Information processing device, information processing method, and program

Abstract

To accurately perform long-term weather prediction, including extreme high or low temperatures, heavy rainfall and water shortage in one or more months ahead as well as the end of the rainy season, the number of approaching typhoons and the like.SOLUTION: An index acquisition unit 51 acquires an index for a remote region as prediction data (index) from an index computation device 2 that executes a numerical prediction model that calculates an index focusing on a weather phenomenon correlated between a prediction region and the remote region, and provides the index to a weather prediction unit 52. The weather prediction unit 52 performs long-term weather prediction in the prediction region on the basis of the index (prediction data) of the remote region. The weather prediction unit 52 uses a regression coefficient 61 of a prediction equation obtained from predetermined machine learning to execute the long-term weather prediction of the prediction region. The regression coefficient 61 of the prediction equation is generated or updated by a learning unit 54.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present invention relates to an information processing apparatus, an information processing method, and a program.

Background Art

[0002] Long-term weather forecasts for more than one month ahead (see, for example, Patent Document 1) are in high demand in various fields and are also disclosed in, for example, Patent Document 1.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, there are physical limitations in long-term weather prediction. The weather prediction in the mid-high latitudes by numerical weather prediction models has low accuracy, and it is impossible to predict extreme high temperatures, low temperatures, heavy rain, and droughts more than one month ahead. Therefore, the guidance that supports the forecasting work by correcting the grid point values directly above the prediction target area in the numerical weather prediction model, as used by the Japan Meteorological Agency, has low accuracy more than one month ahead. In addition, with such guidance using only the grid point values directly above the numerical weather prediction model, it is impossible to represent the timing of the end of the rainy season and the number of approaching typhoons, which have a great impact on the economy.

[0005] The present invention has been made in view of such a situation, and an object thereof is to accurately perform long-term weather prediction including extreme high temperatures, low temperatures, heavy rain, and droughts more than one month ahead, as well as the timing of the end of the rainy season and the number of approaching typhoons.

Means for Solving the Problems

[0006] To achieve the above object, an information processing apparatus according to an aspect of the present invention is Index acquisition means for acquiring the index of the remote area from calculation means for calculating an index focusing on a weather phenomenon correlated between a prediction area and the remote area; Weather prediction means for performing long-term weather prediction in the prediction area based on the index of the remote area; and comprises.

Advantages of the Invention

[0007] According to the present invention, long-term weather prediction can be accurately performed, including extreme high temperatures, low temperatures, heavy rains, and droughts more than one month in advance, as well as the timing of the end of the rainy season and the number of approaching typhoons.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Modes for Carrying Out the Invention

[0009] Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing a configuration example of an information processing system including a weather prediction apparatus as an embodiment of the information processing apparatus of the present invention.

[0010] The information processing system shown in FIG. 1 is configured by connecting a weather prediction apparatus 1, an index calculation apparatus 2 using a numerical weather prediction model, and a weather observation data providing apparatus 3 for a prediction target to each other via a predetermined network such as the Internet.

[0011] The index calculation apparatus 2 using a numerical weather prediction model is an information processing apparatus that calculates an index focusing on a weather phenomenon correlated between a prediction area and a remote area. The weather prediction apparatus 1 is an embodiment of the information processing apparatus to which the present invention is applied, and is an information processing apparatus that performs long-term weather prediction in a prediction area based on an index of a remote area calculated by the index calculation apparatus 2 using a numerical weather prediction model.

[0012] Here, long-term weather prediction means weather prediction in a predetermined period more than one month ahead, and in one aspect, it may be two months or more ahead, and in one aspect, it may be three months or more ahead. The prediction area is an area where long-term weather prediction is performed, and for example, in addition to Japan, the United States, Europe, etc. are assumed. The remote area is an area that is distanced from the prediction area to such an extent that the weather itself in the same time zone has no relevance. For example, if the prediction area is Japan, the subtropical and tropical Indian Ocean and near the Philippines become the remote area. Also, for example, if the prediction area is the United States, the El Niño monitoring sea area is the remote area.

