Crop carbon sink calculation model, crop carbon credit calculation method and storage medium

Abstract

The invention relates to the technical field of carbon sink calculation, in particular to a crop carbon sink calculation model, a crop carbon credit calculation method and a storage medium. The crop carbon sink calculation model is a multi-task learning model based on a Transform architecture, and the multi-task learning model comprises a synchronous prediction carbon sink quantity module, a soil carbon sequestration potential module and a carbon emission factor module; the training data of the model comprises the carbon content of the crop physiology knowledge machine; the crop carbon sink calculation model is used for estimating the carbon sink amount of the target farmland area, multi-dimensional calculation of the carbon sink amount is achieved, the model generalization ability and the calculation precision are ensured, meanwhile, the manual dependency degree is reduced, and the calculation efficiency is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon sink calculation, in particular to a crop carbon sink calculation model, a crop carbon credit calculation method, and a storage medium. Background Art

[0002] Carbon emissions and carbon sequestration have become a focus of attention. Carbon sinks refer to the total amount of carbon dioxide removed from the atmosphere by ecosystems through biological uptake, chemical adsorption, and physical retention. Accurately calculating carbon sinks is crucial for assessing ecosystem services and developing carbon reduction policy tools.

[0003] Current crop carbon credit calculations rely primarily on manual field surveys or single remote sensing data. However, manual field surveys require significant manpower, take weeks or even months, and lack coverage of crops over wide areas. Carbon sink calculations based on remote sensing data ignore the dynamics of soil organic carbon and the impact of crop growth stages, leading to high error rates. Furthermore, remote sensing-based carbon sink calculations rely heavily on spectral data and fail to account for factors that significantly impact carbon sinks, such as extreme weather and agricultural tillage, severely impacting accuracy. Summary of the Invention

[0004] The purpose of this application is to provide a crop carbon sink calculation model, a crop carbon credit calculation method, and a storage medium. Through a multi-task learning model based on the Transformer architecture, combined with a simultaneous prediction carbon sink module, a soil carbon storage calculation module, and a carbon emission factor module, a multi-dimensional crop carbon sink calculation model is established to solve the high dependence on manual labor and the inability to adapt to large-scale crop carbon monitoring needs. At the same time, each module covers all major aspects that affect changes in crop carbon reserves, ensuring coverage and improving the accuracy of carbon sink calculation.

[0005] The present application provides a crop carbon sink calculation model, including a multi-task learning model based on the Transformer architecture, wherein the multi-task learning model includes: a simultaneous carbon sink prediction module, a soil carbon storage calculation module, and a carbon emission factor module; the training data of the model includes crop physiology knowledge.

[0006] Furthermore, an interpretability module is included for generating a heat map of carbon sequestration feature importance to display the weight distribution of carbon sequestration contribution factors.

[0007] Furthermore, it also includes a dynamic correction module for adjusting carbon credits in combination with sudden environmental factors.

[0008] On the other hand, the present application also provides a method for calculating crop carbon credits based on a crop carbon sink calculation model, which uses any of the above-mentioned crop carbon sink calculation models to calculate the carbon sink amount of the target farmland area.

[0009] Further, the following steps are included:

[0010] Use the multi-source data acquisition module to collect basic data of the target farmland area;

[0011] Performing spatiotemporal alignment, anomaly detection and fusion processing on the basic data using a data fusion module;

[0012] The basic data processed by the data fusion module is input into the crop carbon sink calculation model to calculate the carbon sink amount.

[0013] Furthermore, the multi-source data acquisition module is used to collect basic data of the target farmland area, including:

[0014] Using a satellite remote sensing unit to obtain remote sensing information of a target farmland area, the remote sensing information includes crop vegetation index, surface temperature, and evaporative heat dissipation information;

[0015] Use drone inspection units to obtain information on agricultural conditions in target farmland areas;

[0016] Acquire meteorological information of the target farmland area using a meteorological data unit, wherein the meteorological information includes temperature, precipitation, wind speed, and CO2 concentration;

[0017] Use the field IoT unit to obtain soil moisture, soil temperature, and organic carbon content in the target farmland area.

