A coastal wetland carbon storage dynamic monitoring method based on time series remote sensing

Abstract

The present application relates to the technical field of wetland data monitoring, and discloses a coastal wetland carbon storage dynamic monitoring method based on time series remote sensing, comprising: collecting the vegetation condition of the coastal wetland, estimating the carbon storage of the coastal wetland, judging whether the carbon storage is abnormal, and adjusting the detection scheme accordingly, the coastal wetland carbon storage dynamic monitoring method based on time series remote sensing comprehensively considers remote sensing monitoring, ground investigation, laboratory analysis and other methods to ensure the comprehensiveness and accuracy of carbon storage monitoring, through twice judgment, combined with historical data and verification conclusion, dynamically adjusting the comparison threshold, adapting to the carbon storage change situation of different sub-regions, from regional division to overall analysis, forming a systematic monitoring process, which can effectively identify abnormal conditions in the target area, through comparison and verification and collection verification, the double verification mechanism improves the reliability of carbon storage evaluation, and provides reliable data support for subsequent monitoring and management.

Description

Technical Field

[0001] This invention relates to the field of wetland data monitoring technology, specifically to a method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing. Background Technology

[0002] Coastal wetlands, including mangroves, salt marshes, and seagrass beds, are important carbon sinks, playing an indispensable role in the global carbon cycle. Through photosynthesis and the accumulation of organic matter, they effectively fix and store atmospheric carbon dioxide, mitigating climate change. Accurate and dynamic monitoring of the carbon storage and changes in coastal wetlands is crucial for assessing their ecological value, developing effective wetland protection and restoration strategies, and quantifying their contribution to carbon neutrality goals. The core of time-series remote sensing-based dynamic monitoring methods for coastal wetland carbon storage is to fully utilize the rich information in time-series remote sensing data, combining multi-source data fusion and advanced modeling techniques (statistical, machine learning, and ecological models) to invert and track the dynamic changes in coastal wetland carbon storage. The overall approach is to acquire continuous, high-frequency remote sensing data to capture key information such as phenological changes, vegetation growth, and land cover changes in wetland ecosystems. This data is then combined with a carbon storage estimation model to achieve spatiotemporal dynamic monitoring of carbon storage (including vegetation carbon storage and soil carbon storage). The aim is to address the shortcomings of existing coastal wetland carbon storage monitoring methods in terms of accuracy, spatiotemporal coverage, dynamism, and adaptability to complex ecosystems, and to achieve high-precision, full-cycle, and high-timeliness monitoring of coastal wetland carbon storage and its changing processes.

[0003] Existing methods for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing cannot reasonably divide the area into several sub-regions based on the vegetation conditions in images of coastal wetlands. They cannot infer the carbon storage conditions and changes in carbon storage of each sub-region based on the vegetation conditions of each sub-region, nor can they determine whether there are similar sub-regions within the overall area, or whether the inferred carbon storage conditions and changes in carbon storage of the current sub-region are significantly inaccurate based on the carbon storage conditions and changes in carbon storage of similar sub-regions. Furthermore, they cannot determine whether there are anomalies in the overall carbon storage monitoring based on the carbon storage conditions and changes in carbon storage of each sub-region. These limitations can easily reduce the accuracy of monitoring and assessment, increase the false judgment rate, and thus limit their practicality. Summary of the Invention

[0004] This invention provides a method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing, which helps to solve the problems mentioned in the background art.

[0005] This invention provides the following technical solution: a method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing, comprising:

[0006] Obtain the set of regions within the target area;

[0007] Perform parametric analysis sequentially on each sub-region within the region set;

[0008] For each sub-region within the region set, a preliminary judgment is made as to whether the parameter changes are normal, and a preliminary judgment conclusion is generated;

[0009] For each sub-region within the region set, query whether it contains similar regions;

[0010] If similar regions exist, perform a comparison verification.

[0011] If no similar region exists, perform data collection and verification.

[0012] Based on the verification conclusions obtained from comparative verification or data collection verification, a secondary judgment is made on whether the parameter changes in the sub-region are normal, and a secondary judgment conclusion is generated.

[0013] The target region is analyzed as a whole by combining the preliminary and secondary judgment results of each sub-region.

[0014] As an optional scheme of the time-series remote sensing-based dynamic monitoring method for carbon storage in coastal wetlands described in this invention, wherein: obtaining a set of regions of the target area specifically involves: acquiring images of the target area, denoted as... Use image segmentation algorithms to process images Vegetation classification was performed to obtain a vegetation classification map. According to the vegetation classification map The target area is divided into several sub-regions. Check each sub-region The distribution of vegetation types within the region; integrating all sub-regions yields the final region set, denoted as . .

[0015] As an optional scheme of the time-series remote sensing-based dynamic monitoring method for carbon storage in coastal wetlands described in this invention, the following steps are performed: Parameter analysis is sequentially performed on each sub-region within the region set, specifically: Images of the sub-regions are extracted, denoted as... Use image segmentation algorithms to segment sub-regions of the image. The vegetation types were classified to obtain a vegetation classification map, denoted as... Using field survey data The classification results are corrected to obtain a corrected vegetation classification map, denoted as... Calculate vegetation indices to assess vegetation growth characteristics; retrieve vegetation cover. The study assesses vegetation growth status and carbon storage potential; a remote sensing estimation model for carbon storage is constructed based on vegetation indices and field sample data, denoted as [model name missing]. Use the constructed model Estimate the carbon storage of the sub-region.