[0013] Specifically, for example, the index calculation apparatus 2 using a numerical weather prediction model calculates the intensity of convective activity near the Indian Ocean and the Philippines as an index. The weather prediction apparatus 1 predicts the sunshine hours in Japan for a predetermined month (for example, July for the March time point) more than one month ahead from a predetermined time point based on the index. For example, the exponential operation device 2 using the numerical weather prediction model calculates using the change in the sea surface temperature in the El Niño monitoring sea area as an indicator. The weather prediction device 1 predicts the precipitation in the United States for a predetermined month (for example, July for the March time point) more than one month after a predetermined time point based on the said indicator. For a specific example of the long-term weather prediction in the prediction area using such an indicator in a remote area, refer to FIGS. 4 to 7 and it will be described later.

[0014] Here, the weather prediction device 1 of the present embodiment can perform long-term weather prediction in the prediction area using a model (for example, the regression coefficient 61 of the prediction formula in FIG. 3 described later) obtained as a result of predetermined machine learning. When a model is generated or updated by such predetermined machine learning, the weather observation data providing device 3 for the prediction target provides the actual weather observation data in the prediction area as part of the learning data. Regarding the details of the predetermined machine learning, it will be described later as the explanation of the regression coefficient 61 of the prediction formula in FIG. 3. Note that the regression coefficient of the prediction formula here is not limited to the calculation method.

[0015] FIG. 2 is a block diagram showing an example of the hardware configuration of the weather prediction device in the information processing system shown in FIG. 1.

[0016] The CPU 11 executes various processes according to the programs recorded in the ROM 12 or the programs loaded from the storage unit 18 to the RAM 13. In the RAM 13, data and the like necessary for the CPU 11 to execute various processes are also appropriately stored.

[0017] The CPU 11, ROM 12, and RAM 13 are interconnected via the bus 14. An input / output interface 15 is also connected to this bus 14. An input unit 16, an output unit 17, a storage unit 18, a communication unit 19, and a drive 20 are connected to the input / output interface 15.

[0018] The input unit 16 is composed of, for example, a keyboard or the like and inputs various information. The output unit 17 is composed of a display such as a liquid crystal display, a speaker, etc., and outputs various information as images and sounds. The storage unit 18 is composed of a DRAM (Dynamic Random Access Memory), etc., and stores various data. The communication unit 19 communicates with other devices (for example, the exponentiation operation device 2 based on the numerical weather prediction model and the weather observation data providing device 3 for the prediction target in FIG. 1) via a network N including the Internet.

[0019] A removable medium 30 made of a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, etc. is appropriately mounted on the drive 20. The program read from the removable medium 30 by the drive 20 is installed in the storage unit 18 as necessary. In addition, the removable medium 30 can also store various data stored in the storage unit 18 in the same manner as the storage unit 18.

[0020] Although not shown in the figure, the exponentiation operation device 2 based on the numerical weather prediction model and the weather observation data providing device 3 for the prediction target in FIG. 1 can also have a configuration basically the same as the hardware configuration shown in FIG. 2. Therefore, the description of the hardware configuration of the exponentiation operation device 2 based on the numerical weather prediction model and the weather observation data providing device 3 for the prediction target is omitted.

[0021] By the cooperation of various hardware and various software that make up the information processing system of FIG. 1 including the weather prediction device 1 of FIG. 2, a long-term weather prediction in the prediction area can be executed based on an index focusing on weather phenomena that are correlated between the prediction area and the remote area.

[0022] FIG. 3 is a functional block diagram showing an example of the functional configuration of the weather prediction device 1 of FIG. 2 in the information processing system of FIG. 1.

[0023] As shown in FIG. 3, in the CPU 11 of the weather prediction device 1, an index acquisition unit 51, a weather prediction unit 52, a weather observation data acquisition unit 53, and a learning unit 54 function. In an area of the storage unit 18, a regression coefficient 61 of a prediction formula is arranged.

[0024] The index acquisition unit 51 acquires, as prediction data (index), the index of the remote area from the index calculation device 2 using a numerical weather prediction model that calculates an index focusing on weather phenomena correlated between the prediction area and the remote area, and provides it to the weather prediction unit 52. The weather prediction unit 52 performs long-term weather prediction in the prediction area based on the index (prediction data) of the remote area.

[0025] The weather prediction unit 52 of the present embodiment performs long-term weather prediction in the prediction area using the regression coefficient 61 of the prediction formula obtained as a result of predetermined machine learning. The prediction formula is represented by Y = F(X, a), where Y is the weather observation data to be predicted, X is the index of the numerical weather prediction model, and a is the regression coefficient. The regression coefficient 61 of this prediction formula is generated or updated by the learning unit 54. Therefore, the functional configuration in the learning phase will be described below.