[0018] Furthermore, the data fusion module is used to perform spatiotemporal alignment, anomaly detection and fusion processing on the basic data, including:

[0019] Use a spatiotemporal alignment engine to unify the timestamps and spatial resolutions of different data sources;

[0020] Use anomaly detection units to detect and remove abnormal data;

[0021] A multi-source data fusion unit was used to establish a nonlinear mapping relationship among vegetation index, soil carbon content and meteorological data.

[0022] Furthermore, it also includes a dynamic correction module for adjusting carbon credits in combination with sudden environmental factors.

[0023] Furthermore, it also includes outputting and displaying the carbon sink amount using the result output and evidence storage module.

[0024] In another aspect of the present application, a computer-readable storage medium is provided, which stores a computer program, and when the program is executed by a processor, the method as described in any one of the above items is implemented.

[0025] The crop carbon sink calculation model, crop carbon credit calculation method, and storage medium provided in this application adopt a multi-task learning model based on the Transformer architecture, and the multi-task learning model includes: a simultaneous prediction carbon sink module, a soil carbon storage calculation module, and a carbon emission factor module; the training data of the model includes the design of crop physiology knowledge. Through the multi-task learning model based on the Transformer architecture, the carbon sink can be calculated from multiple aspects such as the simultaneous carbon sink of various above-ground organisms and soil carbon storage. The carbon sink of activities such as fertilization that affect the carbon sink is also calculated in combination with the carbon emission factor module, thereby realizing multi-faceted calculation of carbon sinks. The crop carbon sink calculation model based on multiple dimensions solves the problem of high manual dependence and inability to adapt to large-scale crop carbon monitoring needs; at the same time, each module covers all major aspects that affect changes in crop carbon storage, ensuring coverage and improving the accuracy of carbon sink calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0027] Figure 1 This is a flowchart of an embodiment of a method for calculating crop carbon credits based on a crop carbon sink calculation model provided by this application. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0029] Carbon sinks refer to changes in carbon produced by activities, processes, or mechanisms that absorb atmospheric carbon dioxide and store it in organisms and soils through natural processes such as plant photosynthesis. For example, forest carbon sinks can be represented by the sum of changes in forest carbon storage over a given period. Carbon credits are greenhouse gas emission reduction or sequestration enhancement projects developed in accordance with the rules, procedures, and methodologies established by relevant standards or mechanisms, and are issued after verification, verification, and certification by an independent third-party organization. Each carbon credit represents the reduction or removal of one ton of carbon dioxide equivalent from the air.

[0030] This application aims to solve the problems of existing crop carbon sink calculation methods: 1) low efficiency, manual collection of farmland data (such as soil carbon content, crop type) is time-consuming and labor-intensive, and it is difficult to cover large areas of farmland; 2) insufficient accuracy, a single model ignores multidimensional dynamic factors (such as climate change, crop growth cycle), resulting in large deviations in carbon sink estimation, with an error rate generally exceeding 20%; 3) single data dimension, relying on vegetation indices collected by satellites, without integrating key factors affecting carbon sinks such as precipitation and temperature; 4) lack of dynamics, unable to respond in a timely manner to the impact of extreme weather, pests and diseases on carbon sinks; 5) low transparency: the source of key parameters cannot be traced, affecting the credibility of carbon credit certification, and other issues. The approach is to establish a crop carbon sink calculation model based on multiple dimensions, which covers the simultaneous prediction of carbon sink modules, soil carbon storage calculation modules, and carbon emission factor modules, so that the model covers all aspects of the main carbon sources and main carbon consumption in the crop area. Moreover, the model-based crop carbon credit calculation method provided in this application collects multi-modal basic data, and the basic data is aligned, fused, and other processes before being analyzed by the model, ensuring the comprehensiveness and effectiveness of the input data, and thus achieving the accuracy and comprehensiveness of the output results.

