ｎ＋２)’は、次の(8)式により算出することができる。
ここで、Ｌは奇数である。
【００２８】
【数６】

(4)熱帯性低気圧進行方向の予測
前記の(1)で算出した熱帯性低気圧中心位置より、球面三角法を使用して、次の(9)式により、熱帯性低気圧の進行方向を算出する。
【００２９】
cos∠ＡＢＣ＝tanＢＣ／tanＡＢ ・・・(9)
前記の計算により、例えば２２５°＜∠ＡＢＣ≦２４７．５°の時には熱帯性低気圧進行方向は北東、２４７．５°＜∠ＡＢＣ≦２７０°の時は北北東と予測される。
【００３０】
以上の(1)〜(4)の処理により、７２時間先までの１時間毎の予測情報を算出して、前記ステップＳ２の処理を終了する。
【００３１】
次に、図２のステップＳ３に進み、過去の類似熱帯性低気圧の抽出処理を行う。風速予測を行う進路予測情報（＝１２時間後の予測情報）の熱帯性低気圧中心位置（緯度、経度）を基に、緯度、経度各±α°の範囲にて、過去の熱帯性低気圧実績データをデータベースから抽出する。例えば、図５のように、１２時間後の予測熱帯性低気圧中心位置Ｓから抽出した過去熱帯性低気圧実績点を、Ａ，Ｂ，Ｃ，Ｄ，ＥおよびＦとする。また、気象情報の観測地点であるＺを風速予測地点（例えば、アメダス地点）とする。
【００３２】
図２のステップＳ４では、類似熱帯性低気圧は、過去に存在したか否かの判断がなされる。この判断が肯定の場合にはステップＳ５に進み、否定の場合にはステップＳ１１に進む。
【００３３】
ステップＳ５では、前記ステップＳ３で抽出した過去熱帯性低気圧実績点と今回の熱帯性低気圧の熱帯性低気圧中心位置、中心気圧、進行速度、進行方向の差の２乗和の平方根（距離）を算出し、距離の小さい順に、例えば５個の熱帯性低気圧を選定する。選定する際、例えば以下の条件を満たす過去の類似熱帯性低気圧を選定する。
(a)同一熱帯性低気圧が重複しないようにする。（ただし，過去の類似熱帯性低気圧が５個以下の熱帯性低気圧の場合は、この限りではない）
(b)風速予測地点Ｚと過去の熱帯性低気圧との距離がｒ０未満ものは、過去の類似熱帯性低気圧として選定しない。この理由は、風速予測地点と過去の熱帯性低気圧との距離がｒ０未満のものは、熱帯性低気圧の目に入り誤差が大きくなる可能性があるからである。
または、前記選定において、過去の熱帯性低気圧と該熱帯性低気圧との各特性値毎の積をとり、その和が大きい熱帯性低気圧を類似熱帯性低気圧とする。
【００３４】
または、該熱帯性低気圧の中心位置からのある距離以内、あるいは該熱帯性低気圧の中心位置を中心としたある距離を辺とする四角の枠内から類似熱帯性低気圧をまず選定し、その後前記距離にて選定を行う。
【００３５】
ここでは、選定された５個の熱帯性低気圧を、過去熱帯性低気圧実績点Ａ，Ｃ，Ｄ，ＥおよびＦとする。
【００３６】
ステップＳ６では、選定した過去熱帯性低気圧実績点Ａ，Ｃ，Ｄ，ＥおよびＦにおいて風速予測地点Ｚで観測された風速をデータベースより抽出する。抽出した風速予測実績風速を、Ｖ_Ａ，Ｖ_Ｃ，Ｖ_Ｄ，Ｖ_ＥおよびＶ_Ｆとする。
【００３７】
ステップＳ７に進むと、モデル風速値の算出が行われる。該モデル風速値の算出は、今回の熱帯性低気圧および選定した類似熱帯性低気圧のそれぞれの熱帯性低気圧情報を基に、次の(10)、(11)式の静止熱帯性低気圧モデル式、あるいは(11)式の風速に吹き込み角θ3（風向き）を作用させ、(12)式,(13)式の進行熱帯性低気圧モデル式を用いて、方向を持つ風速ベクトルＶ'の算出が行われる。
【００３８】
Ｖ＝Ｖmax・exp［−｛（lnｒ）^θ１−（lnｒ_０）^θ１｝^２］ ・・・(10)
ここで、Ｖmaxは熱帯性低気圧最大風速、ｒは熱帯性低気圧距離（ｋｍ）、θ_１は距離に対する風速の減衰度分である。なお、前記熱帯性低気圧最大風速Ｖmaxは、下記の(11)式で表すことができる。該(11)式中のｐは中心気圧（hPa)である。
【００３９】
Ｖmax＝ａ_１×（ｐ_１−ｐ）^１／２ ・・・(11)
ただし、ｐ_１は気圧を示す定数でａ_１は比例定数である。
【００４０】
熱帯性低気圧距離ｒがｒ_０ｋｍ以下の場合の進行熱帯性低気圧モデル式は(12)式、熱帯性低気圧距離ｒがｒ_０ｋｍ以上の場合の進行熱帯性低気圧モデル式は(13)式で表される。
【００４１】
Ｖ'＝Ｓi・θ_２・ｒ^２／ｒ_０ ^２・Ｓv・exp［−｛π／（１０・ｒ_０）・ｒ｝］ ・・・(12)
Ｖ'＝Ｓi・θ_２・Ｓv・exp［−｛π／（１０・ｒ_０）・ｒ｝］ ・・・(13)
ここで、Ｓvは熱帯性低気圧進行速度（ｍ／ｓ）、ｒは熱帯性低気圧距離（ｋｍ）、Ｓiは熱帯性低気圧位置と観測地点の位置による影響率、θ_２は進行速度の反映度合いを示す変数である。なお、θ_２・Ｓv・exp（−π／１０・ｒ_０）は、進行速度の影響を示す宮崎の式であり、例えば、平成８年度「各種気象情報による台風風速予測に関する研究」の論文に示されている。
【００４２】
まず、上記(10)、(12)および(13)式のいずれかにて算出した今回の熱帯性低気圧の風速予測地点Ｚにおけるモデル風速値Ｖを求める。続いて、過去の類似熱帯性低気圧Ａ，Ｃ，Ｄ，ＥおよびＦにおける熱帯性低気圧中心気圧、風速予測地点Ｚと熱帯性低気圧中心位置との距離ｒ等を上記(10)、(12)または(13)式に当てはめて、過去の類似熱帯性低気圧Ａ，Ｃ，Ｄ，ＥおよびＦにおける風速予測地点のモデル風速値Ｖａ，Ｖｃ，Ｖｄ，ＶｅおよびＶｆを求める。