[0026] The index acquisition unit 51 acquires, as learning data (index), simulation data of the index of the remote area up to a predetermined period later by a past numerical weather prediction model from the index calculation device 2 using the numerical weather prediction model, and provides it to the learning unit 54. Specifically, for example, the index acquisition unit 51 acquires, as learning data (index), simulation data of the sea surface water temperature and convective activity in the remote area up to 7 months later by a past numerical weather prediction model from the index calculation device 2 using the numerical weather prediction model, and provides it to the learning unit 54.

[0027] The weather observation data acquisition unit 53 acquires the actual weather observation data in the prediction area as learning data from the weather observation data providing device 3 for the prediction target, and provides it to the learning unit 54.

[0028] The learning unit 54 analyzes the relationship between the learning data (index) acquired by the index acquisition unit 51 and the learning data (actual meteorological observation data) acquired by the meteorological observation data acquisition unit 53 by a predetermined machine learning, and generates or updates the regression coefficient 61 of the prediction formula as the relational expression that explains the result of the long-term meteorological observation in the prediction area by the index in the remote area.

[0029] Thereby, in the prediction / operation phase, the index acquisition unit 51 and the weather prediction unit 52 function as described above. Specifically, for example, when the index acquisition unit 51 acquires the prediction of the sea surface temperature and convective activity in the remote area up to 7 months ahead by a new numerical weather prediction model as the prediction data (index), it provides it to the weather prediction unit 52. The weather prediction unit 52 incorporates the index (prediction data) in the remote area into the regression coefficient 61 (relational expression) of the prediction formula and performs long-term weather prediction in the prediction area.

[0030] Next, several examples will be described as specific examples of the long-term weather prediction performed by the weather prediction apparatus having the functional configuration of FIG. 3. FIGS. 4 to 7 are diagrams showing the first to fourth specific examples of the long-term weather prediction performed by the weather prediction apparatus having the functional configuration of FIG. 3, respectively.

[0031] In FIG. 4, the result 101 of the long-term weather prediction in the first example to which the present embodiment is applied (hereinafter referred to as "the result 101 of the first example of the present embodiment") shows the precipitation amount (percentage compared to the average year) in July predicted at the time of January of each year (from 1993 to 2016) with the western Pacific side of Japan as the prediction area. In the result 101 of the first example of the present embodiment, as the index in the remote area, the prediction of the convective activity in the subtropical and tropical Indian Oceans and near the Philippines by the numerical weather prediction model is adopted. In the result 101 of the first example of the present embodiment, the thick solid line indicates the observation (the value of the actual meteorological observation data). The thin solid line indicates the prediction (the result of the long-term weather prediction). The broken line indicates the 70% prediction interval.

[0032] Figure 4 shows the results 102 of long-term weather prediction of the conventional method (prediction using the precipitation amount at the grid point directly above the numerical prediction model as an indicator) for comparison with the results 101 of the first case of this embodiment (hereinafter referred to as "conventional method results 102").

[0033] In the conventional method results 102, it was not possible to predict extreme heavy rain or little rain six months ago. However, in the results 101 of the first case of this embodiment, it can be seen that the number of cases where extreme heavy rain or little rain can be predicted has increased. Also, in the results 101 of the first case of this embodiment, it can be seen that the range of assumed errors can now be grasped depending on the prediction interval.

[0034] In Figure 5, the results 103 of long-term weather prediction for the second case to which this embodiment is applied (hereinafter referred to as "results 103 of the second case of this embodiment") show the predicted date of the end of the rainy season (normal year difference) at the time of January for each year (from 1993 to 2022) with the Kanto-Koshinetsu region of Japan as the prediction area. In the results 103 of the second case of this embodiment, as an indicator of a remote area, the prediction of convective activities in the subtropical and tropical Indian Ocean and near the Philippines by a numerical weather prediction model is adopted. In the results 103 of the third case of this embodiment, the thick solid line indicates the observation (the value of the actual weather observation data), the thin solid line indicates the prediction (the results of long-term weather prediction), and the dashed line indicates the 70% prediction interval.

[0035] Although not shown in Figure 5, in the guidance of the conventional method (the method using the grid point value index directly above the numerical prediction model), the "time" of the end of the rainy season could not be expressed. On the other hand, as shown in the results 103 of the second case of this embodiment, it can be seen that by applying the weather prediction device 1 of this embodiment, it has become possible to predict the time such as the end of the rainy season.