[0031] An embodiment of the present application provides a crop carbon sink calculation model, including a multi-task learning model based on a Transformer architecture, wherein the multi-task learning model includes: a simultaneous carbon sink prediction module, a soil carbon storage calculation module, and a carbon emission factor module; the training data of the model includes crop physiology knowledge.

[0032] In this embodiment, after being trained with crop physiology knowledge, the multi-task learning model based on the Transformer architecture can be processed as input by the Normalized Difference Vegetation Index (NDVI) and surface temperature information collected by remote sensing satellites, crop growth image information collected by drone scanning devices, temperature, precipitation, CO2 concentration and other information obtained from the meteorological data interface, and soil moisture and organic carbon content obtained by field Internet of Things devices. The synchronous carbon sequestration prediction module calculates the simultaneous carbon sequestration of various plants in the target farmland area based on the input, and the soil carbon storage calculation module calculates the soil carbon sequestration capacity of the target farmland area based on the input. At the same time, the carbon emission factor calculates the carbon emissions of activities such as fertilization and tillage. In this embodiment, the Transformer architecture is a deep learning model based on the self-attention mechanism. This model is a prior art and will not be described in detail here.

[0033] The crop carbon sink calculation module provided in this embodiment also includes an interpretability module for generating a heat map of carbon sink feature importance to display the weight distribution of carbon sink contribution factors. In this embodiment, the interpretability module uses the SHAP (SHapley Additive exPlanations) analysis model. The output carbon sink displays the weight distribution of each factor in the form of a heat map to enhance carbon sink transparency.

[0034] The crop carbon sink calculation module provided in this embodiment also includes a dynamic correction module for adjusting carbon credits in combination with sudden environmental factors. In this embodiment, the dynamic correction module includes a collection unit for collecting extreme weather events and agricultural management activities, and a reinforcement learning method triggered when extreme weather or agricultural management activities are collected. The reinforcement learning method dynamically adjusts the parameters of the synchronous prediction carbon sink module, the soil carbon storage calculation module, and the carbon emission factor module. The crop carbon sink calculation model is adaptively corrected in combination with historical data to improve the calculation accuracy. The extreme weather events described in this embodiment include heavy rain and drought; the agricultural management activities include fertilization and plowing. The collection content that can be adopted by the collection unit includes but is not limited to images collected based on remote sensing satellites and sensor collection data. Those skilled in the art can select the collection content and collection method as needed, which will not be elaborated here.

[0035] Another embodiment of the present application further provides a crop carbon credit calculation method based on a crop carbon sink calculation model, which uses the above-mentioned crop carbon sink calculation model to calculate the carbon sink amount of a target farmland area.

[0036] like Figure 1 As shown, the crop carbon credit calculation method provided in this embodiment includes the following steps:

[0037] S1: Use the multi-source data acquisition module to collect basic data of the target farmland area;

[0038] S2: Using a data fusion module to perform spatiotemporal alignment, anomaly detection and fusion processing on the basic data;

[0039] S3: Inputting the basic data processed by the data fusion module into the crop carbon sink calculation model to calculate the carbon sink amount.

[0040] In the crop carbon credit calculation method provided in this embodiment, the multi-source data acquisition module is used to collect basic data of the target farmland area, including:

[0041] Using a satellite remote sensing unit to obtain remote sensing information of a target farmland area, the remote sensing information includes crop vegetation index, surface temperature, and evaporative heat dissipation information;

[0042] Use drone inspection units to obtain information on agricultural conditions in target farmland areas;

[0043] Acquire meteorological information of the target farmland area using a meteorological data unit, wherein the meteorological information includes temperature, precipitation, wind speed, and CO2 concentration;

[0044] Use the field IoT unit to obtain soil moisture, soil temperature, and organic carbon content in the target farmland area.

[0045] The crop vegetation index NDVI described in this embodiment can be calculated using NDVI=(near infrared reflectivity-red light reflectivity) / (near infrared reflectivity+red light reflectivity).