[0016] As an optional scheme of the time-series remote sensing-based dynamic monitoring method for carbon storage in coastal wetlands described in this invention, the following steps are taken: for each sub-region within the region set, a preliminary judgment is made as to whether its parameter changes are normal, and a preliminary judgment conclusion is generated. Specifically, this involves extracting the current carbon storage assessment result of the sub-region. ; Obtain historical carbon storage data for the sub-region, denoted as ; Calculate the average value of historical carbon storage data, denoted as Calculate the standard deviation of historical carbon storage data, denoted as . Based on the statistical characteristics of historical data, determine the comparison threshold range, denoted as . ; through preliminary determination function A preliminary assessment is needed to determine whether the carbon storage in this sub-region is normal; if... If so, the parameters are initially determined to be abnormal; if If so, the parameters are preliminarily determined to be normal.

[0017] As an optional scheme of the time-series remote sensing-based dynamic monitoring method for carbon storage in coastal wetlands described in this invention, the method involves: querying each sub-region within the region set to determine if a similar region exists; specifically, extracting the image data of the current sub-region, denoted as... Compare the image of this sub-region with the image of each sub-region in the region set; for each sub-region in the region set, calculate its relationship with the current sub-region. The similarity is denoted as For the current sub-region Generate its set of contrasting subregions, denoted as For each sub-region in the region set, a similarity comparison function is used. Determine its relationship with the current sub-region. Do similar regions exist? If If, then it is determined that similar regions exist; if If the condition is met, then it is determined that no similar regions exist.

[0018] As an optional scheme of the time-series remote sensing-based dynamic monitoring method for carbon storage in coastal wetlands described in this invention, the following is included: performing comparative verification, specifically: acquiring historical carbon storage data for the current sub-region, denoted as... Extract image data of the current sub-region. and comparison sub-region set For the set of contrasting subregions For each comparison sub-region, calculate the similarity between its historical carbon storage data and the historical carbon storage data of the current sub-region, denoted as . ; through historical comparison function Determine whether the similarity between the comparison sub-region and the historical data of the current sub-region is high; if If the historical data similarity is low, then data collection and verification will be performed on the current sub-region; if... If the historical data similarity is high, then carbon storage assessment is performed on the current sub-region and the highly similar comparison sub-regions respectively, and the carbon storage assessment results are recorded as follows: and ; Calculate the similarity of carbon storage changes between the current sub-region and each similar comparison sub-region, denoted as . By changing the comparison function Determine whether the carbon storage changes in the comparison sub-region are highly similar to those in the current sub-region; if If the similarity of the changes is high, then the data error is small; therefore, the verification conclusion is determined. ;like If the similarity of the changes is low, then the data error is large; if the similarity is low, then data collection and verification will be performed on the current sub-region.

[0019] As an optional scheme of the dynamic monitoring method for carbon storage in coastal wetlands based on time-series remote sensing described in this invention, the following is an example: performing data acquisition and verification, specifically: acquiring image data of the current sub-region. and historical carbon storage data The study involved: establishing sampling points within the coastal wetland to collect vegetation and soil samples; performing blanching, drying, weighing, crushing, and carbon content analysis on the collected vegetation samples to calculate the vegetation carbon storage per unit area; analyzing the carbon content and density of soil samples to assess soil carbon storage; monitoring microbial residual carbon in the soil using the amino sugar biomarker method to assess the role of microorganisms in carbon sequestration; classifying vegetation types in the current sub-region and extracting their spatial distribution information; assessing vegetation growth characteristics using vegetation indices; retrieving vegetation cover using vegetation indices combined with methods such as pixel binarization to assess vegetation growth status and carbon storage potential; constructing a remote sensing estimation model for carbon storage based on vegetation indices and field sample data to estimate the carbon storage of coastal wetland vegetation; identifying the main carbon pools in the coastal wetland; assessing greenhouse gases produced by soil microbial metabolism and considering their impact on carbon storage; and combining the overall carbon storage situation of the current sub-region. and historical carbon storage data Assess changes in carbon storage. Output the overall carbon storage situation of the current sub-region. and changes in carbon reserves .

[0020] As an optional scheme of the dynamic monitoring method for coastal wetland carbon storage based on time series remote sensing described in this invention, the method involves: based on the verification conclusions obtained from comparative verification or data collection verification, a secondary judgment is made on whether the parameter changes in the sub-region are normal, and a secondary judgment conclusion is generated. Specifically, this involves: obtaining the comprehensive carbon storage situation of the current sub-region. Historical carbon storage data and changes in carbon reserves and initial comparison threshold ; Calculate the average value of historical carbon storage data Calculate the standard deviation of historical carbon storage data, denoted as . Based on the overall situation of carbon reserves and historical data, the comparison threshold is adjusted and recorded as follows: ; through a quadratic decision function A second check is performed to determine whether the parameter changes in the current sub-region are normal; if... If the secondary judgment parameters are normal; if If the secondary judgment parameters are abnormal, then the error will occur.

