Environmental load prediction system and environmental load prediction method
The environmental load prediction system addresses the inability of existing methods to predict future CO2 absorption by identifying key factors and generating models to forecast CO2 absorption amounts, enhancing the management of carbon credit projects.
Patent Information
- Application Number
- JP2024109547
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-07-08
- Publication Date
- 2026-01-21
AI Technical Summary
Existing methods, such as those described in Patent Document 1, can calculate current CO2 absorption amounts but fail to predict future CO2 absorption amounts accurately, and are influenced by external factors like temperature fluctuations and pests.
An environmental load prediction system that includes a storage device for environmental parameters and influence factors, identifies key factors affecting CO2 absorption, generates models to predict future CO2 absorption based on historical data, and outputs prediction results.
Accurately predicts future CO2 absorption amounts considering external factors, enabling effective management of carbon credit projects.
Smart Images
Figure 2026009571000001_ABST
Abstract
Description
[Technical Field]
[0001] The present invention relates to an environmental load prediction system and an environmental load prediction method. [Background technology]
[0002] Carbon credits are known as a financial system that addresses the increasing environmental burden on the Earth (global environmental issues). Under the carbon credit system, companies can issue credits for the amount of greenhouse gases they reduce or absorb.
[0003] As carbon credits become more widespread, it is expected that the number of investors who use carbon credits for investments will increase. For example, investors will invest in projects such as reforestation by businesses in return for receiving carbon credits that will be created in the future. In this case, in order for investors to obtain the credits they originally planned, it is important that the businesses' projects reduce greenhouse gas emissions as originally planned (for example, forests continue to absorb carbon dioxide (CO2)).
[0004] Therefore, to ensure that investors receive carbon credits, they must determine the effectiveness of the project, specifically, the amount of greenhouse gases that will be reduced by the project.
[0005] As a method for estimating the amount of CO2 absorbed by forests, Patent Document 1 describes measuring the current amount of CO2 absorbed by forestry projects by measuring CO2 concentrations using satellites. [Prior art documents] [Patent documents]
[0006] [Patent Document 1] US Patent Application Publication No. 2022 / 0341907 Summary of the Invention [Problem to be solved by the invention]
[0007] However, while Patent Document 1 makes it possible to calculate the current CO2 absorption amount, it is not possible to predict the future CO2 absorption amount. Furthermore, the CO2 absorption amount may be affected by various external factors such as temperature fluctuations, changes in solar radiation, and pests.
[0008] The present invention has been made in view of the above circumstances, and its purpose is to provide an environmental load prediction system and an environmental load prediction method that are capable of accurately predicting the amount of environmental load that may be affected by external factors. [Means for solving the problem]
[0009] One aspect of the present invention for solving the above problem is an environmental load prediction system comprising: a storage device that stores values of environmental parameters, which are parameters that represent environmental loads at a specified location at each past timing, and values of multiple types of influence factors, which are parameters that represent the state of the specified location at each of the past timings; an influence factor identification process that identifies, among the multiple types of influence factors, a key influence factor that is the influence factor that most strongly leads the value of the environmental parameter to a specified abnormal value, based on the stored values of the environmental parameters and the values of the influence factors; a model generation process that generates, based on the stored values of the environmental parameters and the values of the influence factors, a first model that represents the relationship between the value of the identified key influence factor and the value of the environmental parameter in the state indicated by the value of the key influence factor; and an absorption amount change calculation process that acquires the value of the key influence factor and inputs the acquired value of the key influence factor into the first model, thereby predicting the value of the environmental parameter under the conditions indicated by the acquired value of the key influence factor, and outputs information on the predicted value of the environmental parameter to an output device. [Effects of the Invention]
[0010] According to the present invention, it is possible to accurately predict the amount of environmental load that may be affected by external factors. Configurations and effects other than those described above will become apparent from the following description of the embodiments. [Brief explanation of the drawings]
[0011] [Figure 1] 1 is a diagram illustrating an example of the configuration of an environmental load prediction system according to an embodiment of the present invention. [Figure 2] FIG. 2 is a diagram illustrating an example of a functional unit included in the absorption amount change prediction device. [Figure 3] FIG. 10 is a diagram showing an example of data stored in the absorption amount change prediction device. [Figure 4] FIG. 10 is a diagram illustrating an example of a time lag setting value DB. [Figure 5] FIG. 10 is a diagram illustrating an example of a threshold data DB. [Figure 6] 10 is a diagram illustrating an outline of a process performed by an absorption amount change prediction device. FIG. [Figure 7] FIG. 10 is a diagram showing an example of an absorption amount data DB. [Figure 8] FIG. 10 is a diagram illustrating an example of an influence factor data DB. [Figure 9] FIG. 10 is a process flow diagram illustrating details of an absorption amount abnormality determination process. [Figure 10] FIG. 10 is a process flow diagram illustrating details of a normal-abnormal correlation coefficient calculation process. [Figure 11] FIG. 10 is a diagram illustrating an example of an abnormality correlation coefficient data DB. [Figure 12] FIG. 10 is a diagram illustrating an example of a normal correlation coefficient data DB. [Figure 13] 10A and 10B are diagrams illustrating an example of calculation of an abnormal state correlation coefficient in the normal state / abnormal state correlation coefficient calculation process. [Figure 14] 10A and 10B are diagrams illustrating an example of calculation of an abnormal state correlation coefficient in the normal state / abnormal state correlation coefficient calculation process. [Figure 15] FIG. 10 is a process flow diagram illustrating details of the influencing factor identification and model generation process. [Figure 16] FIG. 10 is a diagram illustrating an example of an impact analysis data DB. [Figure 17] FIG. 10 is a diagram illustrating an example of an approximate relationship model DB. [Figure 18] FIG. 10 is a processing flow diagram illustrating details of an influencing element change detection process. [Figure 19] FIG. 10 is a process flow diagram illustrating details of an absorption amount change calculation process. [Figure 20] FIG. 10 is a diagram illustrating an example of a forecast report summary screen. DETAILED DESCRIPTION OF THE INVENTION
[0012] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described with reference to the drawings.
