JP4818079B2 - Weather forecast data analysis apparatus and weather forecast data analysis method - Google Patents

Description

  The present invention relates to a weather prediction data analysis apparatus and a weather prediction data analysis method for analyzing prediction data output from a weather prediction system.

In a conventional weather prediction system, weather prediction is performed by calculating the atmospheric flow using observation data obtained by a weather radar or the like, or GPV (Grid Point Value) data provided by the Japan Meteorological Agency. Weather forecast information is not only familiar to people but also important information related to life and property, such as typhoons and torrential rains, so that the reliability of forecast data is improved. For example, a method has been proposed that allows the accuracy of prediction data to be continuously maintained even when the time interval for obtaining observation data is long (see, for example, Patent Document 1 or 2).
JP 2004-109001 A JP 2003-090888 A

  However, the weather prediction information is only a prediction, and has uncertainties accompanied by temporal and spatial deviations. However, the value of prediction information exists. For example, if the possibility of the occurrence of torrential rain is predicted in advance, the information is sufficiently valuable even if the possibility is low. In other words, the way of providing weather forecast information determines the value, and there is a problem in the providing method.

  The present invention has been made paying attention to the above circumstances, and its object is to provide a weather prediction data analysis apparatus and a weather prediction capable of providing an accurate index for judging various effects due to weather phenomena. It is to provide a data analysis method.

  In order to achieve the above object, a weather prediction data analysis apparatus according to the present invention includes weather observation data and acquisition means for acquiring weather prediction data calculated based on a weather prediction model based on the weather observation data; Storage means for storing a conditional probability table (CPT) representing a conditional probability in a dependency relationship of a plurality of weather judgment elements for a specific weather phenomenon; and the obtained weather forecast data and weather observation data; And determining means for determining each of a plurality of weather determination elements, and calculating the occurrence probability of the specific weather phenomenon based on the conditional probability table based on the determination result of each of the determined weather determination elements And a notification means for notifying the probability information of the specific weather phenomenon when the calculated occurrence probability exceeds a threshold value. To.

  Further, the weather prediction data analysis method obtains weather observation data and weather prediction data calculated based on a weather prediction model based on the weather observation data, and a plurality of weather determination elements for a specific weather phenomenon. A conditional probability table representing conditional probabilities in the dependency relationship is stored, a plurality of weather determination elements are determined based on the acquired weather prediction data and weather observation data, and each of the determined weather determination elements is determined. Based on the determination result, the occurrence probability of the specific weather phenomenon is calculated based on the conditional probability table, and the probability information of the specific weather phenomenon is notified when the calculated occurrence probability exceeds a threshold value It is characterized by doing.

  The weather prediction data analysis apparatus and the weather prediction data analysis method configured as described above have a conditional probability table that represents conditional probabilities in the dependency relationships of a plurality of weather determination elements for a specific weather phenomenon. Each weather determination element is determined based on the weather prediction data and observation data calculated based on the weather prediction model, and a specific weather phenomenon is obtained based on the conditional probability table based on the determination result. Probability is calculated. In this way, for example, it is possible to notify the probability of a heavy rain that causes a hazard such as flood damage, etc., so that it is possible to provide an accurate index in determining various effects due to weather phenomena It becomes possible.

  Therefore, according to the present invention, it is possible to provide a weather prediction data analysis apparatus and a weather prediction data analysis method capable of providing an accurate index for judging various effects due to weather phenomena.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a block diagram showing a configuration of an embodiment of a weather prediction system according to the present invention. This weather prediction system is connected to the Japan Meteorological Agency data server DS0 and radar site servers DS1, DS2 via a network NT. The weather prediction system includes a communication interface 12 connected to the network NT, a communication processing unit 11, an observation data storage unit 13, a weather prediction model calculation unit 14, a prediction data storage unit 15, and a prediction data analysis unit 16. The data determination / distribution unit 17 is provided.

  The communication processing unit 11 obtains observation data / prediction data (GPV data) based on weather prediction from the Meteorological Agency data server DS0 and the radar site servers DS1, DS2 via the network NT via the communication interface 12. The weather observation data obtained by the communication processing unit 11 is stored in the observation data storage unit 13 and selectively sent to the weather prediction model calculation unit 14 in response to a request from the weather prediction model calculation unit 14.

  The weather prediction model calculation unit 14 is activated when the data that is the basis of the weather prediction is stored in the observation data storage unit 13, and performs the weather prediction calculation. The obtained weather forecast data is stored in the forecast data storage unit 15. In addition, when new observation data is input to the observation data storage unit 13, the weather prediction model calculation unit 14 starts again and re-executes the weather prediction calculation in order to correct the difference between the observation and the prediction.