[0021] As an optional scheme of the dynamic monitoring method for carbon storage in coastal wetlands based on time-series remote sensing described in this invention, the following is included: performing an overall analysis of the target area, specifically: obtaining a set of regions. For each sub-region, combining the preliminary judgment conclusion and the secondary judgment conclusion, a comprehensive judgment conclusion is calculated and denoted as follows: ; Iterate through all sub-regions to arrive at the comprehensive judgment conclusion, count the number of abnormal sub-regions, and denot them as ; through the overall decision function To determine the overall carbon storage situation in the target area; if If the target area parameters are abnormal, then manual monitoring is performed; if... If the parameters of the target area are normal, then monitoring should continue.

[0022] The present invention has the following beneficial effects:

[0023] 1. This time-series remote sensing-based dynamic monitoring method for carbon storage in coastal wetlands inputs image data of the target area. Based on the vegetation cover in the images, the target area is divided into several sub-regions. During the division process, the same plant species are grouped into the same region as much as possible. If it is not possible to group all plants into the same region, the plant species is encouraged to occupy the majority of the region. The division of regions is also made reasonable. For each sub-region in the region set, remote sensing monitoring methods are used to classify and determine the vegetation type and distribution, extracting spatial distribution information. The classification results are then corrected using field survey data. The results are evaluated using vegetation indices (such as NDVI, EVI, etc.). The study estimates vegetation growth characteristics, uses vegetation indices combined with pixel binarization and other methods to invert vegetation cover, assesses vegetation growth status and carbon storage potential, and constructs a remote sensing estimation model for carbon storage based on vegetation indices and field sample data to estimate the carbon storage of coastal wetland vegetation. The target area is divided into multiple sub-regions to facilitate refined management and monitoring, ensuring that each sub-region has a consistent area and is adjacent to each other, thereby improving the rationality and operability of monitoring. Through remote sensing monitoring methods, combined with vegetation type classification and vegetation index assessment, the carbon storage of each sub-region is accurately assessed. Using vegetation indices and remote sensing estimation models, the study scientifically assesses vegetation cover and carbon storage potential.

[0024] 2. This time-series remote sensing-based dynamic monitoring method for carbon storage in coastal wetlands acquires historical data for each sub-region in the regional set. Based on the comparison threshold corresponding to each historical data point and the assessed carbon storage situation of the sub-region, a new comparison threshold is determined. The method then determines whether the assessed carbon storage situation of the sub-region falls within the comparison threshold range. If it falls outside the range, a preliminary judgment of parameter anomaly is made; if it falls within the range, a preliminary judgment of parameter normality is made. Each sub-region is validated, and the parameter changes are reassessed based on the validation results. For each validated sub-region, the comparison threshold is adjusted based on the overall carbon storage situation, historical carbon storage data, and the changes in carbon storage after validation. The method then determines whether the overall carbon storage situation of the sub-region falls within the adjusted comparison threshold range. If it falls outside the adjusted range, a second judgment of parameter anomaly is made; if it falls within the adjusted range, a second judgment is made. If the parameters are normal, and considering the two judgments for each sub-region, a comprehensive analysis of the target region is performed. For each sub-region, a comprehensive judgment is calculated based on the preliminary and secondary judgments. The number of sub-regions with abnormal comprehensive judgments is counted, and it is determined whether the number of sub-regions with abnormal comprehensive judgments exceeds a preset number. If it exceeds the preset number, the target region's parameters are determined to be abnormal, and manual monitoring is initiated. If it does not exceed the preset number, the target region's parameters are determined to be normal, and monitoring continues. By using historical data and comparison thresholds, a preliminary judgment is made on whether the carbon storage changes are normal, providing early warnings and quickly screening potentially abnormal sub-regions to improve monitoring efficiency. Based on the verification conclusions, the comparison thresholds are dynamically adjusted to adapt to changes in carbon storage. Combining historical data and verification conclusions, a secondary judgment is made on whether the parameter changes are normal to improve the accuracy of the judgment. By combining the preliminary and secondary judgment results for each sub-region, a comprehensive assessment of the target region's carbon storage situation is conducted, providing important decision-making basis for subsequent monitoring and management, and ensuring that the carbon storage monitoring and management of the target region is scientific and reasonable.

