Forest grass wet carbon sink metering method and system

Abstract

The invention discloses a forest grass wet carbon sink metering method and system, and relates to the field of carbon sink metering, and the method comprises the following steps: 1, collecting and processing multi-source data, 2, fusing the data, aligning the multi-source data, adopting a geographic registration and time synchronization technology, and unifying a spatial resolution and a timestamp, and 3, extracting features. 4, constructing a carbon sink dynamic model; and 5, carrying out carbon sink visualization and decision support. According to the forest and grass wet carbon sink metering method and system, multi-source data collaboration is achieved, optical, radar and laser radar data are fused, the defect that traditional carbon sink metering depends on ground sample plot survey or single remote sensing data is overcome, carbon sink contributions of forests, grass and wet lands are accurately distinguished, the overall spatial resolution is higher, cloud and mist interference is avoided, the single data source error is solved, and the measurement accuracy is improved. Through multi-source data fusion and dynamic model optimization, high-precision measurement and real-time monitoring of forest grass wet carbon sink are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon sink measurement, and in particular to a forest and grassland wet carbon sink measurement method and system. Background Art

[0002] Forest and grassland carbon sink measurement refers to the process of quantifying and evaluating the carbon dioxide absorbed and stored by ecosystems such as forests, grasslands, and wetlands over a period of time. Its basic principles are based on the carbon cycle and biogeochemical processes. The main steps include determining the carbon sink type, selecting the measurement method, collecting data, calculating carbon storage, and considering the dynamic changes of carbon sinks.

[0003] For example, the patented forestry carbon sink dynamic measurement method with publication number CN119888472A includes the following steps: S100, obtaining a remote sensing image of the land to be measured; S200, processing the remote sensing image to extract characteristic data for calculating the carbon sink of the land to be measured; S300, inputting the characteristic data into a calculation model to calculate the carbon sink of the land to be measured. The forestry carbon sink dynamic measurement method obtains remote sensing data by data processing, extracts data from multiple aspects of the remote sensing image, establishes a calculation model, and thus realizes accurate monitoring of carbon sinks. At the same time, it greatly overcomes the shortcomings of traditional carbon sink monitoring, such as time-consuming and labor-intensive, and small scope, and is also suitable for monitoring forest carbon reserves and carbon sinks on a large spatial scale.

[0004] Another example is the patented ecological slope carbon sink measurement method with publication number CN119577580A. The ecological slope is divided into multiple independent areas according to the terrain characteristics, and divided into different slope grades according to the slope; within each partition, it is further classified into different vegetation type areas according to the vegetation type, namely classification areas; representative sample plots are selected in each classification area, and the size of the sample plots is adjusted according to different vegetation types; the carbon sequestration amount of the vegetation in each sample plot is calculated; the measurement results are corrected according to different slopes and vegetation types; the carbon sink amounts of all sample plots are summarized. The provided ecological slope carbon sink measurement method comprehensively considers the terrain characteristics and vegetation types of the ecological slope, is suitable for carbon sink measurement schemes under complex environmental conditions, and improves the accuracy, reliability and operability of carbon sink measurement in ecological slope environments;

[0005] For example, the patent CN117151921B discloses a carbon sink measurement system and method for planting agriculture, which includes: a carbon total fixation estimation module, which obtains planting-related information of plants in a target area and estimates the total carbon fixation value of the target area during the planting period; a carbon consumption estimation module, which obtains planting consumption information of plants in the target area and estimates the carbon consumption value of the target area during the planting period; a waste treatment module, which obtains planting recycling information and determines the planting recycling carbon value and the planting return carbon value; a carbon sink measurement module, which determines the carbon sink measurement value of the target area during the planting period and estimates the total carbon fixation value, carbon consumption value, planting recycling carbon value and termination return carbon value of plants in the target area, thereby realizing carbon sink measurement estimation of planting agriculture within a regional scope;