  Further, the prediction value output from the weather prediction model calculation unit 14 is input to the prediction data analysis unit 16. The prediction data analysis unit 16 holds a conditional probability table (CPT) that represents a conditional probability in a dependency relationship among a plurality of weather determination elements for a specific weather phenomenon. The predicted data analysis unit 16 calculates a probability value of a specific weather phenomenon based on the input predicted value based on the conditional probability table. Details of the probability calculation process will be described later. The calculated probability value is supplied to the data determination / distribution unit 17 and is converted into probability information as necessary, and then notified to the user by e-mail or the like.

Next, the operation of the prediction data analysis unit 16 that is a feature of the present invention will be described.
For the probability calculation process, a Bayesian Network method is used. A Bayesian network is a method of counting the interaction of probabilities for a chain of uncertain events, and its application is spreading to cognitive science, data mining, etc. as a “computation model for handling uncertainty”. An example of a Bayesian network in the present invention is shown in FIG. In the present embodiment, the occurrence probability of heavy rain is obtained as an example.

  In any judgment area / time, (1) it is judged whether or not heavy rain is predicted by the weather forecast model, and at the same time, the weather forecast model such as the heavy rain forecast of the weather forecast model is shifted in time and space. (5) Calculate the probability of torrential rain. For this purpose, for example, the following nodes (2) to (4) are determined.

(2) Whether there is heavy rain in the surrounding area
At the determination time, the meteorological radar observation data around the determination area is searched to determine whether there is heavy rain that is likely to approach the determination area.

(3) Whether the predicted atmospheric conditions are unstable
Generally, when local heavy rain falls, the atmospheric state is often unstable globally. In such a case, although the weather prediction model is calculated including various other factors, the position and time may be shifted. Therefore, by determining whether the atmospheric state predicted by the weather prediction model is unstable, it is determined whether there is a possibility of heavy rain in the determination region, and this determination result is used as a determination material for calculating the heavy rain probability. To do. Specifically, even if heavy rain is not predicted in the determination region, the heavy rain probability is high if the atmospheric state is unstable. Even if heavy rain is predicted in the determination area, the probability of heavy rain is low unless the entire atmosphere is unstable.

(4) Whether the accuracy of the weather forecast model is high
The accuracy of weather forecast models changes from day to day. In other words, there are days when the prediction is true and not. Even if heavy rain is predicted in the judgment area, if the prediction accuracy of the day is bad, the heavy rain prediction is not reliable. By taking this feature into consideration, it becomes a judgment material for calculating the heavy rain probability in the judgment area. In this determination, rainfall data and weather radar data in the weather forecast data are used.

  The above (1) to (4) are evidence nodes that can be determined at the determination time, and the heavy rain probability is determined based on the probability value corresponding to the conditional probability table of FIG. Even if the evidence node is not determined, the heavy rain probability value is determined by the calculation process using the Bayes' theorem using the probability table of each node and the conditional probability table of FIG.

  The heavy rain probability value calculated in this manner is notified to the user by e-mail or the like as probability information when it is determined by the data determination / distribution unit 17 to be equal to or greater than a predetermined threshold. FIG. 5 shows an example of probability information distributed by the data determination / distribution unit 17. Moreover, the prediction data analysis part 16 can determine by the arbitrary time width about the prediction time part which the prediction data calculated by the weather prediction model calculation part 14 have. For example, as shown in FIG. 5, probabilities such as within 1 hour, within 3 hours, and within 12 hours can be calculated.

  As described above, in the above-described embodiment, the prediction data analysis unit 16 holds a conditional probability table that represents the conditional probabilities in the dependency relationships of a plurality of weather determination elements for a specific weather phenomenon (for example, heavy rain). . The prediction data analysis unit 16 acquires the weather observation data and the weather prediction data calculated by the weather prediction model calculation unit 14, and a plurality of weather determination elements based on the acquired weather prediction data and the weather observation data. Judge each one. The probability of heavy rain is calculated based on the conditional probability table based on the determination result of each determined weather determination element. The data judgment / distribution unit 17 notifies the probability information of heavy rain when the calculated probability of heavy rain exceeds a threshold value.

  Therefore, according to the above-described embodiment, it is possible to provide an accurate index for judging various effects due to weather phenomena while eliminating uncertainty of weather forecast data as much as possible. For example, by notifying the probability of heavy rain, the value of supporting the judgment for users who need to judge the risk of weather phenomena, such as evacuation guidance for residents, operation of dam water discharge, and operation of pumping stations in urban sewers High information can be provided.