[0036] In Figure 6, the results 104 of long-term weather prediction for the third case to which this embodiment is applied (hereinafter referred to as "results 104 of the third case of this embodiment") show the number of typhoons predicted in September at the time of June for each year (from 1993 to 2023) with Japan as the prediction area. The result 105 of the long-term weather forecast for the third case to which this embodiment is applied (hereinafter referred to as "the result 105 of the third case of this embodiment") shows the number of typhoons approaching the main island of Japan in September predicted at the time of June for each year (from 1993 to 2023), with Japan as the prediction area. Here, the main island means Hokkaido, Honshu, Shikoku, and Kyushu. In the results 104 and 105 of the third case of this embodiment, as an index for a remote area, predictions of convective activities in the subtropical and tropical Indian Ocean and near the Philippines by a numerical weather prediction model are adopted. In the results 104 and 105 of the third case of this embodiment, the thick solid line indicates the observation (the value of actual weather observation data), the thin solid line indicates the prediction (the result of the long-term weather forecast), and the dashed line indicates the 70% prediction interval.

[0037] As shown in the results 104 and 105 of the third case of this embodiment, it can be seen that by applying the weather prediction device 1 of this embodiment, the tendency of the number of typhoons generated and approaching, whether it is large or small, can be predicted.

[0038] In FIG. 7, the performance 106 shown for comparison with the fourth case (hereinafter referred to as "the performance 106 of the fourth case") shows the sunshine hours in July for each year (from 2019 to 2022) with Japan as the prediction area. The result 107 of the long-term weather forecast for the fourth case to which this embodiment is applied (hereinafter referred to as "the result 107 of the fourth case of this embodiment") shows the sunshine hours in July predicted at the time of April for each year (from 2019 to 2022) with Japan as the prediction area. In the result 107 of the fourth case of this embodiment, as an index for a remote area, predictions of convective activities in the subtropical and tropical Indian Ocean and near the Philippines by a numerical weather prediction model are adopted.

[0039] Comparing the performance 106 of the fourth case with the result 107 of the fourth case of this embodiment, it can be seen that by applying the weather prediction device 1 of this embodiment, the regional tendencies of sunny and cloudy areas can be generally grasped.

[0040] Although one embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment, and modifications, improvements, etc. within the scope that can achieve the object of the present invention are considered to be included in the present invention.

[0041] For example, in the cases shown in FIGS. 4 to 7, the prediction region is Japan, and the index of the remote region is the convective activity in the subtropical and tropical Indian Ocean and near the Philippines, but it is not particularly limited thereto. Specifically, for example, as described above, the United States may be adopted as the prediction region, and the sea surface temperature in the El Niño monitoring sea area may be adopted as the index of the remote region.

[0042] For example, the hardware configuration of the weather prediction device 1 shown in FIG. 2 is merely an example for achieving the object of the present invention and is not particularly limited.

[0043] Also, the functional block diagram shown in FIG. 3 is merely an example and is not particularly limited. That is, it is sufficient to have a function capable of executing the above-described processing as a whole, and the functional blocks used for realizing this function are not particularly limited to the example of FIG. 3.

[0044] Also, the locations of the functional blocks and the database are not limited to FIG. 3 and may be arbitrary. Specifically, for example, the learning unit 54, the index acquisition unit 51 for learning data acquisition, and the weather observation data acquisition unit 53 may be transferred to a dedicated learning information processing device (not shown). Also, for example, the function of the index calculation device 2 by the numerical weather prediction model and the weather prediction unit 52 may be provided in one information processing device (not shown). Also, for example, the regression coefficient 61 of the prediction formula may be stored in the storage device of another information processing device (not shown).

[0045] Also, the above-described series of processes can be executed by hardware or by software. In addition, one functional block may be configured by hardware alone, software alone, or a combination thereof.

[0046] When a series of processes are to be executed by software, the program constituting the software is installed in a computer or the like from a network or a recording medium. The computer may be a computer incorporated in dedicated hardware. Also, the computer may be a computer capable of executing various functions by installing various programs, for example, a general-purpose smartphone or personal computer in addition to a server.

[0047] A recording medium containing such a program is not only constituted by a removable medium (not shown) distributed separately from the apparatus main body for providing the program to the user, but also constituted by a recording medium or the like provided to the user in a state pre-installed in the apparatus main body.