【００４３】
次に、図２のステップＳ８に進み、風速の補正値を算出する。風速予測地点Ｚにおける風速の補正値、すなわち、１２時間後の予測情報と過去の類似熱帯性低気圧との、中心気圧などの各変数の相違による風速の補正値は、前記１２時間後の予測情報のモデル風速値Ｖと前記過去の熱帯性低気圧のモデル風速値Ｖａ，Ｖｃ，Ｖｄ，ＶｅおよびＶｆから、次のように算出する。
【００４４】
過去の熱帯性低気圧Ａに関する補正値σａ＝Ｖ−Ｖａ
過去の熱帯性低気圧Ｃに関する補正値σｃ＝Ｖ−Ｖｃ
過去の熱帯性低気圧Ｄに関する補正値σｄ＝Ｖ−Ｖｄ
過去の熱帯性低気圧Ｅに関する補正値σｅ＝Ｖ−Ｖｅ
過去の熱帯性低気圧Ｆに関する補正値σｆ＝Ｖ−Ｖｆ
次に、ステップＳ９に進んで、風速予測地点Ｚにおける予測風速を算出する。１２時間後の風速予測地点Ｚにおける予測風速Ｖ’_Ａ，Ｖ’_Ｃ，Ｖ’_Ｄ，Ｖ’_ＥおよびＶ’_Ｆは、過去の類似熱帯性低気圧Ａ，Ｃ，Ｄ，ＥおよびＦの風速予測地点Ｚにおける実績風速値Ｖ_Ａ，Ｖ_Ｃ，Ｖ_Ｄ，Ｖ_ＥおよびＶ_Ｆと、前記補正値σａ，σｃ，σｄ，σｅおよびσｆから次のように算出する。
【００４５】
Ｖ’_Ａ ＝Ｖ_Ａ ＋σａ
Ｖ’_Ｃ ＝Ｖ_Ｃ ＋σｃ
Ｖ’_Ｄ ＝Ｖ_Ｄ ＋σｄ
Ｖ’_Ｅ ＝Ｖ_Ｅ ＋σｅ
Ｖ’_Ｆ ＝Ｖ_Ｆ ＋ σｆ
次いで、これらの平均を取って１２時間後の風速予測地点Ｚにおける予測風速Ｖ’とする。
【００４６】
Ｖ’＝（Ｖ’_Ａ＋Ｖ’_Ｃ＋Ｖ’_Ｄ＋Ｖ’_Ｅ＋Ｖ’_Ｆ）／５
次に、図２のステップＳ１０に進んで、予測風向きを算出する。すなわち、図６に示されているように、まず、風速予測地点Ｚにおける熱帯性低気圧の接線方向の風向きを算出する。次に、該算出した風向きを、吹き込み角θ_３だけ傾けた風向きを予測風向きとする。
【００４７】
次に、図２のステップＳ４の判断が否定の時、すなわち類似熱帯性低気圧が過去に全くなかった時の動作を説明する。ステップＳ１１では、前記(10)式を用いて、風速を予測する地点Ｚにおける１２時間後のモデル風速値Ｖを算出する。また、ステップＳ１２で、前記のようにして、予測風向きＧを求める。
【００４８】
次に、ステップＳ１３で、１２時間後の予測情報の予測風速を、前記モデル風速値Ｖと予測風向きＧを用いて、次の式から算出する。
【００４９】
１２時間後の予測情報の予測風速＝Ｖ×Ｘ（Ｇ）
ここで、Ｘ（Ｇ）は、風向きＧにおける風向き係数である。
次に、ステップＳ１４で、現在の熱帯性低気圧特性情報（図５の熱帯性低気圧現在位置の熱帯性低気圧特性）を過去の熱帯性低気圧として蓄積する。これにより、過去の熱帯性低気圧の個数を増やし、類似熱帯性低気圧の選択自由度を多くする。
【００５０】
なお、前記の説明では、風速予測地点の予測風速を求めるものであったが、本発明を適用すれば、任意の地点の熱帯性低気圧の風速を求めることができるようになることは明らかである。
【００５１】
次に、本発明の風速・被害予測装置３ｂ（図１参照）の本発明に関する機能のブロック図を、図７に示す。前記ＬＡＮ４等の通信線を通って熱帯性低気圧特性予測情報が送られて来ると、通信装置１１はこれを受信する。受信した情報は記憶部１２に記憶される。熱帯性低気圧特性予測情報演算部１３は、該熱帯性低気圧特性予測情報を基に、一時間毎の熱帯性低気圧特性予測情報を算出し、該記憶部１２に格納する。該記憶部１２に、過去の熱帯性低気圧の特性（経路、各地点の中心気圧、風速など）等のデータが記憶されているとすると、類似熱帯性低気圧選択部１４は、今回の熱帯性低気圧の例えば１２時間後の予測中心位置に関し、その近い位置の過去のある時刻の熱帯性低気圧を類似熱帯性低気圧として複数個選択する。モデル風速値演算部１５は、今回の熱帯性低気圧および類似熱帯性低気圧の特性値を基に、ある地点（例えば、風速予測地点）における今回の熱帯性低気圧および類似熱帯性低気圧のモデル風速値Ｖ，Ｖａ，Ｖｃ，・・・，Ｖｆを算出する。補正値演算部１６は、該モデル風速値を基に、該類似熱帯性低気圧の風速補正値σａ，σｃ，・・・，σｆを算出する。次に、予測風速演算部１７は、前記記憶部１２に記憶されている前記ある地点における類似熱帯性低気圧の実績風速値Ｖ_Ａ，Ｖ_Ｃ，Ｖ_Ｄ，Ｖ_ＥおよびＶ_Ｆ を、前記風速補正値σａ，σｃ，・・・，σｆで補正し、これらの平均を取って、予測風速Ｖ’を求める。
その後、通信機１１で受信した現在の熱帯性低気圧特性情報（図５の熱帯性低気圧現在位置の熱帯性低気圧特性）を過去の熱帯性低気圧の特性（経路、各地点の中心気圧など）として前記記憶部１２に蓄積する。これにより、自動的に過去の熱帯性低気圧の数が多くなり、類似熱帯性低気圧の選択自由度が増加する。
【００５２】
【発明の効果】
以上の説明から明らかなように、本発明によれば、人間の経験あるいは判断によることなく、正確に、特定各地点の地域特性を考慮に入れて、熱帯性低気圧の風速を予測することができるようになる。また、このため、熱帯性低気圧の風速予測の未経験者あるいは経験の浅い者でも、容易に、熱帯性低気圧の風速予測をすることができるようになる。
【図面の簡単な説明】
【図１】 本発明が適用されるシステム例を示す図である。
【図２】 本発明の一実施形態のフローチャートである。