[0006] Most of the above-mentioned existing technologies have improved their overall structure. However, in the actual application of existing forest, grassland and wetland carbon sink measurement methods, traditional carbon sink measurement relies on ground sample surveys or single remote sensing data, which makes it difficult to accurately distinguish the carbon sink contributions of forests, grasslands and wetlands, and the spatial resolution is insufficient. For example, optical remote sensing is easily affected by clouds and fog, and existing models are mostly based on static estimates of historical data, which cannot dynamically reflect the impact of extreme climate and human interference such as deforestation and fire on carbon sinks. Summary of the Invention

[0007] The purpose of the present invention is to provide a forest, grassland and wetland carbon sink measurement method and system to solve the problems raised in the above background technology that traditional carbon sink measurement relies on ground sample surveys or single remote sensing data, which makes it difficult to accurately distinguish the carbon sink contributions of forests, grasslands and wetlands, and the spatial resolution is insufficient. For example, optical remote sensing is easily interfered by clouds and fog, and existing models are mostly based on static estimates of historical data and cannot dynamically reflect the impact of extreme climate and human interference such as deforestation and fire on carbon sinks.

[0008] To achieve the above-mentioned objectives, the present invention provides the following technical solutions: a method and system for measuring forest and grassland wetland carbon sinks, comprising the following steps: 1. Acquisition and processing of multi-source data: a. Collection of satellite remote sensing data, acquisition of multispectral (Sentinel-2), radar (Sentinel-1) and lidar (GEDI) data, extraction of parameters such as vegetation coverage, biomass, and surface humidity; b. Reception of ground-measured data, collection of sample plot carbon storage, soil respiration rate, etc., through Internet of Things devices such as soil carbon flux meters and tree growth ring sensors; and c. Auxiliary data collection, integration of meteorological data, such as temperature, precipitation, and land use change data (deforestation, wetland degradation).

[0009] Step 2: Data fusion, through multi-source data alignment, using georeferencing and time synchronization technology to unify spatial resolution and timestamp;

[0010] Step 3: Feature extraction;

[0011] Step 4: Carbon sink dynamic model construction, a: Machine learning model training; Data input: Multi-source feature data + ground-truth carbon sink (label); Generate model: Use LightGBM algorithm to build carbon sink prediction model, and combine SHAP value to analyze feature contribution; b: Dynamic calibration mechanism; Introduce time series analysis (LSTM) to predict carbon sink change trend; Combine meteorological data to adjust model parameters (such as drought index influence coefficient) in real time;

[0012] Step 5: Carbon sink visualization and decision support.

[0013] Further preferably, the step 3: feature extraction includes the following steps:

[0014] a: Extracting texture features (such as gray-level co-occurrence matrix) and vegetation indices (such as EVI and LSWI) from remote sensing images; b: Generating a canopy height model (CHM) based on LiDAR point cloud data; c: Inverting surface moisture and biomass using SAR data.

[0015] Further preferably, the step 5: carbon sink visualization and decision support includes the following steps: a: generating a carbon sink spatial distribution heat map and a time series change curve; b: setting a threshold warning (such as triggering an alarm when the annual carbon sink decline rate is >5%); c: outputting carbon sink enhancement strategy recommendations (such as identifying priority areas for wetland restoration).

[0016] Furthermore, the overall accuracy is improved, and the carbon sink prediction error is reduced by 40% compared with the traditional NDVI model (measured RMSE ≤ 0.8tC / ha); the overall real-time performance is guaranteed, supporting monthly updates, and generating carbon sink loss assessment reports within 72 hours after extreme events (such as typhoons).

[0017] Further optimization can be applied to the government side: policy effect evaluation, ecological compensation fund distribution; enterprise side: carbon sink trading project development (such as CCER); scientific research side: research on the mechanism of global change affecting carbon sinks.

[0018] Further optimization is to achieve multi-source data collaboration: integrate optical, radar, and lidar data to solve the errors of single data sources; dynamic model optimization: combine real-time meteorological data with historical trends to achieve dynamic calibration of carbon sinks; enhance interpretability: quantify the contribution of each feature to carbon sinks through SHAP values to support scientific decision-making.