[0025] 3. This time-series remote sensing-based dynamic monitoring method for carbon storage in coastal wetlands involves querying the existence of similar regions for each sub-region in the region set. For each sub-region, the image is compared with that of every other sub-region in the set. Sub-regions with extremely high similarity are extracted and designated as comparison sub-regions. The method then determines whether comparison sub-regions exist. If they do, a comparison verification is performed; otherwise, a data collection verification is performed. For sub-regions with comparison sub-regions, the comparison sub-regions are extracted, and historical data for both the sub-region and the comparison sub-region are extracted. The similarity between the historical data of the two sub-regions is then determined. If the historical data are similar... If the similarity is low, data collection and verification will be performed on the sub-region. If the historical data similarity is high, the carbon storage changes in the two sub-regions will be inferred based on the assessed carbon storage conditions obtained from the remote sensing monitoring method. The assessed carbon storage changes in the two sub-regions will be compared. If the similarity is high, the data error is considered small. Based on the environmental changes in the sub-region during the period between the most recent historical data collection and the current data collection, combined with the historical data, inferred carbon storage changes, and assessed carbon storage conditions, the carbon storage conditions of the sub-region will be determined. If the similarity is low, the data error is considered large, and data collection and verification will be performed on the sub-region. To conduct sampling verification in the designated sub-regions, ground surveys and laboratory analyses were carried out. Sampling points were established within the coastal wetlands to collect vegetation and soil samples. The collected vegetation samples underwent blanching, drying, weighing, pulverizing, and carbon content analysis to calculate the vegetation carbon storage per unit area. Soil samples were analyzed for carbon content and density to assess soil carbon storage. Microbial residual carbon in the soil was monitored using the aminoglycoside biomarker method to assess the role of microorganisms in carbon sequestration. A stratified sampling method was used to collect soil samples from different depths to determine soil organic carbon content and assess soil carbon storage. Soil property analysis was used to evaluate soil carbon sequestration capacity and to assess how much of the carbon fixed by vegetation could be utilized. Buried and permanently stored, the main carbon pools of coastal wetlands, including underground biomass and soil organic carbon, are identified. A comprehensive assessment is conducted to evaluate greenhouse gases produced by soil microbial metabolism and consider their impact on carbon storage, forming a comprehensive carbon storage situation for this sub-region. Combined with historical carbon storage data for this sub-region, the changes in carbon storage in this sub-region are assessed. Image comparison is used to find similar areas, providing a basis for subsequent comparative verification, ensuring the reliability of comparative verification, and avoiding misjudgments due to regional differences. The combination of comparative verification and data collection verification improves the accuracy of carbon storage assessment. Verification methods can be flexibly selected based on the existence of similar areas to adapt to different situations. Attached Figure Description

[0026] Figure 1 This is a flowchart of the dynamic monitoring method for carbon storage in coastal wetlands based on time-series remote sensing, as described in this invention. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] Example 1: A method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing, see reference. Figure 1 ,include:

[0029] Obtain the set of regions within the target area;

[0030] Perform parametric analysis sequentially on each sub-region within the region set;

[0031] For each sub-region within the region set, a preliminary judgment is made as to whether the parameter changes are normal, and a preliminary judgment conclusion is generated;

[0032] For each sub-region within the region set, query whether it contains similar regions;

[0033] If similar regions exist, perform a comparison verification.

[0034] If no similar region exists, perform data collection and verification.

[0035] Based on the verification conclusions obtained from comparative verification or data collection verification, a secondary judgment is made on whether the parameter changes in the sub-region are normal, and a secondary judgment conclusion is generated.

[0036] The target region is analyzed as a whole by combining the preliminary and secondary judgment results of each sub-region.

[0037] Specifically, obtaining the set of regions for the target region involves:

[0038] Acquire an image of the target area, denoted as ;

[0039] Use image segmentation algorithms to process the image. Vegetation classification was performed to obtain a vegetation classification map. :

[0040] ;

[0041] in, This is an image segmentation function used to output a vegetation classification map by taking the input image and a vegetation classification threshold. This is the threshold for vegetation classification, used to distinguish different vegetation types;

[0042] According to the vegetation classification map The target area is divided into several sub-regions. :

[0043] ;

[0044] in, This is a function for region partitioning, which takes a vegetation classification map, sub-region area, and sub-region number as input, and outputs the sub-regions. This refers to the area of ​​each sub-region, i.e., the area of ​​each pre-defined sub-region. This is a sub-region index with values ​​of 1~ , The number of sub-regions into which the target region is divided;

[0045] Check each sub-region The distribution of vegetation types within the area is analyzed to ensure a reasonable division. If the distribution of vegetation types in a certain sub-region is unreasonable, the division is readjusted.

[0046] ;

[0047] in, This is a function for re-dividing subregions, which takes a vegetation classification map, subregion area, and subregion number as input and outputs new subregions. This is a validation function used to determine whether the vegetation type distribution of an input sub-region is reasonable. If reasonable, it outputs the result. If it is unreasonable, output The determination of whether the vegetation type distribution of a sub-region is reasonable includes: calculating the area proportion of the main vegetation type in the sub-region; if the proportion of a certain vegetation type exceeds a preset threshold, it is considered to meet the first reasonable requirement, otherwise it is considered not to meet the first reasonable requirement; checking whether the boundary of the sub-region coincides with the boundary of the vegetation type to avoid the same vegetation type being divided into multiple sub-regions; if the boundary of the sub-region coincides with the boundary of the vegetation type, it is considered to meet the second reasonable requirement, otherwise it is considered not to meet the second reasonable requirement; checking whether the sub-region contains a complete ecological unit, such as a complete intertidal profile from high tide to low tide; if it contains a complete ecological unit, it is considered to meet the third reasonable requirement, otherwise it is considered not to meet the third reasonable requirement; if it meets two or more reasonable requirements, the vegetation type distribution of the sub-region is considered reasonable, otherwise the vegetation type distribution of the sub-region is considered unreasonable.

[0048] By integrating all sub-regions, we obtain the final set of regions, denoted as . :

[0049] ;

[0050] in, The number of sub-regions into which the target region is divided. This represents the union, which integrates all sub-regions.