【図３】 提供される熱帯性低気圧特性予測情報と補間後の熱帯性低気圧特性予測情報を示す図である。
【図４】 熱帯性低気圧速度算出の説明図である。
【図５】 熱帯性低気圧風速算出の説明図である。
【図６】 予測風向き算出の説明図である。
【図７】 本発明の一実施形態の機能ブロック図である。
【符号の説明】
２・・・気象表示装置、３・・・熱帯性低気圧風速・被害予測装置、３ａ・・・熱帯性低気圧処理装置、３ｂ・・・風速・被害予測装置、４・・・ＬＡＮ、１１・・・通信装置、１２・・・記憶部、１３・・・熱帯性低気圧特性予測情報演算部、１４・・・類似熱帯性低気圧選択部、１５・・・モデル風速値演算部、１６・・・補正値演算部、予測風速演算部。[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a tropical cyclone wind speed prediction method and apparatus, and the characteristics of tropical cyclones after several hours given in advance (position, central pressure, traveling speed, etc.) Therefore, the present invention relates to a tropical cyclone wind speed prediction method and apparatus that can be predicted without human judgment.
[0002]
[Prior art]
Tropical cyclones occur in areas around the earth's equator and, in some cases, cause significant damage to the surrounding area. Conventionally, each region from the tropical cyclone center position, central atmospheric pressure, traveling speed, direction, strong winds and storms predicted several hours later by information providers such as TV and radio broadcasting weather information The wind speed of tropical cyclones in Japan was obtained with reference to past experiences of tropical cyclones, and measures such as reinforcing buildings were being implemented.
[0003]
[Problems to be solved by the invention]
However, there is a problem that it is difficult to predict the wind speed of the tropical cyclone by analyzing the difference in characteristics between the tropical cyclone experienced by humans in the past and the tropical cyclone currently approaching. . Moreover, the prediction of the wind speed by the conventional tropical cyclone largely depends on the experience of the predictor, and the prediction of the inexperienced person is extremely difficult. Another problem is that wind speed prediction based on human experience is not always accurate.
[0004]
The present invention has been made in view of the above-described problems of the prior art, and its purpose is to accurately consider the regional characteristics of each specific point without relying on human experience or judgment. An object of the present invention is to provide a tropical cyclone wind speed prediction method and apparatus for predicting a wind pressure of a cyclone.