[0019] Further preferably, a measurement system for the carbon sink measurement method is also disclosed, and the specific construction architecture is as follows, including a data acquisition layer, a data processing layer, a model calculation layer and an application layer, wherein: the data acquisition layer: collects data through a satellite remote sensing receiving terminal, a ground Internet of Things sensor, and a meteorological data interface;

[0020] Data processing layer: Multi-source data cleaning, fusion and feature extraction are achieved through distributed computing clusters (Spark);

[0021] Model calculation layer: Model calculation is performed through the LightGBM carbon sink prediction model combined with the LSTM dynamic trend prediction model;

[0022] Application layer: Supports third-party system call processing through WebGIS platform visualization and API interface.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] This forest, grassland, and wetland carbon sink measurement method and system achieves multi-source data collaboration and integrates optical, radar, and lidar data. This avoids the defects of traditional carbon sink measurement that rely on ground sample surveys or single remote sensing data, accurately distinguishes the carbon sink contributions of forests, grasslands, and wetlands, and has higher overall spatial resolution. It avoids cloud and fog interference, solves the error of a single data source, and achieves high-precision measurement and real-time monitoring of forest, grassland, and wetland carbon sinks through multi-source data fusion and dynamic model optimization.

[0025] Dynamic model optimization is performed, combining static estimates with historical data, while also dynamically reflecting the impacts of extreme climate and human disturbances, such as deforestation and fires, on carbon sinks. Combining real-time meteorological data with historical trends enables dynamic calibration of carbon sinks, such as rapid assessment of carbon losses after fires.

[0026] The overall interpretability is enhanced, and the contribution of various features to carbon sequestration, such as canopy height and soil moisture, is quantified through SHAP values. The collaborative calibration of ground-based measured data and remote sensing data is efficient, combined with an automated fusion mechanism to support scientific decision-making and improve overall practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a flow chart of the forest and grass wet carbon sink measurement method of the present invention;

[0028] Figure 2 It is a flow chart of the forest and grass wet carbon sink measurement system of the present invention. DETAILED DESCRIPTION

[0029] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0030] Example 1: Please refer to Figure 1-Figure 2The present invention provides the following technical solutions: A forest and grass wet carbon sink measurement method, comprising the following steps: Step 1: Collection and processing of multi-source data

[0031] a: Satellite remote sensing data collection, acquiring multispectral (Sentinel-2), radar (Sentinel-1) and lidar (GEDI) data, extracting parameters such as vegetation cover, biomass, and surface moisture; b: Ground-based measured data reception, collecting plot carbon storage, soil respiration rate, etc. through IoT devices such as soil carbon flux meters and tree growth ring sensors; c: Auxiliary data collection, integrating meteorological data such as temperature, precipitation, and land use change data (deforestation, wetland degradation).

[0032] Step 2: Data Fusion

[0033] By aligning multi-source data, using georeferencing and time synchronization technology, the spatial resolution and time stamp are unified;

[0034] Step 3: Feature extraction;

[0035] It includes the following steps:

[0036] a: Extracting texture features (such as gray-level co-occurrence matrix) and vegetation indices (such as EVI and LSWI) from remote sensing images; b: Generating a canopy height model (CHM) based on LiDAR point cloud data; c: Inverting surface moisture and biomass using SAR data

[0037] Step 4: Construction of carbon sink dynamic model

[0038] a: Machine learning model training; Data input: multi-source feature data + ground-truth carbon sink (label); Generate model: Use LightGBM algorithm to build carbon sink prediction model, and combine SHAP value to analyze feature contribution;

[0039] b: Dynamic calibration mechanism; introduce time series analysis (LSTM) to predict carbon sink trends; combine meteorological data to adjust model parameters (such as drought index influence coefficient) in real time;

[0040] Step 5: Carbon sink visualization and decision support, including the following steps: a: Generate carbon sink spatial distribution heat map and time series change curve; b: Set threshold warning (such as triggering an alarm when the annual carbon sink decline rate is >5%); c: Output carbon sink enhancement strategy recommendations (such as identifying priority areas for wetland restoration).