[0013] Fig. 1 is a diagram showing an example of the configuration of an environmental load prediction system 100 according to this embodiment. When a business operator 6 is implementing an operation (project) to reduce environmental load within a predetermined area and there is an investor 5 investing in the project, the environmental load prediction system 100 is an information processing system that predicts the future effects of the project for the investor 5. The environmental load prediction system 100 is applied, for example, to a case where the business operator 6 has issued or is about to issue carbon credits to the investor 5.
[0014] In this embodiment, the project is a project (hereinafter referred to as a forestry project) that reduces the concentration of greenhouse gases (here, CO2 (carbon dioxide)) in the atmosphere by planting trees, etc. at each point or area within a specified range (hereinafter referred to as the target area).
[0015] That is, the environmental load prediction system 100 includes information processing devices, namely, an absorption change prediction device 1, an investor communication terminal 2, and an implementer communication terminal 3.
[0016] The investor communication terminal 2 is an information processing device managed by an investor 5 of the project.
[0017] The implementer communication terminal 3 is an information processing device managed by the business operator 6 that implements the project.
[0018] The absorption change prediction device 1 is an information processing device managed by a business operator 4 that predicts the effects of a project. Specifically, the absorption change prediction device 1 predicts the future CO2 absorption amount by forests in the target area based on measured values of CO2 concentration, etc. The absorption change prediction device 1 transmits information on the prediction results to the investor communication terminal 2 and the implementer communication terminal 3.
[0019] That is, the effect of the project is represented by a predetermined environmental parameter (here, CO2 absorption amount). The CO2 absorption amount changes depending on various external factors. In this embodiment, such factors (hereinafter referred to as influencing factors) are assumed to be the density of pests living in the target area, temperature, and amount of solar radiation. The absorption amount change prediction device 1 predicts the future CO2 absorption amount in the target area while taking such influencing factors into consideration. Note that the influencing factors listed here are just examples, and other factors such as humidity, weather, soil characteristics, etc. may also be used.
[0020] The absorption amount change prediction device 1 includes a CPU 11 (Central Processing Unit) (arithmetic device), a memory 12 such as a RAM (Random Access Memory) or a ROM (Read Only Memory), a storage device 13 such as a HDD (Hard Disk Drive) or an SSD (Solid State Drive), an output device 14 such as a display or a printer, an input device 15 such as a keyboard, a mouse, or a touch panel, and a communication interface 16 configured by a NIC (Network Interface Card), a wireless communication module, a USB (Universal Serial Interface) module, a serial communication module, or the like.
[0021] The investor communication terminal 2 includes a CPU 21, a memory 22 such as RAM or ROM, a storage device 23 such as HDD or SSD, an output device 24 such as a display or printer, an input device 25 such as a keyboard, mouse, or touch panel, and a communication interface 26 consisting of a NIC, a wireless communication module, a USB module, or a serial communication module, etc.
[0022] The implementer communication terminal 3 includes a CPU 31, a memory 32 such as RAM or ROM, a storage device 33 such as HDD or SSD, an output device 34 such as a display or printer, an input device 35 such as a keyboard, mouse, or touch panel, and a communication interface 36 consisting of a NIC, a wireless communication module, a USB module, or a serial communication module, etc.
[0023] The absorption amount change prediction device 1, the investor communication terminal 2, and the implementer communication terminal 3 are communicatively connected via a wired or wireless communication path 7, such as the Internet, a LAN (Local Area Network), a WAN (Wide Area Network), or a dedicated line.
[0024] 2 is a diagram showing an example of functional units included in the absorption change prediction device 1. The absorption change prediction device 1 includes functional units of an absorption prediction knowledge accumulation unit 8 and a forest project absorption change prediction unit 9.
[0025] The absorption amount prediction knowledge accumulation unit 8 accumulates basic data for predicting future CO2 absorption amounts in the target area based on measurement data of past CO2 absorption amounts in the target area and measurement data of influencing factors, and constructs a mathematical model (hereinafter referred to as the absorption amount prediction model or first model) for predicting future CO2 absorption amounts.
[0026] The forest project absorption change prediction unit 9 predicts the CO2 absorption amount in the target area based on the absorption amount prediction model.
[0027] The absorption prediction knowledge accumulation unit 8 and the forest project absorption change prediction unit 9 are executed repeatedly at predetermined timings (for example, at predetermined times, at predetermined time intervals (every week)).
[0028] Here, the absorption amount prediction knowledge accumulation unit 8 includes a CO2 concentration data acquisition unit 111, an absorption amount calculation unit 112, an absorption amount abnormality determination unit 113, an influencing factor data acquisition unit 114, a normal / abnormal correlation coefficient calculation unit 115, and an influencing factor identification / model generation unit 116.
[0029] The CO2 concentration data acquisition unit 111 acquires the CO2 concentration in the area targeted by the project (target area). In this embodiment, the absorption amount change prediction device 1 acquires the CO2 concentration from a predetermined satellite system that measures the CO2 concentration in the target area or from a database that accumulates measurement values from that satellite system. However, the method of acquiring the CO2 concentration is not limited to this, and any method can be used.
[0030] The absorption amount calculation unit 112 calculates the CO2 absorption amount by the project based on the CO2 concentration acquired by the CO2 concentration data acquisition unit 111.
[0031] The absorption amount abnormality determination unit 113 determines whether there is a period (abnormal period) in which the CO2 absorption amount is abnormal (low effect, i.e., CO2 is not sufficiently absorbed).The absorption amount abnormality determination unit 113 also identifies periods other than the abnormal period (normal periods, periods in which CO2 is sufficiently absorbed).
[0032] In this embodiment, it is assumed that the effect of the project (amount of CO2 absorbed) normally has a periodicity on an annual basis.
[0033] The influencing factor data acquisition unit 114 acquires past data of a plurality of types of influencing factors.
[0034] The normal / abnormal correlation coefficient calculation unit 115 calculates, for each influencing element, an abnormal correlation (abnormal correlation coefficient), which is the correlation (correlation coefficient) between the value of the influencing element when the value of the CO2 absorption amount is an abnormal value (an abnormal period during which the value of the CO2 absorption amount is an abnormal value) and the value of the CO2 absorption amount in that case. Furthermore, the influencing element identification / model generation unit 116 calculates, for each influencing element, an abnormal correlation (abnormal correlation coefficient), when it is assumed that the value of the influencing element is affected after a predetermined time has passed (after a predetermined time lag) since the change in the value of the CO2 absorption amount.