[0005]
[Means for Solving the Problems]
In order to achieve the above-described object, the present invention provides a tropical cyclone wind speed prediction method for predicting wind speed at a certain point due to a tropical cyclone, and the tropical cyclone after t (t> 0) hours. Due to the past similar tropical cyclone at a certain point selected from past similar tropical cyclones based on the predicted values of characteristics (center position, central pressure, traveling direction, traveling speed, etc.) Applying the correction value of the wind speed calculated based on the model wind speed value at the certain point obtained by using the characteristics of the tropical cyclone and the past similar tropical cyclone to the actual wind speed value The first characteristic is that the predicted wind speed resulting from the tropical cyclone at the point is obtained.
[0006]
In addition, the present invention provides a communication device that receives tropical cyclone characteristic prediction information, a means for performing an interpolation operation on tropical cyclone characteristic prediction information received by the communication device, and a tropical property determined by the interpolation operation. Means for storing the cyclone characteristic prediction information and past tropical cyclone data, means for selecting a past similar tropical cyclone, the present tropical cyclone and the selected past similar tropical cyclone Means for calculating a model wind speed value based on the characteristic value of the atmospheric pressure; means for calculating a correction value based on the model wind speed value; the actual wind speed value at the time of the previous similar tropical cyclone and the correction value; The second feature is that a tropical cyclone wind speed predicting device is provided that includes a means for calculating the predicted wind speed of the tropical cyclone this time.
[0007]
According to the first and second features, since the predicted wind speed caused by the tropical cyclone is obtained using the characteristics of the similar tropical cyclone in the past, it is based on human experience or judgment. However, it becomes possible to accurately predict the wind speed of tropical cyclones.
[0008]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, the present invention will be described in detail with reference to the drawings. FIG. 1 is a schematic system diagram of a tropical cyclone wind speed / damage prediction system to which the tropical cyclone wind speed prediction method of the present invention is applied. The term “tropical cyclone” in the present invention includes typhoons, hurricanes, cyclones and the like.
[0009]
For example, a weather information provider 2 that displays weather news, such as the location of a tropical cyclone, and local governments, insurance companies, electric power companies, etc. that need wind speed prediction for damage prediction etc. For example, a tropical cyclone wind speed / damage prediction device 3 of Kyushu Electric Power Co., Inc. is connected by a communication line such as LAN 4. The tropical cyclone wind speed / damage prediction device 3 includes a tropical cyclone processing device 3a that mediates data and a wind speed / damage prediction device 3b that performs wind speed / damage prediction.
[0010]
When a tropical cyclone occurs, the weather display device 2 provides tropical cyclone characteristic prediction information. The tropical cyclone characteristic prediction information includes the tropical cyclone center position (latitude and longitude) and the central atmospheric pressure after 1, 6, 12, 24, 36, 48, and 72 hours from the time of issue (actual information). Is provided. The wind speed / damage prediction device 3b receives the tropical cyclone characteristic prediction information via the LAN 4, and predicts the tropical cyclone wind speed according to the tropical cyclone wind speed prediction method of the present invention.
[0011]
Below, one Embodiment of the tropical cyclone wind speed prediction method of this invention is described with reference to FIG.
FIG. 2 is a flowchart showing an outline of the tropical cyclone wind speed prediction operation of the wind speed / damage prediction apparatus 3b.
[0012]
In step S1, tropical cyclone characteristic prediction information is acquired from the weather display device 2. For example, as shown in FIG. 3 (a), the tropical cyclone center position (latitude, longitude) after 1, 6, 12, 24, 36, 48, and 72 hours from the issue time (actual information) ), And central pressure.
[0013]
In step S2, tropical cyclone characteristic prediction information (tropical cyclone center position (latitude, longitude), central atmospheric pressure, traveling speed, traveling direction) for each hour based on the received tropical cyclone characteristic prediction information. Is calculated.
[0014]
Hereinafter, a method of calculating tropical cyclone characteristic prediction information after 2 hours as shown in FIG. 3B will be described.
(1) Prediction of tropical cyclone center position (latitude, longitude) Prediction of tropical cyclone center position (latitude, longitude) is calculated using a linear interpolation method. Assuming that the time one hour after the time t _{n in the} live status information is t _{n + 1} and the time after m hours is t _{n + m} , for example, the latitude at the time t _{n + 2} two hours later is calculated by the following equation (1). .
[0015]
[Expression 1]

Here, lat (t) represents the latitude at time t. As for the longitude, the longitude after 2 hours can be calculated by the same expression as the expression (1).
[0016]
The tropical cyclone center position (latitude, longitude) of 3 hours, 4 hours, 5 hours, 7 hours, and so on is obtained by the linear interpolation method similar to the above equation (1).
(2) Prediction of tropical cyclone central pressure Predict tropical tropical cyclone central pressure using linear interpolation and k-th order moving average filter as shown in the following procedures (1) and (2). .
[0017]
(1) Linear interpolation method Assuming that the time one hour after the time t _n at the time of live information is t _{n + 1} and the time after m hours is t _{n + m} , the tropical cyclone central pressure at the time t _{n + 2} two hours later is It is calculated by the following equation (2).
[0018]
[Expression 2]

Here, hPa (t) represents the tropical cyclone central pressure at time t.