[0041] The overall accuracy is improved. Compared with the traditional NDVI model, the carbon sink prediction error is reduced by 40% (measured RMSE ≤ 0.8tC / ha). The overall real-time performance is guaranteed, supporting monthly updates, and generating carbon sink loss assessment reports within 72 hours after extreme events (such as typhoons).

[0042] It can be applied to the government side: policy effect evaluation, ecological compensation fund allocation; the enterprise side: carbon sink trading project development (such as CCER); the scientific research side: research on the mechanism of global change affecting carbon sinks.

[0043] Achieve multi-source data collaboration: integrate optical, radar, and lidar data to resolve errors in a single data source; dynamic model optimization: combine real-time meteorological data with historical trends to achieve dynamic calibration of carbon sinks; enhance interpretability: quantify the contribution of each feature to carbon sinks through SHAP values to support scientific decision-making.

[0044] Specifically:

[0045] When applied to forest carbon sink measurement:

[0046] First, data collection: Sentinel-2 multispectral imagery (10m resolution) and GEDI lidar data are acquired;

[0047] Ground sensors were deployed to monitor the tree diameter at breast height and soil organic carbon content in the sample plots.

[0048] Model training: Input: canopy height (LiDAR), EVI index (Sentinel-2), soil moisture (SAR) + ground-truth carbon storage;

[0049] Output: Train the LightGBM model. SHAP analysis shows that the canopy height contributes 45%;

[0050] Application output: Identify carbon sink hotspots in a protected area and recommend prioritizing the protection of high-carbon-density old-growth forests;

[0051] Realize multi-source data collaboration, integrate optical, radar, and lidar data, avoid the defects of traditional carbon sink measurement relying on ground sample surveys or single remote sensing data, accurately distinguish the carbon sink contributions of forests, grasslands, and wetlands, have higher overall spatial resolution, avoid cloud and fog interference, solve the error of single data source, and realize high-precision measurement and real-time monitoring of forest, grassland, and wetland carbon sinks through multi-source data fusion and dynamic model optimization.

[0052] Example 2: Based on Example 1, a measurement system for a carbon sink measurement method is also disclosed. The specific construction architecture is as follows, including a data acquisition layer, a data processing layer, a model calculation layer, and an application layer, wherein:

[0053] Data collection layer: Data collection is carried out through satellite remote sensing receiving terminals, ground IoT sensors, and meteorological data interfaces;

[0054] Data processing layer: Multi-source data cleaning, fusion and feature extraction are achieved through distributed computing clusters (Spark);

[0055] Model calculation layer: Model calculation is performed through the LightGBM carbon sink prediction model combined with the LSTM dynamic trend prediction model;

[0056] Application layer: Supports third-party system call processing through WebGIS platform visualization and API interface.

[0057] Example 3: Based on Examples 1 and 2, this solution is used for early warning of wetland carbon sink loss. It combines Sentinel-1 radar data to monitor wetland water level changes and predict the risk of carbon release caused by drought.

[0058] After the model warning, the management department implemented ecological water replenishment to avoid a 15% decrease in carbon sinks. It combined historical data for static estimation and simultaneously dynamically reflected the impact of extreme climate and human interference, such as deforestation and fire, on carbon sinks. It combined real-time meteorological data with historical trends to achieve dynamic calibration of carbon sinks, such as rapid assessment of carbon loss after a fire. It quantified various characteristics, such as canopy height and soil moisture, and their contribution to carbon sinks through SHAP values. The collaborative calibration of ground-based measured data and remote sensing data was efficient, and combined with an automated fusion mechanism to support scientific decision-making.