[0046] The drone inspection unit in this embodiment acquires agricultural information through multispectral and thermal imaging. The drone inspection unit in this embodiment collects images of the target farmland area and analyzes the collected images to obtain agricultural information such as plant types. In this embodiment, the analysis of the target farmland area images can be performed by equipping the drone with an image recognition core, which can identify crop type, density, and other information.

[0047] The meteorological data unit in this embodiment is an interface connected to a national meteorological station or a local meteorological station, and obtains temperature, precipitation, wind speed, and CO2 concentration information through data from the meteorological station.

[0048] In this embodiment, the field IoT units include distributed temperature and humidity sensors and organic carbon content sensors buried underground in the target farmland area. The data collected by the sensors is connected to the cloud via low-power wireless transmission (such as LoRa or NB-IoT) for real-time storage. The data is then transmitted from the cloud to the crop carbon sequestration calculation model to calculate the carbon sequestration amount.

[0049] In the crop carbon credit calculation method provided in this embodiment, the data fusion module performs spatiotemporal alignment, anomaly detection, and fusion processing on the basic data, including:

[0050] Use a spatiotemporal alignment engine to unify the timestamps and spatial resolutions of different data sources;

[0051] Use anomaly detection units to detect and remove abnormal data;

[0052] A multi-source data fusion unit was used to establish a nonlinear mapping relationship among vegetation index, soil carbon content and meteorological data.

[0053] The spatiotemporal alignment engine described in this embodiment uses an interpolation algorithm and a time series alignment method to unify the timestamps and spatial resolutions of different data sources. The anomaly detection unit uses an isolation forest or a variational autoencoder (VAE) to detect and remove anomalous data. Those skilled in the art can choose either of these methods for anomaly data detection. Using these two methods to analyze anomaly data is well known in the art and will not be elaborated on here.

[0054] In this embodiment, the data fused by the multi-source data fusion unit has become an integrated whole and is then fed into a transformer-based multi-task learning model for carbon sequestration calculation, ensuring the integrity and comprehensiveness of the calculated carbon sequestration. This transformer-based multi-task learning model utilizes a shared parameter mechanism, allowing different task learning models to share data features, improving the model's generalization capabilities.

[0055] The crop carbon credit calculation method provided in this embodiment further includes a dynamic correction module for adjusting the carbon credit in combination with sudden environmental factors.

[0056] The dynamic carbon sink correction module in this embodiment uses reinforcement learning (RL) to adjust model parameters in response to extreme weather events (such as rainstorms and droughts) and agricultural management activities (such as fertilization and tillage). The adjustment method is based on historical data.

[0057] like Figure 1 As shown, the crop carbon credit calculation method provided in this embodiment also includes S4: using the result output and evidence storage module to output and display the carbon sink amount.

[0058] In this embodiment of the agricultural crop carbon credit calculation method, the output and evidence storage module includes a blockchain interface connected to the blockchain. The blockchain interface calculates the hash value of the carbon sink and sends it to the blockchain. The output carbon sink can be standardized and authenticated to generate a carbon credit report.

[0059] In this embodiment, the carbon sink amount is displayed by displaying the carbon sink spatial distribution and historical trend through a web terminal or a mobile terminal.

[0060] In this embodiment, the satellite remote sensing unit can use the Sentinel-2 satellite to acquire images with a resolution of 10 meters to 50 meters. In this embodiment, the vegetation index NDVI can be calculated using NDVI=(near infrared reflectivity-red light reflectivity) / (near infrared reflectivity+red light reflectivity).

[0061] In this embodiment, the drone inspection unit uses a multispectral or hyperspectral camera carried by the drone to obtain degraded grassland information with a resolution of 0.5 to 1 meter. The multispectral here includes multiple bands, each of which is used to obtain different information: blue light is used to obtain information related to water and soil characteristics, green light is used to obtain information related to vegetation health and chlorophyll concentration, red light (620-750nm) is used to obtain vegetation type and NDVI related information, red edge (700-740nm) is used to monitor plant stress and growth status information, and near infrared is used to obtain vegetation activity and NDVI related information. Thermal infrared and other methods can also be used to obtain soil moisture and temperature change information.