[0019]
▲ 2 ▼ k following moving average time of the filter 2 hours later and _{t n + 2,.} 1 to from the time _{t n + 2} of the (k-1) / 2 hours prior to the time _{t n + 2- (k-1} ) / 2, ·· .., T _{n + 1} , 1 to (k−1) / 2 hours later, t _{n + 3} ,..., T _{n + 2 + (k−1) / 2} , then tropical cyclone central pressure hPa ( t _{n + 2} ) ′ can be calculated by the following (3).
Here, k is an odd number.
[0020]
[Equation 3]

When there is no data for one or two hours before the actual information, the tropical cyclone central atmospheric pressure one or two hours ago is calculated using the above equation (2). In the same manner as described above, the tropical cyclone central pressure of 3 hours, 4 hours, 5 hours, 7 hours, and so on is obtained.
(3) Prediction of tropical cyclone advancement speed Prediction of tropical cyclone advancement speed is calculated by calculating the speed of each prediction point from the tropical cyclone characteristic prediction information acquired from the weather display device 2. Is calculated using a linear interpolation method and an L-th order moving average filter.
[0021]
(1) Calculation of tropical cyclone advancement speed The tropical cyclone advancement speed is calculated from the tropical cyclone center position of each prediction point as follows. For example, in FIG. 4, the tropical cyclone center position after 1 hour as tropical cyclone characteristic prediction information is A, the tropical cyclone center position after m hours is B, and the latitude line passing through A and B are passed. The intersection of the meridian is C, and the north pole is N. A spherical trigonometry is applied to this spherical triangle NAB, and the angular distance ⌒AB and distance R between AB are obtained by the following equations (4) and (5).
[0022]
[Expression 4]

R = R _e × AB (5)
Here, ⌒ the angular distance, _{R e} represents a earth radius [miles].
[0023]
Based on the distance R and the movement time t, the speed sp is obtained by the following equation (6).
[0024]
Speed sp = R / t (6)
(2) Linear Interpolation Method Next, the tropical cyclone traveling speed at time t _{n + 2} after 2 hours, _where t _{n + 1} is the time 1 hour after the time t _n in the live situation information, and t _{n +} m is the time after m hours. Is calculated by the following equation (7).
[0025]
[Equation 5]

Here, sp (t) indicates the tropical cyclone traveling speed at time t.
[0026]
In addition, the actual data of 1 to 2 hours before is extracted from the actual condition information, and if there is no data, the tropical cyclone traveling speed of 1 to 2 hours before is calculated using the equation (7).
[0027]
▲ 3 ▼ to L next moving average filter following, the time after two hours and _{t n + 2,.} 1 to from the time _{t n + 2 (L-1} ) / 2 hours prior to the time _{t n + 2- (L-1} ) / 2 , ..., t _{n + 1} , 1 to (L-1) / 2 hours later, _assuming that t _{n + 3} , ..., t _{n + 2 + (L-1) / 2} , tropical cyclone progression after 2 hours The speed sp (t _{n + 2} ) ′ can be calculated by the following equation (8).
Here, L is an odd number.
[0028]
[Formula 6]

(4) Prediction of tropical cyclone travel direction From the tropical cyclone center position calculated in (1) above, using spherical trigonometry, the following formula (9) Is calculated.
[0029]
cos∠ABC = tanBC / tanAB (9)
According to the above calculation, for example, when 225 ° <∠ABC ≦ 247.5 °, the tropical cyclone traveling direction is predicted to be northeast, and when 247.5 ° <∠ABC ≦ 270 °, it is predicted to be northeast.
[0030]
Through the above processes (1) to (4), the prediction information for every hour up to 72 hours ahead is calculated, and the process of step S2 is ended.
[0031]
Next, it progresses to step S3 of FIG. 2, and the extraction process of the past similar tropical cyclone is performed. Based on the tropical cyclone center position (latitude and longitude) of the course prediction information (prediction information after 12 hours) for wind speed prediction, past tropical cyclones in the range of latitude and longitude ± α ° Extract actual data from the database. For example, as shown in FIG. 5, the past tropical cyclone achievement points extracted from the predicted tropical cyclone center position S after 12 hours are defined as A, B, C, D, E, and F. Further, Z, which is an observation point of weather information, is set as a wind speed prediction point (for example, AMeDAS point).
[0032]
In step S4 of FIG. 2, it is determined whether a similar tropical cyclone has existed in the past. If this determination is affirmative, the process proceeds to step S5, and if negative, the process proceeds to step S11.
[0033]
In step S5, the square root (distance) of the sum of squares of the difference between the past tropical cyclone record point extracted in step S3 and the tropical cyclone center position, central atmospheric pressure, traveling speed, and traveling direction of the current tropical cyclone. ), And, for example, five tropical cyclones are selected in ascending order of distance. When selecting, for example, a past similar tropical cyclone that satisfies the following conditions is selected.
(a) Ensure that the same tropical cyclone does not overlap. (However, this does not apply if the past tropical cyclone has 5 or less similar tropical cyclones.)
(b) If the distance between the wind speed prediction point Z and the past tropical cyclone is less than r0, it is not selected as the past similar tropical cyclone. The reason for this is that if the distance between the wind speed prediction point and the past tropical cyclone is less than r0, there is a possibility that the error will enter the tropical cyclone and increase the error.