[0035] In addition, the normal / abnormal correlation coefficient calculation unit 115 calculates, for each influencing factor, a normal correlation (normal correlation coefficient), which is the correlation (correlation coefficient) between the value of the influencing factor when the value of the CO2 absorption amount is not an abnormal value (a normal period when the value of the CO2 absorption amount is not an abnormal value) and the value of the environmental parameter when that value is an abnormal value.
[0036] The influencing factor identification / model generation unit 116 identifies, from among multiple types of influencing factors, important influencing factors that most strongly shift the value of the CO2 absorption amount to a predetermined abnormal value. That is, the influencing factor identification / model generation unit 116 calculates, for each influencing factor, the correlation between the value of the influencing factor when the value of the CO2 absorption amount is an abnormal value and the value of the CO2 absorption amount in that case, and identifies important influencing factors based on the calculated correlation for each influencing factor.
[0037] Specifically, the influencing factor identification / model generation unit 116 calculates the difference between the abnormal time correlation and the normal time correlation for each influencing factor, thereby identifying the important influencing factor.
[0038] Furthermore, the influence factor identification / model generation unit 116 generates a model (absorption amount prediction model, first model) that represents the relationship between the value of the identified significant influence factor and the value of the CO2 absorption amount in the state indicated by the value of the significant influence factor.
[0039] The influence factor identification / model generation unit 116 further generates a model (time lag model, second model) that represents the relationship between the value of the identified significant influence factor and the time (time lag) required for the value of the significant influence factor to affect the value of the CO2 absorption amount.
[0040] Next, the forestry project absorption change prediction unit 9 includes an influencing factor change detection unit 117, an absorption change calculation unit 118, and an absorption prediction result report creation unit 119.
[0041] After generating the absorption amount prediction model, the influencing factor change detection unit 117 identifies a period (abnormal period) in which the value of the important influencing factor is abnormal from the time series data of the important influencing factor acquired by the influencing factor data acquisition unit 114.
[0042] The absorption amount change calculation unit 118 predicts future changes in CO2 absorption amount based on the identified abnormal period and absorption amount prediction model. That is, the absorption amount change calculation unit 118 inputs the values of the significant influencing factors for the abnormal period into the absorption amount prediction model, and thereby predicts the value of CO2 absorption amount in the target area under the conditions indicated by the values of the significant influencing factors.
[0043] The absorption amount prediction result report creation unit 119 generates a screen (prediction report summary screen) showing the important influencing factors in the target area and future changes in CO2 absorption amount.
[0044] 3 is a diagram showing an example of data stored in the absorption amount change prediction device 1. The absorption amount change prediction device 1 stores an absorption amount data DB131 which is a database related to the history of CO2 absorption amounts, an influence element data DB132 which is a database related to numerical data related to each influence element, a time lag setting value DB133 which is a database related to time lags, an abnormal-state correlation coefficient data DB134 which is a database related to abnormal-state correlation coefficients, a normal-state correlation coefficient data DB135 which is a database related to normal-state correlation coefficients, a threshold data DB136 which is a database related to post-thresholds for determining important influence elements, an influence analysis data DB137, and an approximate relationship model DB138 which is a database related to absorption amount prediction models.
[0045] (Time lag setting value DB) 4 is a diagram showing an example of the time lag setting value DB 133. In the time lag setting value DB 133, a plurality of types of time lag values are set.
[0046] (Threshold data DB) 5 is a diagram illustrating an example of the threshold data DB 136. The threshold data DB 136 includes data on a correlation coefficient threshold 136a, which is a threshold related to the correlation, and data on a difference threshold 136b, which is a threshold related to the difference between the abnormal correlation and the normal correlation.
[0047] The functional units of the absorption amount change prediction device 1 described above are realized by the CPU 11 of the absorption amount change prediction device 1 reading out programs from the memory 12 or the storage device 13. Each program can be recorded on, for example, a portable or fixed recording medium and distributed. All or part of each program in the absorption amount change prediction device 1 may be realized using virtual information processing resources provided using virtualization technology, process space separation technology, or the like, such as a virtual server provided by a cloud system. All or part of these programs may be realized by a service provided by a cloud system via an API (Application Programming Interface), for example. Next, the processing performed by the environmental load prediction system 100 will be described.
[0048] 6 is a diagram illustrating an outline of the process performed by the absorption amount change prediction device 1. This process is started, for example, when the absorption amount change prediction device 1 acquires the CO2 concentration from the satellite system.
[0049] The CO2 concentration data acquisition unit 111 acquires time-series data of CO2 concentration in the target area (S111).
[0050] For example, the CO2 concentration data acquisition unit 111 acquires the CO2 concentration in a developed area (area where afforestation is being carried out) and the CO2 concentration in an undeveloped area (area where afforestation is not being carried out) in the target area.
[0051] Then, the absorption amount calculation unit 112 calculates time series data of the CO2 absorption amount by the project in the target area based on the CO2 concentration acquired in S111 (S112). For example, the absorption amount calculation unit 112 subtracts the CO2 concentration of the maintenance area from the CO2 concentration of the maintenance area in the abandoned area. The absorption amount calculation unit 112 registers the calculated time series data of the CO2 absorption amount in the absorption amount data DB131.
[0052] (Absorption data DB) 7 is a diagram showing an example of the absorption amount data DB 131. The absorption amount data 131 is a database that manages environmental parameters (CO2 absorption amount) in a target area. The absorption amount data DB 131 includes the following data: ID 131a of each project, forest tier 131b related to each project, measurement date 131c of CO2 absorption amount per unit area for each project, CO2 absorption amount per unit area 131d on measurement date 132c for each project, and abnormal value flag 131e, which is associated with each project and is set when the CO2 absorption amount is an abnormal value.
[0053] The term "stratum" refers to the type of plant that makes up a forest, and for example, plants with the same botanical characteristics are classified into the same stratum. In other words, different stratums have different CO2 absorption patterns.
[0054] Next, as shown in FIG. 6, the absorption amount abnormality determination unit 113 executes an absorption amount abnormality determination process S113 for identifying an abnormal period during which the CO2 absorption amount is an abnormal value, based on the absorption amount data DB 131.