[0062] In this embodiment, each spectrum can be captured using a multi-lens multispectral camera, with each band using a separate lens to capture images, ensuring high spectral resolution. Naturally, the multiple lenses require time synchronization of the images captured, with the spatial alignment between the lenses controlled by a registration system. Spatial registration of images from multiple lenses is well known in the art and will not be discussed further here. Alternatively, an integrated multispectral camera can be used, with a single lens capturing images by switching between wavelengths using a beam splitter prism or filter wheel. While this type of camera may have slightly lower wavelength capture accuracy, it offers advantages such as cost savings and ease of operation. Those skilled in the art can select the camera type as needed, and this will not be discussed further here.

[0063] In order to reduce manual intervention in this embodiment, a route can be set for the drone, and each time the drone patrols, it patrols according to a highly overlapping route.

[0064] In this embodiment, the basic data obtained and processed by the data fusion module are input into the crop carbon sequestration calculation model in the form of a standardized data set. The standardized data set can avoid data loss.

[0065] In this embodiment, the spatiotemporal alignment engine is used to align the drone data with the remote sensing imagery, downsample the drone camera data to the same resolution as the remote sensing imagery, and unify the temporal and spatial resolutions of different data sources, making them comparable and fusible at the same scale. Establishing a mapping relationship between satellite imagery and ground sensor data involves modeling and matching the remote sensing imagery with the real-time ground sensor data in terms of space and attributes. Possible mapping models include linear / nonlinear regression models, or machine learning models such as decision trees (CART, XGBoost), random forests, support vector machines (SVM), and neural networks (MLP, CNN). Those skilled in the art can select these models as needed, and this will not be elaborated on here.

[0066] This application first uses a spatiotemporal alignment engine to unify data of varying resolutions and temporal dimensions. Machine learning and modeling algorithms then establish a mapping relationship between remote sensing data and storefront data. Based on this comprehensive spatial and temporal mapping, the application estimates uninhabited areas, improving the accuracy of data on uninhabited areas. This estimation of uninhabited areas ensures the integrity of the entire forest data, which helps improve the accuracy of carbon sink calculations.

[0067] A computer-readable storage medium stores a computer program, which implements the method described above when executed by a processor.

[0068] The crop carbon sink calculation model, crop carbon credit calculation method, and storage medium provided in this embodiment achieve full automation of the entire process from data acquisition to carbon credit certification through multimodal data collection, data fusion and quality control, large-scale agricultural model calculation, dynamic correction, and result storage. This achieves:

[0069] 1. Efficient automation: No manual intervention is required from data collection to carbon sink output, improving efficiency by more than 90%.

[0070] 2. High-precision calculation: Multimodal data fusion reduces the error rate of carbon sink estimation to less than 5% (traditional methods are 20%).

[0071] 3. Dynamic adaptability: The real-time correction module has a response time of less than 1 hour to extreme weather conditions.

[0072] 4. Transparency and traceability: The explainability module supports layer-by-layer verification of the basis for carbon credit generation.

[0073] 5. Cross-regional generalization: Pre-trained large models adapt to different crop types and geographical environments.

[0074] The present invention covers the entire process of data collection and analysis through multi-source data fusion and intelligent algorithm optimization, realizing the deep integration and dynamic calibration of multimodal data (remote sensing, meteorology, and Internet of Things devices); the interpretability module in the model makes the calculation basis of the output carbon sink or converted carbon credit transparent; it also supports real-time updates and generalization to improve calculation accuracy and credibility.

[0075] It provides an efficient, accurate and transparent solution for calculating crop carbon credits, which can be widely used in carbon market transactions.

[0076] The crop carbon sink calculation method provided in this embodiment can adopt a variety of optimization schemes for different application scenarios.