Or in the said selection, the product for every characteristic value of the past tropical cyclone and this tropical cyclone is taken, and the tropical cyclone with the big sum is made into a similar tropical cyclone.
[0034]
Alternatively, first select a similar tropical cyclone from within a certain distance from the center position of the tropical cyclone, or within a rectangular frame with a side centered on the distance from the central position of the tropical cyclone, After that, selection is made at the distance.
[0035]
Here, the selected five tropical cyclones are designated as past tropical cyclone achievement points A, C, D, E, and F.
[0036]
In step S6, the wind speed observed at the wind speed prediction point Z at the selected past tropical cyclone achievement points A, C, D, E, and F is extracted from the database. The extracted wind speed prediction actual wind speeds are defined as V _A , V _C , V _D , V _E and V _F.
[0037]
In step S7, the model wind speed value is calculated. The model wind speed value is calculated based on the tropical cyclone information of the current tropical cyclone and the selected similar tropical cyclone based on the following (10) and (11) The wind speed vector V ′ with direction is calculated using the model formula or the advanced tropical cyclone model formula of formulas (12) and (13) by acting the blowing angle θ3 (wind direction) on the wind speed of formula (11). Calculation is performed.
[0038]
V = Vmax · exp [- { (lnr) θ1 - (lnr 0) θ1} 2] ··· (10)
Here, Vmax is tropical storm maximum wind, r is Tropical Storm distance (km), θ ₁ is the attenuation amount of wind speed versus distance. The tropical cyclone maximum wind speed Vmax can be expressed by the following equation (11). In the formula (11), p is a central pressure (hPa).
[0039]
_{Vmax = a 1 × (p 1} -p) 1/2 ··· (11)
However, p ₁ is a ₁ in the constant indicating pressure is a proportionality constant.
[0040]
When the tropical cyclone distance r is less than r ₀ km, the advanced tropical cyclone model formula is (12), and when the tropical cyclone distance r is more than r ₀ km, the advanced tropical cyclone model is ( It is expressed by equation (13).
[0041]
V ′ = Si · θ ₂ · r ² / r ₀ ² · Sv · exp [− {π / (10 · r ₀ ) · r}] (12)
V ′ = Si · θ ₂ · Sv · exp [− {π / (10 · r ₀ ) · r}] (13)
Where Sv is the tropical cyclone travel speed (m / s), r is the tropical cyclone distance (km), Si is the influence rate due to the tropical cyclone position and the position of the observation point, and θ ₂ is the travel speed It is a variable that indicates the degree of reflection. Note that θ ₂ · Sv · exp (−π / 10 · r ₀ ) is Miyazaki's formula indicating the influence of the traveling speed. For example, in the paper of “Research on Typhoon Wind Speed Prediction by Various Weather Information” in 1996 It is shown.
[0042]
First, the model wind speed value V at the wind speed prediction point Z of the current tropical cyclone calculated by any one of the above formulas (10), ( 12 ) and ( 13 ) is obtained. Subsequently, the tropical cyclone central pressure in the past similar tropical cyclones A, C, D, E and F, the distance r between the wind speed prediction point Z and the tropical cyclone center position, etc. are described in (10), ( The model wind speed values Va, Vc, Vd, Ve, and Vf at the wind speed prediction points in the past similar tropical cyclones A, C, D, E, and F are obtained by applying to the equation ( 12 ) or ( 13 ).
[0043]
Next, the process proceeds to step S8 in FIG. 2, and a wind speed correction value is calculated. The correction value of the wind speed at the wind speed prediction point Z, that is, the correction value of the wind speed due to the difference of each variable such as the central pressure between the prediction information after 12 hours and the past similar tropical cyclone is the prediction after 12 hours. From the model wind speed value V of the information and the model wind speed values Va, Vc, Vd, Ve, and Vf of the past tropical cyclone, it is calculated as follows.
[0044]
Correction value σa = V−Va for the past tropical cyclone A
Correction value σc = V−Vc for the past tropical cyclone C
Correction value σd = V−Vd for the past tropical cyclone D
Correction value σe = V−Ve for the past tropical cyclone E
Correction value σf = V−Vf for the past tropical cyclone F
Next, it progresses to step S9 and the prediction wind speed in the wind speed prediction point Z is calculated. The predicted wind speeds V ′ _A , V ′ _C , V ′ _D , V ′ _E and V ′ _F at the wind speed prediction point Z after 12 hours are the wind speeds of similar tropical cyclones A, C, D, E and F in the past. The actual wind speed values V _A , V _C , V _D , V _E and V _F at the predicted point Z and the correction values σa, σc, σd, σe and σf are calculated as follows.
[0045]
V ′ _A = V _A + σa
V ′ _C = V _C + σc
V ′ _D = V _D + σd
V ′ _E = V _E + σe
V ′ _F = V _F + σf
Next, the average of these is taken as the predicted wind speed V ′ at the wind speed prediction point Z after 12 hours.
[0046]
V ′ = (V ′ _A + V ′ _C + V ′ _D + V ′ _E + V ′ _F ) / 5
Next, it progresses to step S10 of FIG. 2, and calculates a predicted wind direction. That is, as shown in FIG. 6, first, the wind direction in the tangential direction of the tropical cyclone at the wind speed prediction point Z is calculated. Next, a wind direction obtained by inclining the calculated wind direction by the blowing angle θ ₃ is set as a predicted wind direction.