[0055] On the other hand, the influence factor data acquiring unit 114 acquires time series data of each influence factor from a predetermined device or database etc. (S114a). The influence factor data acquiring unit 114 registers the time series data of each influence factor in the influence factor data DB132.
[0056] When the influencing factor is pest density, for example, the business operator 6 installs insect capture means such as an insect trap or insect paper in the target area (abandoned area, maintenance area) in advance for a predetermined period of time, and a predetermined photography system or the business operator 6 photographs the swarm of insects captured by the insect capture means with a photography device to obtain image data of the swarm of insects in the insect capture means.The influencing factor data acquisition unit 114 then inputs this image data into a predetermined discrimination model (a trained model that determines the number of insects present in the image) to obtain the pest density.
[0057] When the influencing factor is the amount of solar radiation, for example, the influencing factor data acquiring unit 114 acquires data on the amount of solar radiation for each period from a predetermined weather database.
[0058] (Influence factor data DB) 8 is a diagram showing an example of the influencing factor data DB 132. The influencing factor data DB 132 is a database that manages multiple types of influencing factors in a target area. The influencing factor data DB 132 includes each project ID 132a, each project tier 132b, measurement date 132c of pest density in each project, pest density 132d, which is the value of an influencing factor for measurement date 132 in each project, and temperature 132e (winter temperature), which is the value of another influencing factor for measurement date 132 in each project.
[0059] Next, as shown in FIG. 6, the normal / abnormal correlation coefficient calculation unit 115 executes a normal / abnormal correlation coefficient calculation process S115 to calculate an abnormal correlation coefficient and a normal correlation coefficient for each influencing factor based on the absorption amount data DB 131 and the influencing factor data DB 132.
[0060] The influencing factor identification / model generation unit 116 executes an influencing factor identification process S116 to identify important influencing factors based on the calculated abnormal-state correlation coefficient and normal-state correlation coefficient, and to generate an absorption amount prediction model and a time lag model. The influencing factor identification / model generation unit 116 registers each generated model in the approximate relational model DB 138.
[0061] After generating the approximate relationship model DB 138, the influence factor data acquisition unit 114 acquires time-series data of important influence factors from a predetermined device, database, or the like by the same process as in S114a (S114b).
[0062] The influencing factor change detection unit 117 executes an influencing factor change detection process S117 for identifying an abnormal period in which the value of the important influencing factor is an abnormal value from the time-series data of the important influencing factor acquired in S114b.
[0063] The absorption amount change calculation unit 118 executes an absorption amount change calculation process S118 for predicting future changes in CO2 absorption amount by inputting time series data of the important influencing factors for the abnormal period identified in the influencing factor change detection process S117 into the absorption amount prediction model and the time lag model generated by the influencing factor identification / model generation unit 116.
[0064] The absorption amount prediction result report creation unit 119 displays a screen showing the results of the absorption amount change calculation process S118 and the like (S119).
[0065] <Absorption amount abnormality determination process S113> 9 is a process flow diagram illustrating details of the absorption amount abnormality determination process S113. The absorption amount abnormality determination unit 113 acquires past time-series data of the CO2 absorption amount for a certain floor that has been accumulated up to the present time from the absorption amount data DB 131 (S1131).
[0066] The absorption amount abnormality determination unit 113 identifies an abnormal period in which the CO2 absorption amount is an abnormal value and the time series data of the CO2 absorption amount during that abnormal period (S1132).
[0067] For example, the absorption amount abnormality determination unit 113 performs a statistical analysis of the CO2 absorption amount 13d of each record in the absorption amount data DB 131, and identifies the CO2 absorption amount 131d that is not within the 2σ interval and its measurement date 131c. The absorption amount abnormality determination unit 113 sets "1" to the abnormal value flag 131e of the record related to the CO2 absorption amount 131d that is not within the 2σ interval, and sets "0" to the abnormal value flag 131e of the record related to the CO2 absorption amount 131d that is within the 2σ interval.
[0068] The absorption amount abnormality determination unit 113 determines whether or not an abnormal period was identified in S1132 (S1133). If an abnormal period was identified in S1132 (S1133: Yes), the absorption amount abnormality determination unit 113 executes the process of S1134, and if an abnormal period was not identified in S1132 (S1133: No), the absorption amount abnormality determination unit 113 executes the process of S1131 on the updated absorption amount data DB 131.
[0069] In S1134, the absorption amount abnormality determination unit 113 transmits data on the CO2 absorption amount relating to the abnormal period (contents of the record relating to the abnormal period in the absorption amount data DB 131) to the normal / abnormal state correlation coefficient calculation unit 115 (S1134).
[0070] <Calculation of correlation coefficient between normal and abnormal conditions> FIG. 10 is a process flow diagram illustrating details of the normal / abnormal correlation coefficient calculation process S115.
[0071] The normal / abnormal correlation coefficient calculation unit 115 reads the data of the CO2 absorption amount in the abnormal period for the selected project and floor, which is output from the absorption amount abnormality determination unit 113 (S1151).
[0072] In the example of FIG. 7, the normal / abnormal state correlation coefficient calculation unit 115 acquires, for example, from the absorption amount data DB 131, the contents of the record 1311 in which the abnormality flag 1311e is "1" from December 15, 2024 to January 15, 2025, for the project with the project ID "53", and for which the value of the hierarchical ID 131b is "12".
[0073] Furthermore, as shown in FIG. 10, the normal / abnormal state correlation coefficient calculation unit 115 reads the influencing element data DB 132 to obtain time series data of each influencing element (S1152).
[0074] The normal / abnormal correlation coefficient calculation unit 115 acquires, for the same hierarchical ID, time series data of each influencing element during the abnormal period (time series data (a)) and time series data of each influencing element during a period other than the abnormal period (normal period) (time series data (b)) (S1153).
[0075] For example, the normal / abnormal correlation coefficient calculation unit 115 acquires from the influence factor data DB 132 the contents of each record whose measurement date 132c is within the abnormal period, and also acquires the contents of each record whose measurement date 132c is within the abnormal period.
[0076] In the example of FIG. 8, for hierarchical ID "12" in the project with project ID "53", time series data (a) 1322 of pest density for the abnormal period (December 15, 2024 to January 5, 2025) in which the abnormal flag is "1" is acquired. Also, for hierarchical ID "12" in the project with project ID "1", time series data (b) 1321 of pest density for the abnormal period (December 15, 2024 to January 5, 2025) in which the abnormal flag is "0" is acquired.