[0051] This embodiment also provides a holistic analysis of the target area, specifically:

[0052] Get region set Each sub-region contains a preliminary judgment conclusion. Secondary judgment conclusion and preset threshold for the number of abnormal sub-regions ;

[0053] For each sub-region, combining the preliminary judgment conclusion and the secondary judgment conclusion, a comprehensive judgment conclusion is calculated, denoted as... :

[0054] ;

[0055] in, It is a comprehensive judgment function, when the preliminary judgment conclusion is reached. and the conclusion of the second judgment If any one of the conclusions is deemed abnormal, then the overall conclusion is considered. The output is That is, the comprehensive judgment conclusion To determine if something is abnormal, otherwise, a comprehensive judgment conclusion will be made. The output is That is, the comprehensive judgment conclusion To be considered normal;

[0056] The comprehensive judgment results of all sub-regions are traversed, and the number of abnormal sub-regions is counted, denoted as . :

[0057] ;

[0058] in, Represents a set of regions A sub-region in , Indicates the statistical comprehensive judgment conclusion The total number of sub-regions to determine anomalies;

[0059] Based on the overall decision function To determine the overall carbon storage situation in the target area:

[0060] ;

[0061] in, The preset threshold for the number of abnormal sub-regions is used to determine whether the overall carbon storage of the target area is abnormal. If the number of abnormal sub-regions exceeds the threshold, the overall carbon storage of the target area is determined to be abnormal; otherwise, the overall carbon storage of the target area is determined to be normal.

[0062] like If so, the target area parameters are determined to be abnormal;

[0063] Then, manual monitoring will be conducted;

[0064] like If so, the target area parameters are determined to be normal;

[0065] Then, monitoring will continue.

[0066] The above methods comprehensively consider various approaches, including remote sensing monitoring, ground surveys, and laboratory analysis, to ensure the comprehensiveness and accuracy of carbon storage monitoring. Through preliminary and secondary assessments, combined with historical data and verification conclusions, the comparison threshold is dynamically adjusted to adapt to carbon storage changes in different sub-regions. From regional division to overall analysis, a systematic monitoring process has been formed, which can effectively identify anomalies within the target area. Through comparative verification and data collection verification, the dual verification mechanism improves the reliability of carbon storage assessment and provides reliable data support for subsequent monitoring and management.

[0067] Example 2 is an improvement upon Example 1. This method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing performs parameter analysis sequentially on each sub-region within the region set. Specifically:

[0068] Extract the image of the sub-region, denoted as ;

[0069] Image segmentation algorithm is used to segment sub-regions of the image. The vegetation types were classified to obtain a vegetation classification map, denoted as... :

[0070] ;

[0071] in, This is an image segmentation function used to output a vegetation classification map by taking the input image and a vegetation classification threshold. This is the threshold for vegetation classification, used to distinguish different vegetation types;

[0072] Using field survey data The classification results are corrected to obtain a corrected vegetation classification map, denoted as... :

[0073] ;

[0074] in, The data is from on-site surveys and is used to correct the classification results. This is a correction function used to correct classification results based on input vegetation classification maps and field survey data;

[0075] Calculate vegetation indices (such as NDVI and EVI) to assess vegetation growth characteristics:

[0076] ;

[0077] ;

[0078] ;

[0079] in, The Normalized Difference Vegetation Index (NDVI) is used to reflect vegetation greenness and is sensitive to high vegetation cover. This is an enhanced vegetation index used to correct for atmospheric and soil background effects by introducing the blue light band, and is suitable for high biomass areas. This is a parameter identifier used to specify the type of vegetation index to be calculated. Through this parameter, the function can flexibly select different vegetation index calculation formulas according to its needs. This is a vegetation index calculation function, used to output a vegetation index from an input image and calculation parameters, specifically:

[0080] ;

[0081] Vegetation cover can be inverted using methods such as vegetation index combined with pixel binarization. Assess the growth status and carbon storage potential of vegetation:

[0082] ;

[0083] in, This is a vegetation cover assessment function, used to take a vegetation index as input and output vegetation cover. The higher the vegetation cover, the denser the vegetation and the more vigorous the growth. Meanwhile, vegetation cover... It is positively correlated with carbon storage; the higher the coverage, the greater the carbon sequestration potential.

[0084] A remote sensing estimation model for carbon storage was constructed based on vegetation index and field sample data, denoted as [model name missing]. :

[0085] ;

[0086] in, These are the parameters for the carbon storage estimation model, used to construct the carbon storage estimation model. A function is constructed for the remote sensing estimation model, which takes vegetation index, field survey data, and model parameters as input and outputs a carbon storage estimation model. The specific steps for constructing the carbon storage estimation model are as follows:

[0087] S1. Collect remote sensing images and corresponding ground-measured carbon storage data ;

[0088] S2. Feature extraction: Extract vegetation indices (NDVI, EVI), texture features, topographic factors, etc. from images;

[0089] S3. Sample selection: Select training and validation samples based on the principle of representativeness.

[0090] S4. Model selection: Choose machine learning algorithms such as random forest, support vector machine, or neural network.

[0091] S5. Model training: Use training samples to fit the relationship between carbon storage and remote sensing features.

[0092] S6. Model Validation: Use validation samples to evaluate model accuracy and ensure... , ;

[0093] S7. Model optimization: Improve model performance through parameter tuning and feature selection;

[0094] Use the constructed model Estimate the carbon storage of the sub-region:

[0095] ;

[0096] in, This is a carbon storage estimation function used to output carbon storage assessment results by taking a carbon storage estimation model and vegetation cover as input.