[0059] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "connected" and "connection" should be understood in a broad sense. For example, they can refer to fixed connection, detachable connection, or integral connection; mechanical connection, electrical connection; direct connection, or indirect connection through an intermediary. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0060] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for measuring forest and grassland wet carbon sinks, characterized in that: The steps include: Step 1: Collection and processing of multi-source data a: Satellite remote sensing data collection, acquiring multispectral (Sentinel-2), radar (Sentinel-1) and lidar (GEDI) data, extracting parameters such as vegetation cover, biomass, and surface humidity; b: Ground measurement data reception, through IoT devices such as soil carbon flux meters and tree growth ring sensors, to collect carbon storage and soil respiration rate of sample sites; c: Auxiliary data collection, integration of meteorological data, such as temperature, precipitation, and land use change data (deforestation, wetland degradation). Step 2: Data Fusion By aligning multi-source data, using georeferencing and time synchronization technology, the spatial resolution and time stamp are unified; Step 3: Feature extraction; Step 4: Construction of carbon sink dynamic model a: Machine learning model training; Data input: multi-source feature data + ground-truth carbon sinks (labels); Generate model: LightGBM algorithm is used to build a carbon sequestration prediction model, and SHAP value is combined to analyze the feature contribution; b: dynamic calibration mechanism; Introducing time series analysis (LSTM) to predict carbon sink trends; Adjust model parameters (such as drought index influence coefficient) in real time based on meteorological data; Step 5: Carbon sink visualization and decision support.

2. A forest and grass wet carbon sink measurement method according to claim 1, characterized in that: The step 3: feature extraction includes the following steps: a: Extract texture features (such as gray-level co-occurrence matrix) and vegetation index (such as EVI, LSWI) of remote sensing images; b: Generate canopy height model (CHM) based on LiDAR point cloud data; c: Surface moisture and biomass were inverted using SAR data.

3. A forest and grass wet carbon sink measurement method according to claim 2, characterized in that: Step 5: Carbon sink visualization and decision support includes the following steps: a: Generate carbon sink spatial distribution heat map and time series change curve; b: Set threshold warnings (e.g. trigger an alarm when the annual decline rate of carbon sinks exceeds 5%); c: Output carbon sink enhancement strategy recommendations (such as identification of priority areas for wetland restoration).

4. A method for measuring forest and grassland wet carbon sinks according to claim 3, characterized in that: The overall accuracy is improved, and the carbon sequestration prediction error is reduced by 40% compared with the traditional NDVI model (measured RMSE ≤ 0.8tC / ha); Overall real-time performance is guaranteed, supporting monthly updates and generating carbon sink loss assessment reports within 72 hours after extreme events (such as typhoons).

5. The method for measuring forest and grassland wet carbon sinks according to claim 3, characterized in that: It can be applied to the government side: policy effect evaluation, ecological compensation fund allocation; the enterprise side: carbon sink trading project development (such as CCER); the scientific research side: research on the mechanism of global change affecting carbon sinks.

6. The method for measuring forest and grassland wet carbon sinks according to claim 3, characterized in that: Achieve multi-source data collaboration: integrate optical, radar, and lidar data to resolve errors from single data sources; Dynamic model optimization: combining real-time meteorological data with historical trends to achieve dynamic calibration of carbon sinks; Enhanced interpretability: SHAP values are used to quantify the contribution of each feature to carbon sequestration, supporting scientific decision-making.

7. A forest and grassland wet carbon sink measurement method according to claim 6, characterized in that: The carbon sequestration measurement method also discloses a measurement system. The specific construction architecture is as follows, including a data acquisition layer, a data processing layer, a model calculation layer, and an application layer, wherein: Data collection layer: Data collection is carried out through satellite remote sensing receiving terminals, ground IoT sensors, and meteorological data interfaces; Data processing layer: Use distributed computing clusters (Spark) to achieve multi-source data cleaning, fusion and feature extraction; Model calculation layer: Model calculation is performed by combining the LightGBM carbon sink prediction model with the LSTM dynamic trend prediction model; Application layer: Through WebGIS platform visualization and API interface, it supports third-party system call processing.

Images