[0077] Then, the normal / abnormal correlation coefficient calculation unit 115 calculates the correlation coefficient (abnormal correlation coefficient) between the value of the influencing element and the CO2 absorption amount by comparing the time series data (a) of the influencing element during the abnormal period with the time series data of the CO2 absorption amount during the abnormal period (time series data record 1311 of the absorption amount data DB 131) for the same hierarchical ID (S1154).
[0078] Furthermore, for each time lag, the normal / abnormal state correlation coefficient calculation unit 115 converts the time series data (a) of the influencing element during the abnormal period for the same hierarchical ID into time series data delayed by the time lag indicated by the time lag setting value DB 133. Then, the normal / abnormal state correlation coefficient calculation unit 115 compares the converted time series data of the influencing element with the time series data of the CO2 absorption amount during the abnormal period, which are data of the same timing, for each time lag, to calculate the correlation coefficient between the value of the influencing element and the CO2 absorption amount (the abnormal state correlation coefficient when the influencing element is shifted by the time lag) (S1154).
[0079] The normal / abnormal correlation coefficient calculation unit 115 registers the data of the correlation coefficient calculated in S1154 in the abnormal correlation coefficient data DB 134 (S1155).
[0080] Furthermore, the normal / abnormal correlation coefficient calculation unit 115 compares the time series data (b) of the influencing factor related to the normal period with the time series data of the CO2 absorption amount during the abnormal period (record 1311 of the time series data of the record whose abnormality flag in the absorption amount data DB 131 is "1"), which are data at the same timing, for the same hierarchical ID, to calculate the correlation coefficient (normal correlation coefficient) between the value of the influencing factor during the normal period and the CO2 absorption amount during the abnormal period (S1156).
[0081] The normal / abnormal correlation coefficient calculation unit 115 registers the data of the normal correlation coefficient calculated in S1156 in the normal correlation coefficient data DB 135 (S1157).
[0082] Unlike the time series data (a) of the influential factors, the time series data (b) of the influential factors is data from a normal period, and therefore does not contain any abnormal values. Therefore, in S1156, there is no need to calculate the correlation coefficient taking the time lag into account.
[0083] The normal / abnormal state correlation coefficient calculation unit 115 transmits the contents of the abnormal state correlation coefficient data DB 134 and the normal state correlation coefficient data DB 135 to the influencing factor identification / model generation unit 116 (S1158).
[0084] (Abnormal correlation coefficient data DB) 11 is a diagram showing an example of the abnormality correlation coefficient data DB 134. The abnormality correlation coefficient data DB 134 has data on items (correlated items 134a) of influencing factors (for example, pest density) correlated with the CO absorption amount, time lags 134b associated in the correlation, and correlation coefficients 134c (abnormality correlation coefficients) in the correlation.
[0085] For example, record 1341 shown in the figure indicates that there is a correlation between the CO2 absorption amount and pest density during the abnormal period with a time lag of -6 and a correlation coefficient of 0.68.
[0086] (Normal correlation coefficient data DB) 12 is a diagram showing an example of the normal state correlation coefficient data DB 135. The normal state correlation coefficient data DB 135 has data on items (correlated items 135a) of influencing factors (for example, pest density) correlated with the CO absorption amount, and correlation coefficients 135b (normal state correlation coefficients) in the correlation.
[0087] 13 is a diagram illustrating an example of calculation of the abnormal state correlation coefficient in the normal state / abnormal state correlation coefficient calculation process S115. The normal state / abnormal state correlation coefficient calculation unit 115 compares time series data 1302 of the CO2 absorption amount with time series data 1303 of the values of the influencing factors during an abnormal period 1301, and calculates the abnormal correlation coefficient.
[0088] Then, the normal / abnormal correlation coefficient calculation unit 115 generates time series data 1304 by delaying the values in the time series data 1303 of the influencing factors by a time Δt, and compares the generated time series data 1304 with the time series data 1302 of the CO2 absorption amount during the abnormal period 1301 to calculate a normal correlation coefficient.
[0089] 14 is a diagram illustrating an example of calculation of the abnormal state correlation coefficient in the normal state / abnormal state correlation coefficient calculation process S115. The normal state / abnormal state correlation coefficient calculation unit 115 identifies time series data 1701 of the CO2 absorption amount during a normal period, and also acquires time series data 1702 of the values of the influencing elements during that normal period. The normal state / abnormal state correlation coefficient calculation unit 115 compares the time series data 1302 of the CO2 absorption amount during the abnormal period 1301 with the time series data 1702 of the values of the influencing elements during the normal period, and calculates the abnormal state correlation coefficient.
[0090] <Influence factor identification and model generation processing> FIG. 15 is a process flow diagram illustrating the details of the influencing factor identification and model generation process S116.
[0091] The influencing factor identification / model generation unit 116 acquires the contents of the abnormal state correlation coefficient data DB 134 and the normal state correlation coefficient data DB 135 from the normal state / abnormal state correlation coefficient calculation unit 115 (S1161).
[0092] The influencing factor identification / model generation unit 116 also acquires the value of the correlation coefficient threshold 136a from the threshold data DB 136. Then, the influencing factor identification / model generation unit 116 acquires all records in which the value of the correlation coefficient 134c exceeds the value of the correlation coefficient threshold 136a acquired above from the abnormal correlation coefficient data DB 134 acquired in S1161, thereby extracting combinations of influencing factors and time lags that are highly correlated with the CO2 absorption amount during the abnormal period (S1162).
[0093] The influencing factor identification / model generation unit 116 calculates the difference between the abnormal correlation coefficient associated with each combination of influencing factors and time lags extracted in S1162 and the normal correlation coefficient associated with the influencing factors in that combination, thereby calculating the net correlation between the CO2 absorption amount and the value of each influencing factor during the abnormal period (S1163).The influencing factor identification / model generation unit 116 then identifies the combination of influencing factors and time lags with the largest difference value (S1164).