[0047]
Next, the operation when the determination in step S4 in FIG. 2 is negative, that is, when there has been no similar tropical cyclone in the past, will be described. In step S11, the model wind speed value V after 12 hours at the point Z where the wind speed is predicted is calculated using the equation (10). In step S12, the predicted wind direction G is obtained as described above.
[0048]
Next, in step S13, the predicted wind speed of the predicted information after 12 hours is calculated from the following equation using the model wind speed value V and the predicted wind direction G.
[0049]
Predicted wind speed of predicted information after 12 hours = V × X (G)
Here, X (G) is a wind direction coefficient in the wind direction G.
Next, at step S14, the current tropical cyclone characteristic information (the tropical cyclone characteristic at the current position of the tropical cyclone in FIG. 5) is accumulated as the past tropical cyclone. This increases the number of past tropical cyclones and increases the degree of freedom of selection of similar tropical cyclones.
[0050]
In the above description, the predicted wind speed at the wind speed prediction point is obtained, but it is obvious that the wind speed of the tropical cyclone at any point can be obtained by applying the present invention. is there.
[0051]
Next, FIG. 7 shows a block diagram of functions related to the present invention of the wind speed / damage prediction device 3b (see FIG. 1) of the present invention. When tropical cyclone characteristic prediction information is sent through the communication line such as the LAN 4, the communication device 11 receives this information. The received information is stored in the storage unit 12. The tropical cyclone characteristic prediction information calculation unit 13 calculates the tropical cyclone characteristic prediction information for each hour based on the tropical cyclone characteristic prediction information and stores it in the storage unit 12. If the storage unit 12 stores data such as past tropical cyclone characteristics (path, central air pressure at each point, wind speed, etc.), the similar tropical cyclone selection unit 14 For example, a plurality of tropical cyclones at a certain time in the past are selected as similar tropical cyclones for the predicted center position 12 hours after the tropical cyclone. Based on the characteristic values of the current tropical cyclone and the similar tropical cyclone, the model wind speed value calculation unit 15 calculates the current tropical cyclone and the similar tropical cyclone at a certain point (for example, the wind speed prediction point). Model wind speed values V, Va, Vc,..., Vf are calculated. The correction value calculation unit 16 calculates wind speed correction values σa, σc,..., Σf of the similar tropical cyclone based on the model wind speed value. Next, the predicted wind speed calculation unit 17 uses the actual wind speed values V _A , V _C , V _D , V _E and V _F of the similar tropical cyclone stored at the certain point stored in the storage unit 12 as the wind speed. Correction is performed using correction values σa, σc,..., Σf, and an average of these values is obtained to obtain a predicted wind speed V ′.
Thereafter, the current tropical cyclone characteristic information received by the communication device 11 (the tropical cyclone characteristic at the current position of the tropical cyclone in FIG. 5) is converted into the past tropical cyclone characteristics (route, central atmospheric pressure at each point). Etc.) in the storage unit 12. Thereby, the number of past tropical cyclones automatically increases, and the degree of freedom of selection of similar tropical cyclones increases.
[0052]
【The invention's effect】
As is clear from the above description, according to the present invention, it is possible to accurately predict the wind speed of a tropical cyclone, taking into account the regional characteristics of each specific point, without relying on human experience or judgment. become able to. For this reason, even a person who has not or has little experience in predicting wind speeds in tropical cyclones can easily predict wind speeds in tropical cyclones.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating an example of a system to which the present invention is applied.
FIG. 2 is a flowchart of an embodiment of the present invention.
FIG. 3 is a diagram showing provided tropical cyclone characteristic prediction information and tropical cyclone characteristic prediction information after interpolation.
FIG. 4 is an explanatory diagram for calculating tropical cyclone velocity.
FIG. 5 is an explanatory diagram of calculation of tropical cyclone wind speed.
FIG. 6 is an explanatory diagram of predictive wind direction calculation.
FIG. 7 is a functional block diagram of an embodiment of the present invention.
[Explanation of symbols]
2 ... Weather display device, 3 ... Tropical cyclone wind speed / damage prediction device, 3a ... Tropical cyclone processing device, 3b ... Wind velocity / damage prediction device, 4 ... LAN, 11 ... Communication device, 12 ... Storage unit, 13 ... Tropical cyclone characteristic prediction information calculation unit, 14 ... Similar tropical cyclone selection unit, 15 ... Model wind speed value calculation unit, 16 ... Correction value calculation unit, predicted wind speed calculation unit.