[0097] This embodiment also provides a preliminary determination of whether the parameter changes of each sub-region within the region set are normal, generating a preliminary determination conclusion, specifically:

[0098] Extract current carbon storage assessment results for sub-regions ;

[0099] Obtain historical carbon storage data for the sub-region, denoted as The historical carbon storage data includes carbon storage data at multiple time points;

[0100] Calculate the average value of historical carbon storage data, denoted as :

[0101] ;

[0102] in, For the first Historical carbon storage data for each sub-region This refers to the sample size of historical data, i.e., the total number of past monitoring time points;

[0103] The standard deviation of historical carbon storage data is calculated and denoted as . :

[0104] ;

[0105] Based on the statistical characteristics of historical data, a comparison threshold range is determined, denoted as . :

[0106] ;

[0107] in, To compare the lowest value within the threshold range, the specific formula is as follows:

[0108] ;

[0109] In the comparison threshold range, To compare the highest value within the threshold range, the specific formula is as follows:

[0110] ;

[0111] in, This is a multiple of the standard deviation, used to determine the width of the confidence interval, i.e., to control the stringency of the comparison threshold. The larger the value, the wider the threshold range and the more lenient the anomaly detection. For example, a value of 1.96 corresponds to a 95% confidence interval.

[0112] Based on the preliminary judgment function A preliminary assessment is needed to determine whether the carbon storage in this sub-region is normal.

[0113] ;

[0114] in, This indicates that it belongs to, i.e., the current carbon storage. Within the comparison threshold range Inside, at this point, the function is initially determined. The output is , This indicates that it does not belong to the category of current carbon storage. Not within the comparison threshold range Inside, at this point, the function is initially determined. The output is ;

[0115] like If so, it can be preliminarily determined that the parameters are abnormal;

[0116] like If so, the parameters are preliminarily determined to be normal.

[0117] Example 3 is an improvement on Example 2. In this example, each sub-region within the region set is queried to determine if there are similar regions. Specifically:

[0118] Extract the image data of the current sub-region, denoted as ;

[0119] Compare this sub-region with the image of each sub-region within the region set;

[0120] For each subregion in the region set, calculate its relationship with the current subregion. The similarity is denoted as :

[0121] ;

[0122] in, For the first region in the set Image data of each sub-region The similarity calculation function is as follows:

[0123] ;

[0124] in, For image The pixel mean reflects the average brightness of the image. For image The pixel mean reflects the average brightness of the image. For image and The covariance reflects the structural correlation between two images. and To maintain stability and prevent the denominator from being zero, such as for , for , The similarity is quantified by comparing the brightness, contrast, and structural information of two images, taking a range of pixel values. The closer they are to 1, the more similar they are.

[0125] For the current sub-region Generate its set of contrasting subregions, denoted as :

[0126] ;

[0127] in, This is a similarity threshold used to determine whether two sub-regions are similar. If a sub-region in a region set has a similarity threshold... Greater than or equal to the similarity threshold Then add the sub-region to the comparison sub-region set. middle;

[0128] For each sub-region in the region set, a similarity comparison function is used. Determine its relationship with the current sub-region. Are there similar regions?

[0129] ;

[0130] in, To represent the empty set, Represents the set of contrasting subregions An empty value indicates that there is no subregion that is compatible with the current subregion. Similar subregions, at this time the similarity comparison function The output is , Represents the set of contrasting subregions Not empty means it exists in the current sub-region. Similar subregions, at this time the similarity comparison function The output is ;

[0131] like If so, then it is determined that similar regions exist;

[0132] like If the condition is met, then it is determined that no similar regions exist.

[0133] This embodiment also provides a method for performing comparison verification, specifically:

[0134] Obtain the historical carbon storage data for the current sub-region, denoted as ;

[0135] Extract image data of the current sub-region and comparison sub-region set ;

[0136] Among them, the comparison sub-region set Image data containing contrast sub-regions and historical carbon storage data Its formula is ,in, The number of sub-regions to compare, i.e. the total number of other sub-regions similar to the current sub-region;

[0137] For the set of contrasting subregions For each comparison sub-region, calculate the similarity between its historical carbon storage data and the historical carbon storage data of the current sub-region, denoted as . :

[0138] ;

[0139] in, To compare historical carbon storage data for sub-regions, i.e. , The similarity calculation function is as follows:

[0140] ;

[0141] in, This represents the historical average carbon storage of the current sub-region, reflecting the long-term carbon sequestration level. To compare the historical average carbon storage of sub-regions and reflect long-term carbon sequestration levels, The covariance of the carbon storage time series of the current sub-region and the comparison sub-region reflects the synchronicity of their interannual variation trends. This represents the temporal variance of carbon reserves in the current sub-region, reflecting the stability of interannual fluctuations in carbon reserves. To compare the temporal variance of carbon reserves in sub-regions and reflect the stability of interannual fluctuations in carbon reserves, and To ensure stability and prevent the denominator from being zero, this constant is set based on the range of carbon storage data. It is used in stable regions where the variation in carbon storage is very small to prevent instability in SSIM calculations.