[0094] The influencing factor identification / model generation unit 116 registers the combination of the influencing factor and its time lag identified in S1164, and the abnormal-state correlation coefficient and normal-state correlation coefficient associated with the combination, in the impact analysis data 137 (S1165). For example, the influencing factor identification / model generation unit 116 registers the content of the record 1341 in the abnormal-state correlation coefficient data DB 134 in a new record in the impact analysis data DB 137.
[0095] The influence factor identification / model generation unit 116 generates an absorption amount prediction model based on the CO2 absorption amount in the absorption amount data DB 131 and the values of the influence factors in the influence factor data DB 132, and registers the contents of the generated absorption amount prediction model in the approximate relationship model DB 138 (S1166).
[0096] Specifically, the influence factor identification / model generation unit 116 generates, for each influence factor, an approximation function (e.g., a polynomial with the influence factor as a variable) that uses the value of the influence factor as an input variable and the CO2 absorption amount as an output value.
[0097] Furthermore, the influencing factor identification / model generation unit 116 generates, for each influencing factor, an approximation function (for example, a polynomial with the influencing factor as a variable) with the value of the influencing factor as an input variable and the time lag as an output value.
[0098] Furthermore, the influencing factor identification / model generation unit 116 calculates the average value of the abnormal correlation coefficient and the average value of the normal correlation coefficient for each influencing factor based on the contents of the influence analysis data DB 137, and registers each calculated average value in the approximate relationship model DB 138 (S1167).
[0099] (Impact analysis data DB) 16 is a diagram showing an example of the impact analysis data DB 137. The impact analysis data 137 has each data item: an ID 137a of each hierarchical level, an impact factor 137b in each hierarchical level, a difference value 137c related to the impact factor, an average value 137d of the impact factor, a time lag 137e associated with the impact factor, an abnormal state correlation coefficient 137f related to the impact factor, and a normal state correlation coefficient 137g related to the impact factor.
[0100] (Approximate relational model DB) 17 is a diagram showing an example of the approximate relationship model DB 138. The approximate relationship model DB 138 has each piece of data: each influencing element 138a, an ID 138b of each tier, an absorption amount prediction model equation 138c associated with the influencing element and tier, a correlation coefficient 138d calculated in determining the absorption amount prediction model equation, a time lag relational equation 138e associated with the influencing element and tier, a correlation coefficient 138f calculated in determining the time lag relational equation, an average value 138g of abnormal correlation coefficients associated with the influencing element and tier, and an average value 138h of normal correlation coefficients associated with the influencing element and tier.
[0101] The correlation coefficient 138d in the equation of the absorption amount prediction model is a correlation coefficient in the basis data (influence factors and measured values of the CO2 absorption amount) of the equation 138c of the absorption amount prediction model.
[0102] The correlation coefficient 138f in the time lag relational expression is the correlation coefficient in the basis data (value of the influential element and value of the time lag) of the time lag relational expression 138ec.
[0103] The average value 138g of the abnormal state correlation coefficient is the average value of the abnormal state correlation coefficient 137f of the influence analysis data DB 137, which is the basis data for the equation 138c of the absorption amount prediction model.
[0104] The average value 138h of the correlation coefficients under normal conditions is the average value of the correlation coefficients under normal conditions 137g of the influence analysis data DB 137, which is the basis data for the equation 138c of the absorption amount prediction model.
[0105] <Influence factor change detection process S117> FIG. 18 is a processing flow diagram illustrating the details of the influencing element change detection processing S117.
[0106] The influencing factor change detection unit 117 acquires the time-series data of the influencing factors identified in the influencing factor identification / model generation process S116 from the influencing factor data DB 132 output by the influencing factor data acquisition unit 114 (S1171).
[0107] The influencing factor change detection unit 117 identifies, from the time-series data of the influencing factor acquired in S1171, a period in which the value is an abnormal value and the time-series data of the value of the influencing factor in that period (S1172).
[0108] For example, the influencing element change detection unit 117 performs statistical analysis of the pest density 132d of each record in the pest density influencing element data DB 132 for the abnormal period, and identifies the pest density 132d and its measurement date 132c that are not within the 2σ interval.
[0109] If an influence element with an abnormal value is found (S1173: Yes), the influence element change detection unit 117 executes the processing of S1174, and if an influence element with an abnormal value is not found (S1173: No), the influence element change detection unit 117 repeats the processing of S1171 for the time series data of the updated influence element.
[0110] In S1174, the influencing factor change detection unit 117 transmits the abnormal value of the influencing factor to the absorption amount change calculation unit 118. If the abnormal value of the influencing factor exists over multiple consecutive days, the influencing factor change detection unit 117 calculates the average value of the abnormal values of the influencing factor for each day and transmits it to the absorption amount change calculation unit 118.
[0111] <Absorption change calculation process> FIG. 19 is a process flow diagram illustrating the details of the absorption amount change calculation process S118.
[0112] The absorption amount change calculation unit 118 acquires the abnormal value (or the average value) of the influencing factor in the target period from the influencing factor change detection unit 117 (S1181).
[0113] The absorption amount change calculation unit 118 inputs the acquired abnormal values (or their average values) of the influencing factors into an absorption amount prediction model and a time lag model, respectively, to predict the change in the CO2 absorption amount and the time lag (S1182).
[0114] For example, the absorption amount change calculation unit 118 calculates the change in CO2 absorption amount and the time lag by substituting the pest density (pest density 132d of the influence element data DB 132) into the absorption amount relational expression 138c and the time lag relational expression 138f, respectively.
[0115] The absorption amount change calculation unit 118 transmits the calculated change in CO2 absorption amount and information on the time lag to the absorption amount prediction result report creation unit 119.
[0116] (Forecast report summary screen) 20 is a diagram showing an example of a forecast report summary screen 1191. The forecast report summary screen 1191 is displayed on the investor communication terminal 2 or the implementer communication terminal 3, for example.
[0117] The forecast report summary screen 1191 has an influence factor information display field 1192 that displays information about influence factors, an influence factor distribution display field 1193 that displays the distribution of predicted values of influence factors in the target area, a CO2 absorption amount distribution display field 1194 that displays the distribution of predicted values of CO2 absorption amount in the target area, and a time lag distribution display field 1195 that displays the distribution of predicted values of time lags in the target area.
[0118] The influencing factor information display field 1192 displays the influencing factor (e.g., pest density), the hierarchical level, and the average value of the correlation coefficient (normal correlation coefficient and abnormal correlation coefficient) between the value of the influencing factor and the CO2 absorption amount for that hierarchical level.