Claims (15)

In a tropical cyclone wind speed prediction method for predicting wind speed at a certain point due to tropical cyclone,
The past similar tropical cyclone is selected using the characteristic predicted value of the tropical cyclone after t (t> 0) time, and the actual wind speed caused by the past similar tropical cyclone at the certain point By applying a correction value of the wind speed calculated based on the model wind speed value at the certain point obtained by using the characteristic of the tropical cyclone and the past similar tropical cyclone to the value, A tropical cyclone wind speed prediction method characterized in that a predicted wind speed caused by a tropical cyclone is obtained. The tropical cyclone wind speed according to claim 1, wherein the tropical cyclone characteristic after the time t is obtained by interpolating a plurality of provided tropical cyclone characteristic prediction information after a specified time. Prediction method. The characteristics of the tropical cyclone and a past similar tropical cyclone include at least one of a central position (latitude, longitude), a central atmospheric pressure, a traveling speed, and a traveling direction of the tropical cyclone. Item 3. The tropical cyclone wind speed prediction method according to Item 1 or 2. The past similar tropical cyclones are extracted from the past tropical cyclones in ascending order of the distance between the past tropical cyclone achievement points and the current tropical cyclone. The tropical cyclone wind speed prediction method according to 1 or 2. A tropical cyclone with a small sum of squared differences in characteristics between the past tropical cyclone and the current tropical cyclone is used as the previous similar tropical cyclone, or the past tropical cyclone and the present tropical cyclone. The tropical cyclone according to claim 1 or 2, characterized in that a tropical cyclone having a large sum is obtained by taking a product for each characteristic of the tropical cyclone of the tropical cyclone as the similar tropical cyclone in the past. Wind speed prediction method. 6. A frame within a certain distance from the central position of the current tropical cyclone or a rectangular frame with a certain distance centered on the central position of the present tropical cyclone set as a side, and then 3. The tropical cyclone wind speed prediction method according to claim 1 or 2, wherein the past similar tropical cyclone is selected from within the frame by a method. In the extraction of the past similar tropical cyclone, the tropical cyclone characteristic information accumulated as a result from the same similar tropical cyclone is not redundantly selected. Item 7. The tropical cyclone wind speed prediction method according to Item 4, 5 or 6. The model wind speed value is calculated using a tropical cyclone model formula, and the tropical cyclone model formula includes a central position of the tropical cyclone, a distance from the tropical cyclone center position, a maximum wind speed, and a maximum wind speed. The tropical cyclone wind speed prediction method according to claim 1, wherein at least one of a radius and a central atmospheric pressure is used as a parameter. The model wind speed value is calculated by the following formula (1) static tropical cyclone model formula, or the wind speed θ3 (wind direction) acting on the wind speed of formula (1), or formula (2), formula (3) 3. The tropical cyclone wind speed prediction method according to claim 1, wherein a wind speed vector V ′ having a direction is calculated using a progressive tropical cyclone model formula.
V = Vmax · exp [- { (lnr) θ1 - (lnr 0) θ1} 2] ··· (1)
V ′ = Si · θ ² · (r ² / r ₀ ² ) · Sv · exp [− {π · r / (10 · r ₀ )}] (2) (However, the tropical cyclone distance r is r ₀ km or less)
V ′ = Si · θ 2 · Sv · exp [− {π · r / (10 · r ₀ )}] (3) (However, the tropical cyclone distance r is greater than or equal to r ₀ km)
Here, Vmax is tropical storm maximum wind, r is Tropical Storm distance (km), _{r 0} is the maximum wind speed radius (miles), Sv is Tropical Storm traveling speed (m / s), Si tropical Is a variable indicating the degree of reflection of the traveling speed, θ1 is the amount of attenuation of the wind speed with respect to the distance, and θ2 is a variable indicating the degree of reflection of the traveling speed. The predicted wind speed caused by a tropical cyclone at the certain point is averaged based on the predicted wind speed obtained for each of the plurality of similar tropical cyclones in the past, or a weighted average based on the numerical value indicating similarity in claim 5. The tropical cyclone wind speed prediction method according to claim 1, wherein the tropical cyclone wind speed prediction method is obtained. The tropical nature according to claim 1 or 2, wherein a tangential direction of a tropical cyclone at the certain point is obtained, and a direction obtained by adding a blowing angle to the tangential direction is set as a predicted wind direction. Low pressure wind speed prediction method. When the past similar tropical cyclone does not exist, the tropical cyclone at the certain point is calculated based on the model wind speed value at the point obtained using the characteristics of the tropical cyclone and the predicted wind direction. 3. The tropical cyclone wind speed prediction method according to claim 1 or 2, wherein the predicted wind speed resulting therefrom is obtained. It is characterized by having a learning function to accumulate this tropical cyclone as the past tropical cyclone after the wind speed prediction by this tropical cyclone or after passing this tropical cyclone The tropical cyclone wind speed prediction method according to claim 1. In a tropical cyclone wind speed prediction device that predicts wind speed at a certain point caused by tropical cyclone,
A communication device that receives tropical cyclone characteristic prediction information;
Means for interpolating tropical cyclone characteristic prediction information received by the communication device;
Means for storing tropical cyclone characteristic prediction information obtained by the interpolation calculation and past tropical cyclone data;
A means to select past similar tropical cyclones,
Means for calculating a model wind speed value based on the characteristic value of the tropical cyclone of this time and the selected similar tropical cyclone in the past,
Means for calculating a correction value based on the model wind speed value;
A tropical cyclone wind speed prediction apparatus comprising means for calculating a predicted wind speed of the current tropical cyclone from the past actual wind speed value at the time of the similar tropical cyclone and the correction value. 15. The tropical cyclone wind speed prediction according to claim 14, wherein the current tropical cyclone of claim 5 is additionally stored in a storage area of the means for storing the past tropical cyclone data. apparatus.

Images