[0142] By comparing historical data Determine whether the historical data of the comparison sub-region is highly similar to that of the current sub-region:

[0143] ;

[0144] in, This indicates the existence of historical data similarity for any comparison sub-region. Greater than or equal to the similarity threshold , This indicates that there is no historical data similarity among the comparison sub-regions. Greater than or equal to the similarity threshold ;

[0145] like If so, the historical data similarity is low;

[0146] Then, perform data collection and verification on the current sub-region;

[0147] like If so, it is determined that the historical data has a high degree of similarity;

[0148] Then, carbon storage assessments are performed on the current sub-region and the comparable sub-regions with high similarity, respectively, to obtain carbon storage assessment results, which are denoted as follows: and :

[0149] ;

[0150] ;

[0151] in, This is a carbon storage assessment function used to evaluate the carbon storage status of the current sub-region and highly similar comparison sub-regions. This function integrates multi-source data from remote sensing monitoring, ground surveys, and laboratory analysis to achieve a systematic estimation of vegetation carbon pools, soil carbon pools, and microbial carbon pools. Based on a three-layer architecture principle of "remote sensing inversion + ground verification + laboratory analysis," at the remote sensing level, it uses vegetation indices and spectral characteristics for large-scale and rapid estimation. At the ground level, it acquires measured data through sample point deployment for model training and result correction. At the laboratory level, it accurately measures the carbon content of vegetation and soil, providing ground truth values. The final output is the net carbon storage after greenhouse gas emission correction, ensuring that the assessment results reflect the true carbon sink capacity. For the current sub-region, the latest remote sensing imagery and the ground samples collected this time are used for assessment. Greenhouse gas emission correction uses the most recent measured data, and the results are used for current status diagnosis and anomaly detection. For similar sub-regions, remote sensing imagery from the same time period as the current sub-region and historical sample data are used to ensure temporal consistency and comparability. The results are used for historical trend comparison and similarity calculation.

[0152] Calculate the similarity of carbon storage changes between the current sub-region and each similar comparison sub-region, denoted as . :

[0153] ;

[0154] in, The similarity calculation function is as follows:

[0155] ;

[0156] in, This represents the average carbon storage in the current sub-region, reflecting the overall carbon sequestration level. To compare the average carbon storage of sub-regions and reflect the overall carbon sequestration level, The covariance of the current sub-region and the comparison sub-region reflects the spatial covariance of carbon reserves, indicating the strength of their spatial distribution correlation. This represents the spatial variance of carbon reserves in the current sub-region, reflecting the degree of heterogeneity in carbon reserve distribution. To compare the spatial variance of carbon reserves in sub-regions and reflect the degree of heterogeneity in carbon reserve distribution, and To ensure stability and prevent the denominator from being zero, the value is set based on the range of carbon storage data, guaranteeing calculation stability in a uniform region where the carbon storage difference is very small.

[0157] By changing the comparison function To determine whether the carbon storage changes in the comparison sub-region are highly similar to those in the current sub-region:

[0158] ;

[0159] in, This is a similarity threshold for carbon storage changes, used to determine whether the carbon storage changes of the comparison sub-region and the current sub-region are highly similar. This indicates the similarity of carbon storage changes in any comparison sub-region. Greater than or equal to the similarity threshold of carbon storage changes , This indicates that there is no similarity in carbon storage changes among any of the comparison sub-regions. Greater than or equal to the similarity threshold of carbon storage changes ;

[0160] like If the similarity of the changes is high, then the changes are judged to be of high similarity.

[0161] If the data error is small, then the verification conclusion is determined. :

[0162] Let the set of environmental factors be... Environmental factors include temperature, precipitation, and sea level.

[0163] ;

[0164] For each environmental factor , ;

[0165] in, These are the environmental factor values ​​for the current time period. is the mean of environmental factors over a historical period, and is the standard deviation of environmental factors over a historical period;

[0166] Calculate the comprehensive change index of multiple environmental factors, denoted as . :

[0167] ;

[0168] The constraints are as follows: That is, the sum of the weights of each factor is 1;

[0169] Let the historical carbon storage time series be ;

[0170] Calculate the slope of the linear trend:

[0171] ;

[0172] in, For historical data timestamps, This is the historical average. This represents the historical average carbon storage.

[0173] The underlying trend for calculation and prediction: ;

[0174] in, For time intervals, ;

[0175] Computing environment modifications: ;

[0176] in, The environmental sensitivity coefficient is used to control the intensity and scope of the impact of environmental changes on carbon storage prediction results, ensuring the rationality and stability of the prediction.

[0177] Calculate the final predicted value of current carbon storage: ;

[0178] Calculate the predicted value and the remote sensing assessment value Deviation: ;

[0179] like ,but This means the data error is small, indicating the current carbon storage situation in the sub-region. ,in, The preset error tolerance, This is a threshold for environmental changes, used to determine whether the data error is large;

[0180] like ,but If the data error is large, then data collection and verification will be performed.

[0181] like If the similarity of the changes is low, then the change is judged to be low.

[0182] If the data error is large, then data collection and verification will be performed on the current sub-region.

[0183] Specifically, the data collection and verification process includes:

[0184] Get the image data of the current sub-region and historical carbon storage data ;

[0185] Sampling points were established within the coastal wetlands to collect vegetation and soil samples. The sampling points should evenly cover the work area, including vegetation distribution areas at high tide, mid tide, and low tide levels.