[0119] In the influencing factor distribution display field 1193, the CO2 absorption amount distribution display field 1194, and the time lag distribution display field 1195, the values of the influencing factors, CO2 absorption amount, and time lag values in each target area are each represented by a fill pattern or color according to the magnitude of the value.
[0120] The influencing factor information display field 1192 is generated based on the influencing factors 138 a , the hierarchical ID 138 b , the average value 138 g of the abnormal correlation coefficients, and the average value 138 h of the normal correlation coefficients in the approximate relationship model DB 138 .
[0121] The influencing factor distribution display field 1193 is generated based on the influencing factor data DB 132 relating to abnormal values, which is generated by the influencing factor change detection unit 117.
[0122] The CO2 absorption amount distribution display field 1194 is generated based on the change in CO2 absorption amount calculated by the absorption amount change calculation unit 118.
[0123] The time lag distribution display field 1195 is generated based on the time lag calculated by the absorption amount change calculation unit 118 .
[0124] As described above, the environmental load prediction system 100 of this embodiment identifies, from among multiple types of influence factors (pest density, solar radiation, etc.), the important influence factor that most strongly leads the environmental parameter value (CO2 absorption amount) to an abnormal value using the absorption amount data DB131, the influence factor data DB132, etc., generates an absorption amount prediction model that represents the relationship between the value of the identified important influence factor and the CO2 absorption amount under the conditions indicated by the value of the important influence factor, inputs the acquired value of the important influence factor into the absorption amount prediction model, predicts the CO2 absorption amount under the conditions indicated by the acquired value of the important influence factor, and outputs information on the predicted CO2 absorption amount.
[0125] That is, the environmental load prediction system 100 of this embodiment predicts environmental parameters using an absorption amount prediction model for predicting the value of an environmental parameter related to the most important influencing factor that leads the environmental parameter to the most abnormal value among multiple influencing factors that can affect the value of the environmental parameter (such as CO2 absorption amount).
[0126] In this way, the environmental load prediction system 100 of this embodiment can accurately predict the amount of environmental load that may be affected by external factors.
[0127] In addition, the environmental load prediction system 100 of this embodiment calculates, for each influence factor, the correlation between the value of the influence factor when the value of the environmental parameter is an abnormal value and the value of the environmental parameter in that case, and identifies important influence factors based on the calculated correlation for each influence factor.
[0128] In this way, by identifying important influencing factors by utilizing the correlation between the value of an environmental parameter and the value of an influencing factor when the environmental parameter takes an abnormal value, it is possible to accurately identify the influencing factor that leads the value of the environmental parameter to an abnormal value.
[0129] Specifically, the environmental load prediction system 100 of this embodiment calculates, for each influence element, an abnormal correlation, which is the correlation between the value of the influence element when the value of the environmental parameter is an abnormal value and the value of the environmental parameter in that case, and also calculates, for each influence element, a normal correlation, which is the correlation between the value of the influence element when the value of the environmental parameter is a normal value and the value of the environmental parameter when that value is an abnormal value, and identifies important influence elements by calculating the difference between the abnormal correlation and the normal correlation for each influence element.
[0130] In this way, by identifying important influencing factors using the difference between the abnormal correlation when the environmental parameter has an abnormal value and the normal correlation when the environmental parameter has a normal value, it is possible to accurately estimate the net extent to which the influencing factors cause the environmental parameter value to become abnormal.
[0131] More specifically, the environmental load prediction system 100 of this embodiment extracts abnormal correlations of influencing elements whose correlations exceed a predetermined threshold, calculates the difference between the abnormal correlations of each extracted influencing element and the normal correlations of each influencing element, and identifies the influencing element related to the abnormal correlation with the largest difference among the calculated differences as the important influencing element.
[0132] This makes it possible to accurately estimate the net degree to which influencing factors cause the values of environmental parameters to become abnormal values, and to extract appropriate important influencing factors.
[0133] Furthermore, the environmental load prediction system 100 of this embodiment calculates, for each influencing factor, an abnormal correlation when it is assumed that the value of the environmental parameter is affected after a predetermined time lag from a change in the value of the influencing factor.
[0134] Depending on the influencing factor, it may take a considerable amount of time for the environmental parameter to become an abnormal value. The environmental load prediction system 100 of this embodiment can take such a time lag into consideration and calculate the abnormal correlation for each influencing factor to identify important influencing factors, so that it can accurately identify influencing factors that lead the value of the environmental parameter to become an abnormal value, including time factors.
[0135] In addition, the environmental load prediction system 100 of this embodiment further generates a time lag model that represents the relationship between the value of a significant influence factor and the time (time lag) required for the value of that significant influence factor to affect the value of the environmental parameter, and by inputting the value of the significant influence factor into the time lag model, predicts the time lag required for the value of the significant influence factor to affect the value of the environmental parameter, and outputs information on the predicted time lag.
[0136] This makes it possible to predict, for example, when the value of a significant influence factor becomes abnormal, and at what timing this will subsequently affect the value of the environmental parameter.
[0137] Furthermore, the environmental load prediction system 100 of this embodiment employs values of pest density, temperature, or amount of solar radiation as influential factors, and employs CO2 absorption amount as an environmental parameter.
[0138] This allows, for example, investors who invest in forestry projects instead of carbon credits to predict CO2 absorption by forests, taking into account various factors that may affect the amount of CO2 absorption by forests, and can adjust their carbon credit policies early based on the predicted value.On the other hand, forestry project implementers can consider appropriate countermeasures if the predicted CO2 absorption value is on a downward trend and use the information to improve future forestry project policies.
[0139] The present invention is not limited to the above-described embodiments, and can be implemented using any components within the scope of the present invention. The above-described embodiments and modifications are merely examples, and the present invention is not limited to these contents as long as the features of the invention are not impaired. Furthermore, although various embodiments and modifications have been described above, the present invention is not limited to these contents. Other aspects conceivable within the scope of the technical idea of the present invention are also included within the scope of the present invention.
[0140] For example, part of the hardware provided in each device of this embodiment may be provided in another device.
[0141] Furthermore, each program of each device may be provided in another device, a program may consist of multiple programs, or multiple programs may be integrated into one program.