[0048] Note that, in this specification, the step of describing a program recorded on a recording medium includes not only processes performed in time series according to the order thereof, but also processes that are not necessarily processed in time series, but are executed in parallel or individually.

[0049] To summarize the above, it suffices for the information processing apparatus to which the present invention is applied to have the following configuration, and various embodiments can be adopted. That is, the information processing apparatus (for example, the weather prediction apparatus 1 in FIG. 1) to which the present invention is applied an index acquisition means (for example, the index acquisition unit 51 in FIG. 3) that acquires the index of the remote region from a calculation means (for example, the index calculation device 2 by the numerical weather prediction model in FIG. 1) that calculates an index (for example, convective activity in the vicinity of the Indian Ocean and the Philippines in the subtropical and tropical regions) focusing on a meteorological phenomenon correlated between a prediction region (for example, Japan) and a remote region (for example, the subtropical and tropical Indian Ocean and the vicinity of the Philippines); Weather prediction means (for example, the weather prediction unit 52 in FIG. 3) that performs long-term weather prediction (for example, the long-term weather prediction shown in FIGS. 4 to 7) in the prediction area based on the index in the remote area; An information processing device having the above is sufficient.

[0050] As a result, it becomes possible to accurately perform long-term weather prediction including extreme high temperatures, low temperatures, heavy rain, and droughts more than one month in advance, as well as the timing of the end of the rainy season and the number of approaching typhoons.

[0051] Here, for example, the index can be at least one of convective activity and sea surface temperature in the sea in the remote area.

[0052] Further, the information processing device By analyzing the relationship between the simulation data of the index in the remote area up to a predetermined period by a past numerical weather prediction model and the actual weather observation data in the prediction area by a predetermined machine learning, a model (for example, the regression coefficient 61 of the prediction formula in FIG. 3) that explains the result of long-term weather observation in the prediction area by the index in the remote area is generated. The information processing device further includes learning means (for example, the learning unit 54 in FIG. 3), The index acquisition means acquires an estimate of the index in the remote area up to the predetermined period by a new numerical weather prediction model, The weather prediction means inputs the acquired estimate of the index in the remote area into the model, and uses the output of the model as the long-term weather prediction in the prediction area. It can be like this.

Explanation of symbols

[0053] 1... Weather prediction device, 2... Exponential operation device using numerical weather prediction model, 3... Weather observation data providing device for prediction target, 11... CPU, 12... ROM, 13... RAM, 14... Bus, 15... Input / output interface, 16... Input unit, 17... Output unit, 18... Storage unit, 19... Communication unit, 20... Drive, 30... Removable media, 51... Index acquisition unit, 52... Weather prediction unit, 53... Weather observation data acquisition unit, 54... Learning unit, 61... Regression coefficient of prediction formula

Claims

1. An index acquisition means for acquiring the index of the remote area from a calculation means for calculating an index focusing on a weather phenomenon correlated between a prediction area and the remote area, A weather prediction means for performing a long-term weather prediction in the prediction area based on the index of the remote area, An information processing apparatus comprising:

2. The index of the remote area is at least one of convective activity and sea surface water temperature in the sea of the remote area, The information processing apparatus according to claim 1.

3. Further comprising a learning means for generating a model for explaining the result of long-term weather observation in the prediction area by the index of the remote area by analyzing the relationship between simulation data of the index of the remote area up to a predetermined period by a past numerical weather prediction model and actual weather observation data of the prediction area by a predetermined machine learning, The index acquisition means acquires an expectation of the index of the remote area up to the predetermined period by a new numerical weather prediction model, The weather prediction means inputs the acquired expectation of the index of the remote area into the model and uses the output of the model as the long-term weather prediction in the prediction area, The information processing apparatus according to claim 1 or 2.

4. In an information processing method executed by an information processing apparatus, An index acquisition step of acquiring the index of the remote area from a calculation means for calculating an index focusing on a weather phenomenon correlated between a prediction area and the remote area, A weather prediction step of performing a long-term weather prediction in the prediction area based on the index of the remote area, An information processing method including:

5. A computer, An index acquisition step of acquiring the index of the remote area from a calculation means for calculating an index focusing on a weather phenomenon correlated between a prediction area and the remote area, A weather prediction step of performing a long-term weather prediction in the prediction area based on the index in the remote area; A program for executing control processing including the above.

Images