CN-121981599-A - Evaluation method for farmland carbon sink datum line measurement and additional demonstration

CN121981599ACN 121981599 ACN121981599 ACN 121981599ACN-121981599-A

Abstract

The invention discloses an evaluation method for farmland carbon sink datum line measurement and additional demonstration, which relates to the technical field of agricultural environment monitoring management and comprises the steps of establishing an area attribute knowledge graph based on multi-source heterogeneous data of a target farmland area including soil, climate and rotation history; training a carbon sequestration potential model by combining a knowledge graph and an reinforcement learning algorithm to generate a farmland carbon sequestration datum line with regional adaptability, and deploying an intelligent sensing system and edge computing equipment in a target farmland region. According to the invention, the dynamically updated regional attribute knowledge graph is constructed by fusing the multisource heterogeneous data, so that multidimensional information such as soil, climate, crop rotation, management records and the like is effectively integrated, a structured farmland ecological system knowledge base is formed, incremental update can be automatically triggered according to real-time monitoring data and external events, and the baseline modeling and evaluation process is ensured to be always based on the latest regional state.

Inventors

SONG GUOYING
LIU GUOYI
Bian Bazhuoma
ZHOU DI
Secondary overpowering pile
Bian Bazhaxi
Nima Draga

Assignees

西藏自治区农牧科学院农业资源与环境研究所

Dates

Publication Date: 20260505
Application Date: 20260113

Claims (10)

1. An evaluation method for farmland carbon sink baseline measurement and additional demonstration, which is characterized by comprising the following steps: step 1, establishing an area attribute knowledge graph based on multi-source heterogeneous data of a target farmland area including soil, climate and rotation history; step 2, training a carbon sequestration potential model by combining a knowledge graph and an enhanced learning algorithm, and generating a farmland carbon sequestration datum line with regional adaptability; step 3, deploying an intelligent sensing system and edge computing equipment in a target farmland area, and monitoring actual carbon sink data including soil respiration rate, microbial community activity and greenhouse gas flux in real time; Step 4, carrying out local preprocessing and anomaly detection on the monitoring data through edge computing equipment, extracting carbon flux characteristics and uploading the carbon flux characteristics to a cloud collaborative analysis platform in real time; Step5, constructing a carbon leakage risk simulation model at the cloud, fusing monitoring data with a history management record, and simulating carbon leakage paths and strength changes under different farmland management scenes; step 6, calculating the difference value between the actual carbon sink quantity and the reference line carbon sink quantity based on the farmland carbon sink reference line and the real-time monitored actual carbon sink data, and identifying the additional space-time distribution and contribution factors generated by carbon sink; And 7, carrying out systematic correction on the additivity by combining the carbon leakage simulation result, quantifying the influence of the leakage risk on the net carbon sink, and forming an additivity demonstration report.
2. The method for determining a reference line of carbon sink and carrying out additional demonstration according to claim 1, wherein the step1 specifically comprises the following steps: the method comprises the steps of obtaining multisource heterogeneous data of a target farmland area, wherein the multisource heterogeneous data comprise soil physicochemical properties, climate factors, crop rotation histories and farmland management records; Carrying out structural fusion and semantic modeling on the multi-source heterogeneous data to construct a regional attribute knowledge graph of a regional agricultural ecological system; And dynamically updating the regional attribute knowledge graph according to the change of soil, climate and management practice to form a dynamic knowledge base reflecting the regional attribute time-space evolution.
3. The method for determining a reference line of carbon sink and carrying out additional demonstration according to claim 1, wherein the step 2 specifically comprises the following steps: based on the regional attribute knowledge graph, extracting the potential of a soil carbon reservoir, the carbon fixation efficiency of crops and the climate response characteristics as model input; training a carbon sequestration potential model in a simulation environment by using an reinforcement learning algorithm, and optimizing model parameters through multi-round environment states and decision interactions; and outputting a farmland carbon sink datum line with regional adaptability, and reflecting carbon fixation potential changes under different management measures and climatic conditions.
4. The method for determining a reference line of carbon sink for farmland and for additional demonstration according to claim 3, wherein S2 further comprises: Constructing an agricultural ecological system simulation environment containing soil carbon dynamics, crop growth and greenhouse gas emission; strategy exploration and decision optimization are carried out in the simulation environment of the agricultural ecological system through the reinforcement learning intelligent agent, and an optimal carbon fixation management strategy is learned; and combining the historical climate sequence and the rotation mode to generate a carbon sink datum line curve under multiple scenes.
5. The method for determining a reference line of carbon sink and carrying out additional demonstration according to claim 1, wherein the step 3 specifically comprises the following steps: arranging a multi-parameter intelligent sensing system in a target farmland area, and collecting actual carbon sink data including soil respiration, microbial activity and greenhouse gas flux data in real time; And deploying edge computing equipment, realizing local caching, preliminary cleaning and anomaly marking of the monitoring data, and constructing an in-situ high-frequency monitoring data real-time uploading mechanism.
6. The method for determining a reference line of carbon sink and carrying out additional demonstration according to claim 1, wherein the step 4 specifically comprises the following steps: performing time sequence data filtering, missing value interpolation and outlier detection in edge computing equipment to finish local preprocessing of monitoring data; Extracting carbon flux characteristic data comprising soil carbon flux, greenhouse gas emission intensity and microbial activity index to form a carbon flux characteristic sequence, and uploading the processed carbon flux characteristic sequence data to a cloud collaborative analysis platform in real time.
7. The method for determining a reference line of carbon sink and performing additional demonstration according to claim 1, wherein the step 5 comprises the following steps: Constructing a carbon leakage risk simulation model in a cloud collaborative analysis platform, and integrating a soil carbon migration, cultivation disturbance and nutrient management module; Combining real-time farmland carbon process in-situ high-frequency monitoring data, historical management records and regional attribute knowledge maps, and simulating carbon leakage paths under different farmland management measures; And outputting a carbon leakage intensity space-time distribution map under each scene, and quantitatively managing and adjusting the systematic carbon leakage risk caused by the adjustment.
8. The method for determining a reference line of carbon sink and performing additional demonstration according to claim 1, wherein the step 6 specifically comprises the following steps: calculating a difference value between the actual carbon sink quantity and the datum line carbon sink quantity based on the farmland carbon sink datum line and the real-time monitored actual carbon sink data; Dividing high, medium and low extra grades and time periods by adopting a spatial clustering and time sequence analysis method; And identifying dominant driving factors of the additivity of each region and interaction effects thereof through correlation analysis and attribution models, constructing an additivity contribution degree evaluation matrix, and quantifying the relative influence of different management measures on the additivity formation.
9. The method for determining a reference line of carbon sink and additional demonstration according to claim 8, wherein S6 further comprises: Based on the additivity grading result, combining the obtained carbon sink additivity difference matrix, identifying the space-time distribution characteristics of the carbon sink additivity, and extracting main contribution factors including crop types, fertilization modes and irrigation management; generating a carbon sink extra distribution map and a contribution factor weight table, wherein the carbon sink extra space distribution map is colored by taking a field block as a unit, and the contribution factor weight table is a two-dimensional table.
10. The method for determining a reference line of carbon sink and performing additional demonstration according to claim 1, wherein the step 7 specifically comprises the following steps: combining with the carbon leakage risk simulation result, systematically correcting the preliminary identified carbon sink additionally, and deducting the carbon loss caused by leakage; calculating net carbon sink quantity to form a multidimensional assessment index system comprising actual carbon sink, reference line carbon sink, leakage correction and net extra; and generating an additional evaluation report with complete structure, traceable data and transparent demonstration process, and supporting scientific demonstration and decision application of carbon sink projects.