[0142] For example, the absorption amount change prediction device 1 can also be applied to absorption amount changes in other carbon credit natural absorption projects such as farmland projects and marine projects.
[0143] Furthermore, in this embodiment, the carbon credit system is assumed and the case where the environmental parameter representing the environmental load is the CO2 absorption amount (concentration) has been described. However, the present invention is also applicable to a system that makes predictions regarding environmental loads other than greenhouse gases (for example, various harmful liquids, solids, or gases).
[0144] Additionally, the information displayed in the impact factor information display field 1192 on the forecast report summary screen 1191 may be provided to a provider of a rating service for carbon credit creation projects. This can be used to improve the method of rating project risks. [Explanation of symbols]
[0145] 1 Absorption amount change prediction device, 111 CO2 concentration data acquisition unit, 112 Absorption amount calculation unit, 113 Absorption amount abnormality judgment unit, 114 Influence factor data acquisition unit, 115 Normal / abnormal correlation coefficient calculation unit, 116 Influence factor identification / model generation unit, 117 Influence factor change detection unit, 118 Absorption amount change calculation unit, 119 Absorption amount prediction result report creation unit
Claims
1. a storage device that stores values of environmental parameters, which are parameters that represent environmental loads at a predetermined point at each past timing, and values of a plurality of types of influencing factors, which are parameters that represent the state of the predetermined point at each past timing; and an influencing element identification process for identifying, from among the plurality of types of influencing elements, a significant influencing element that is an influencing element that most strongly leads the value of the environmental parameter to a predetermined abnormal value, based on the stored values of the environmental parameter and the influencing element; a model generation process for generating a first model representing a relationship between the value of the identified significant influence factor and the value of an environmental parameter in a state indicated by the value of the significant influence factor, based on the stored values of the environmental parameters and the values of the influence factors; a computing device that executes an absorption amount change calculation process, which acquires values of important influence factors, inputs the acquired values of important influence factors into the first model, thereby predicting values of environmental parameters under conditions indicated by the acquired values of important influence factors, and outputs information on the predicted values of the environmental parameters to an output device. An environmental load prediction system equipped with the above.
2. The computing device In the influencing factor identification process, a correlation between the value of the influencing factor when the value of the environmental parameter is the abnormal value and the value of the environmental parameter in the abnormal value is calculated for each influencing factor, and the important influencing factor is identified based on the calculated correlation related to each influencing factor. The environmental load prediction system according to claim 1 .
3. In the influencing factor identification process, the arithmetic device Calculating an abnormal correlation for each influencing factor, which is a correlation between the value of the influencing factor when the value of the environmental parameter is the abnormal value and the value of the environmental parameter in the abnormal value; Calculating a normal correlation for each influencing factor, which is a correlation between the value of the influencing factor when the value of the environmental parameter is not the abnormal value and the value of the environmental parameter when the value of the influencing factor is the abnormal value; calculating a difference between the abnormal correlation and the normal correlation for each of the influencing factors to identify the important influencing factors; The environmental load prediction system according to claim 2 .
4. In the influencing factor identification process, the arithmetic device extracting, from the calculated correlations of the influencing factors when an abnormality occurs, correlations of the influencing factors when the correlations exceed a predetermined threshold value; calculating a difference between the abnormal correlation of each of the extracted influence factors and the normal correlation of each of the influence factors; identifying an influencing factor related to the abnormal correlation having the largest difference among the calculated differences as the important influencing factor; The environmental load prediction system according to claim 3 .
5. The computing device In the influencing factor identification process, an abnormality correlation is calculated for each influencing factor when it is assumed that the value of the environmental parameter is affected after a predetermined time has elapsed since a change in the value of the influencing factor; calculating a difference between the abnormal correlation and the normal correlation for each of the influencing factors to identify the important influencing factors; The environmental load prediction system according to claim 3 .
6. The computing device In the model generation process, a second model is further generated based on the stored values of the environmental parameters and the influencing factors, the second model representing a relationship between the identified significant influencing factors and the time required for the value of the identified significant influencing factors to affect the value of the environmental parameters; In the absorption amount change calculation process, the value of the acquired significant influence factor is input into the second model to predict a time required for the value of the acquired significant influence factor to affect the value of the environmental parameter, and information on the predicted time is output to an output device. The environmental load prediction system according to claim 5 .
7. The storage device stores the amount of CO2 absorption at each past timing at a predetermined point, and the values of the influencing factors, such as pest density, temperature, or amount of solar radiation, at each past timing, The computing device In the influencing factor identification process, a significant influencing factor that most strongly leads the value of the CO2 absorption amount to a predetermined abnormal value is identified from among the plurality of types of influencing factors based on the stored value of the CO2 absorption amount and the value of the influencing factor; In the model generation process, a first model is generated that represents a relationship between the value of the identified significant influence factor and the CO absorption amount in a state indicated by the value of the significant influence factor; In the absorption amount change calculation process, a value of a significant influence factor is acquired, and the acquired value of the significant influence factor is input into the first model to predict the CO2 absorption amount under the conditions indicated by the acquired value of the significant influence factor, and information on the predicted value of the CO2 absorption amount is output to an output device. The environmental load prediction system according to claim 1 .
8. An environmental load prediction method using an information processing device including a storage device that stores values of environmental parameters, which are parameters that represent environmental loads at a specified point at each past timing, and values of a plurality of types of influencing factors, which are parameters that represent the state of the specified point at each past timing, and a calculation device, The computing device an influencing element identification process for identifying, from among the plurality of types of influencing elements, a significant influencing element that is an influencing element that most strongly leads the value of the environmental parameter to a predetermined abnormal value, based on the stored values of the environmental parameter and the influencing element; a model generation process for generating a first model representing a relationship between the value of the identified significant influence factor and the value of the environmental parameter in a state indicated by the value of the significant influence factor; and executing an absorption amount change calculation process for acquiring values of important influence factors, inputting the acquired values of important influence factors into the first model, thereby predicting values of environmental parameters under conditions indicated by the acquired values of important influence factors, and outputting information on the predicted values of the environmental parameters to an output device. Environmental load prediction methods.
Citation Information
Patent Citations
Environmental emission monitoring system with GHG emission thresholding and related method
US20220341907A1