  The present invention is not limited to the above embodiment. In the above-described embodiment, the calculation method of the probability value of heavy rain is described as an example, but it can also be applied to the calculation of the probability value of the integrated rainfall amount that causes a sediment disaster, for example. The determinations (1) to (4) described above are made by using numerical data and comparing with each threshold value. By changing the threshold value, it is possible to easily calculate probability values such as “strong rain” and “precipitation”.

  The Bayesian network shown above is merely an example, and it is of course possible to calculate probability values related to weather disaster prevention such as “wind”, “wind / rain”, and “snowfall” by constructing other Bayesian networks. However, it is the same in that an event that supports the event to be predicted is included in the node. Furthermore, you may make it update a conditional probability table | surface using the weather observation data actually obtained. In this way, it is possible to calculate a probability value with a higher hit rate.

  In short, the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. Further, various inventions can be formed by appropriately combining a plurality of constituent elements disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment. Furthermore, you may combine suitably the component covering different embodiment.

The block block diagram which shows one Embodiment of the weather prediction system concerning this invention. The figure which shows an example of the Bayesian network for heavy rain prediction in the prediction data analysis part 16. FIG. The figure which shows an example of the conditional probability table corresponding to the Bayesian network of FIG. The figure which shows an example of the probability table | surface of each node. The figure which shows an example of the delivery information delivered from the data determination / distribution part 17. FIG.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 11 ... Communication processing part, 12 ... Communication interface, 13 ... Observation data storage part, 14 ... Weather prediction model calculating part, 15 ... Prediction data storage part, 16 ... Prediction data analysis part, 17 ... Data determination / delivery part, NT ... Network, DS0 ... Japan Meteorological Agency data server, DS1, DS2 ... Radar site data server.

Claims (10)

An acquisition means for acquiring weather observation data;
A computing means for computing weather forecast data relating to a specific weather phenomenon based on a weather forecast model based on the weather observation data;
Storage means for storing: (Conditional Probability Table CPT), the conditional probability table representing the conditional probability in dependence of a plurality of meteorological determination factors for the particular weather phenomena
Respectively determining means for authenticity for the plurality of meteorological determination elements on the basis of said meteorological data and the weather forecast data,
Probability calculation means for calculating the occurrence probability of the specific weather phenomenon based on the conditional probability table based on the determination result of each determined weather determination element;
Notification means for notifying the probability information of the specific weather phenomenon when the calculated occurrence probability exceeds a threshold ,
The weather prediction data analysis apparatus , wherein the weather determination element is determined in consideration of a temporal and spatial shift caused by the weather prediction model .   The weather prediction data analysis apparatus according to claim 1, wherein the conditional probability table is created based on a Bayesian network configured by dependency relationships between the plurality of weather determination elements.   The weather prediction data analysis apparatus according to claim 1, further comprising an update unit that updates the value of the conditional probability based on the weather observation data acquired by the acquisition unit.   When there is a weather determination element that cannot be determined by the determination means, the probability calculation means uses the probability value of the weather determination element prepared in advance and the conditional probability table to determine the specific weather phenomenon. The weather prediction data analysis apparatus according to claim 1, wherein an occurrence probability is calculated.   The weather prediction according to claim 1, wherein the probability calculation means calculates the occurrence probability of the specific weather phenomenon at arbitrary time intervals for the weather prediction data for the prediction time acquired by the acquisition means. Data analysis device. Get the weather observation data,
Calculate weather forecast data related to a specific weather phenomenon based on the weather forecast model based on the weather observation data,
Storing the conditional probability table representing the conditional probability in dependence of a plurality of meteorological determination factors for the particular weather events,
The authenticity determining respective for the plurality of meteorological determination factors based on said weather forecast data of said meteorological data,
Calculate the occurrence probability of the specific weather phenomenon based on the conditional probability table based on the determination result of each determined weather determination element,
Notifying the probability information of the specific weather phenomenon when the calculated occurrence probability exceeds a threshold ,
The weather prediction data analysis method , wherein the weather determination element is determined in consideration of a temporal and spatial shift caused by the weather prediction model .   The weather prediction data analysis method according to claim 6, wherein the conditional probability table is created based on a Bayesian network configured by dependency relationships between the plurality of weather determination elements.   The weather prediction data analysis method according to claim 6, further comprising updating the conditional probability value based on the acquired weather observation data.   When there is a weather determination element that cannot be determined in the determination, calculating a probability of occurrence of the specific weather phenomenon using a probability value of the weather determination element prepared in advance and the conditional probability table The meteorological prediction data analysis method according to claim 6, wherein:   The weather forecast data analysis method according to claim 6, wherein the occurrence probability of the specific weather phenomenon is calculated at arbitrary time intervals for the weather forecast data for the obtained forecast time.

Images