Description

Evaluation method for farmland carbon sink datum line measurement and additional demonstration Technical Field The invention relates to the technical field of agricultural environment monitoring management, in particular to an evaluation method for farmland carbon sink datum line measurement and additional demonstration. Background Along with the increasing serious influence of climate change, carbon sink is widely paid attention to as an effective means for slowing down greenhouse gas emission, in a farmland ecological system, factors such as vegetation coverage, soil management, agricultural practice and the like affect carbon fixation and release together, therefore, accurate determination of carbon sink reference lines of farmlands is crucial for evaluating benefits of different agricultural management measures in terms of carbon sink, wider acceptance and support can be obtained only when carbon sink projects can provide additional environmental benefits, and whether actual carbon emission reduction effects brought by implementing specific agricultural measures exceed the existing emission reduction level can be determined by adopting additional arguments, so that the method has a key meaning for establishing credit and effectiveness of carbon markets. For example, the method for estimating the carbon sink of the farmland ecosystem based on the agricultural input-output data disclosed in China patent publication No. 115797093A improves the scientificity and convenience of the method for measuring the carbon sink of the farmland, facilitates the carbon sink obtained by final calculation to be more close to the actual carbon sink of the farmland, and facilitates the acquisition of more accurate carbon sink of the farmland in a short time. In current farmland carbon sink assessment practice, the carbon sink quantity is estimated mainly by depending on static agricultural input-output data, and a carbon sink reference line reflecting the influences of complex factors such as regional climate, soil physicochemical properties, crop rotation, management history and the like is difficult to dynamically construct, so that the deviation source between the real carbon sink quantity and the reference line is difficult to scientifically identify, further the effective demonstration of influencing the carbon sink additivity is difficult to realize, meanwhile, systematic carbon leakage risks possibly caused by farmland management measure adjustment are less considered, and the additivity assessment result has obvious defects in the aspects of integrity and credibility. Disclosure of Invention In order to solve the technical problems, the invention is realized by the following technical scheme that the evaluation method for determining and additionally proving the reference line of farmland carbon sink comprises the following steps: Step 1, establishing a region attribute knowledge graph capable of being dynamically updated based on multi-source heterogeneous data of a target farmland region including soil, climate and rotation history; Step 2, training a carbon sequestration potential model by combining a knowledge graph and an reinforcement learning algorithm, and generating a farmland carbon sink reference line with regional adaptability by simulating the multi-environment state interaction dynamic optimization carbon sequestration potential model; Step 3, deploying an intelligent sensing system and edge computing equipment in a target farmland area, monitoring actual carbon sink data comprising soil respiration rate, microbial community activity and greenhouse gas flux in real time, and collecting farmland carbon process in-situ high-frequency monitoring data; Step 4, carrying out local preprocessing and anomaly detection on the monitoring data through edge computing equipment, extracting carbon flux characteristics and uploading the carbon flux characteristics to a cloud collaborative analysis platform in real time; Step5, constructing a carbon leakage risk simulation model at the cloud, fusing monitoring data with a history management record, and simulating carbon leakage paths and strength changes under different farmland management scenes; step 6, calculating the difference value between the actual carbon sink quantity and the reference line carbon sink quantity based on the farmland carbon sink reference line and the real-time monitored actual carbon sink data, and identifying the additional space-time distribution and contribution factors generated by carbon sink; and 7, carrying out systematic correction on the additivity by combining the carbon leakage simulation result, and quantifying the influence of the leakage risk on the net carbon sink to form a complete and verifiable additivity demonstration report. Preferably, the step1 specifically includes: acquiring multisource heterogeneous data of a target farmland area, wherein the multisource heterogeneous data comprise soil physicochemical properties, cli