[0077] For small-scale farmland, only field sensors and meteorological data can be used, eliminating satellite remote sensing and drones to reduce deployment costs. For high-value cash crops, higher-resolution drone remote sensing data, combined with a DNA soil microbial analysis module, can improve the accuracy of soil carbon sink estimation. For extreme environments, low-power IoT (LoRa) data transmission can be used in real time to enhance the system's dynamic response capabilities. For cross-regional, generalized scenarios, agricultural big data from different climate zones can be used as training data to train the agricultural carbon sink calculation model.

[0078] As used below, the term “unit,” “sub-unit,” or “module” may be a combination of software and / or hardware that implements a predetermined function.

[0079] Based on the same inventive concept, an embodiment of the present application further provides a computer-readable storage medium storing a computer program, which implements the method described in the above embodiment when executed by a processor.

[0080] Through the description of the above implementation methods, it can be known that those skilled in the art can clearly understand that the present application can be implemented by means of software plus necessary hardware. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, or can be embodied through the implementation process of data migration. All or part of the present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld devices or portable devices, tablet devices, mobile communication terminals, multi-processor systems, microprocessor-based systems, programmable electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc.

[0081] Although the present application has been described with reference to the embodiments, those skilled in the art will appreciate that there are many modifications and variations to the present application without departing from the spirit of the present application. It is intended that the appended claims include these modifications and variations without departing from the spirit of the present application.

Claims

1. A crop carbon sink calculation model, characterized in that , including a multi-task learning model based on the Transformer architecture, the multi-task learning model includes: a simultaneous carbon sequestration prediction module, a soil carbon storage calculation module and a carbon emission factor module; the training data of the model includes crop physiology knowledge.

2. The crop carbon sequestration calculation model according to claim 1, characterized in that: It also includes an interpretability module for generating a heat map of carbon sequestration feature importance to show the weight distribution of carbon sequestration contribution factors.

3. The crop carbon sequestration calculation model according to claim 1 or 2, characterized in that: It also includes a dynamic correction module for adjusting carbon credits based on sudden environmental factors.

4. A method for calculating crop carbon credits based on a crop carbon sink calculation model, characterized in that: The crop carbon sequestration calculation model as claimed in any one of claims 1 to 3 is used to calculate the carbon sequestration amount of the target farmland area.

5. The crop carbon credit calculation method according to claim 4, characterized in that: The following steps are involved: Use the multi-source data acquisition module to collect basic data of the target farmland area; Performing spatiotemporal alignment, anomaly detection and fusion processing on the basic data using a data fusion module; The basic data processed by the data fusion module is input into the crop carbon sink calculation model to calculate the carbon sink amount.

6. The crop carbon credit calculation method according to claim 5, characterized in that: The multi-source data acquisition module is used to collect basic data of the target farmland area, including: Using a satellite remote sensing unit to obtain remote sensing information of a target farmland area, the remote sensing information includes crop vegetation index, surface temperature, and evaporative heat dissipation information; Use drone inspection units to obtain information on agricultural conditions in target farmland areas; Acquire meteorological information of the target farmland area using a meteorological data unit, wherein the meteorological information includes temperature, precipitation, wind speed, and CO2 concentration; Use the field IoT unit to obtain soil moisture, soil temperature, and organic carbon content in the target farmland area.

7. The crop carbon credit calculation method according to claim 5, characterized in that: The data fusion module is used to perform spatiotemporal alignment, anomaly detection and fusion processing on the basic data, including: Use a spatiotemporal alignment engine to unify the timestamps and spatial resolutions of different data sources; Use anomaly detection units to detect and remove abnormal data; A multi-source data fusion unit was used to establish a nonlinear mapping relationship among vegetation index, soil carbon content and meteorological data.

8. The crop carbon credit calculation method according to claim 5, characterized in that: It also includes a dynamic correction module for adjusting carbon credits based on sudden environmental factors.

9. The crop carbon credit calculation method according to any one of claims 4 to 8, characterized in that: It also includes the use of result output and evidence storage modules to output and display carbon sink amounts.

10. A computer-readable storage medium storing a computer program, characterized in that: When the program is executed by a processor, the method according to any one of claims 3 to 9 is implemented.

Images