[0186] ;

[0187] in, This is the collected sample set, including vegetation samples and soil samples. This is a sample acquisition function used to collect data from sub-region images. Under the guidance of [unspecified entity], sample points were set up and samples were collected on-site.

[0188] The collected vegetation samples were subjected to blanching, drying, weighing, crushing, and carbon content analysis to calculate the vegetation carbon storage per unit area.

[0189] ;

[0190] in, For vegetation carbon storage, This is a function for analyzing vegetation samples, used to perform laboratory analysis on vegetation samples and calculate carbon storage.

[0191] Soil samples were analyzed for carbon content and carbon density to assess soil carbon storage.

[0192] ;

[0193] in, For soil carbon storage, This is a function for analyzing soil samples, used to perform carbon content and density analysis on soil samples;

[0194] The role of microorganisms in carbon sequestration was assessed by monitoring microbial residual carbon in soil using an aminoglycoside biomarker method.

[0195] ;

[0196] in, For microbial residual carbon storage, This is a microbial analysis function used to analyze residual carbon in microorganisms using the amino sugar biomarker method;

[0197] Classify the vegetation type in the current sub-region and extract its spatial distribution information:

[0198] ;

[0199] in, The vegetation index value is a dimensionless value. This is a vegetation index calculation function used to calculate indices such as NDVI and EVI based on remote sensing images;

[0200] Assess vegetation growth characteristics using vegetation indices (such as NDVI, EVI, etc.):

[0201] ;

[0202] in, This represents vegetation cover, with values ​​between 0 and 1. This is a vegetation cover calculation function used to invert vegetation cover based on vegetation indices.

[0203] Vegetation cover can be retrieved using vegetation indices combined with methods such as pixel binarization to assess vegetation growth status and carbon storage potential.

[0204] ;

[0205] in, This refers to carbon storage estimation models, i.e., machine learning model objects. A function is built for the model to train the carbon storage estimation model based on sample data;

[0206] A remote sensing model for carbon storage estimation was constructed based on vegetation index and field sample data. The carbon storage of coastal wetland vegetation was estimated using the model.

[0207] ;

[0208] in, For remote sensing-estimated vegetation carbon storage, This is a carbon storage estimation function used to apply the model to remote sensing imagery to estimate carbon storage.

[0209] Identify the main carbon pools in coastal wetlands, including underground biomass and soil organic carbon, while ignoring aboveground biomass and litter, as their carbon storage is significantly affected by tides.

[0210] ;

[0211] in, Total carbon reserves are the sum of all carbon pools.

[0212] Assess greenhouse gases (such as CO2, CH4, N2O) produced by soil microbial metabolism and consider their impact on carbon storage:

[0213] ;

[0214] ;

[0215] in, This study quantifies the amount of carbon lost from soil carbon pools due to microbial decomposition caused by greenhouse gas emissions, reflecting the negative impact of greenhouse gas emissions on carbon storage. The magnitude of this value directly reflects the degree of carbon storage reduction caused by greenhouse gas emissions. The larger the size, the smaller the net carbon storage and the weaker the wetland's carbon sequestration function. The smaller the value, the larger the net carbon storage and the stronger the carbon sequestration function. It serves as a crucial bridge connecting ecosystem processes and carbon storage accounting, reflecting the true carbon sequestration capacity of coastal wetlands as "blue carbon" ecosystems. This is a greenhouse gas assessment function used to quantify the greenhouse gas fluxes produced by soil microbial metabolism. Received Net carbon storage is the amount lost due to greenhouse gas emissions after deducting the portion of total carbon storage from total carbon storage.

[0216] Based on the current overall situation of carbon storage in the sub-region and historical carbon storage data Assess changes in carbon storage. :

[0217] ;

[0218] Output the overall carbon storage situation of the current sub-region and changes in carbon reserves .

[0219] This embodiment also provides that, based on the verification conclusion obtained from comparative verification or data collection verification, a secondary judgment is made as to whether the parameter changes in the sub-region are normal, and a secondary judgment conclusion is generated, specifically as follows:

[0220] Obtain the overall carbon storage situation of the current sub-region Historical carbon storage data and changes in carbon reserves and initial comparison threshold Initial comparison threshold This is the threshold for initially determining whether carbon reserves are abnormal;

[0221] Calculate the average value of historical carbon storage data :

[0222] ;

[0223] in, For the first Historical carbon storage data at various points in time, The number of historical data points;

[0224] The standard deviation of historical carbon storage data is calculated and denoted as . :

[0225] ;

[0226] Based on the overall situation of carbon storage and historical data, the comparison threshold was adjusted and denoted as... :

[0227] ;

[0228] in, This is a multiple of the standard deviation, used to determine the width of the confidence interval and control the stringency of the comparison threshold. The larger the value, the wider the threshold range, and the more lenient the anomaly detection. The value is 1.96, corresponding to a 95% confidence interval. This is an adjustment factor used to account for the impact of changes in carbon reserves. The weight factor value is between 0 and 1, representing the degree of inheritance from the initial threshold. This is a statistical adjustment item based on historical data. This is a dynamic adjustment term based on changes in carbon storage;

[0229] Through the quadratic decision function A second check is performed to determine whether the parameter changes in the current sub-region are normal.

[0230] ;

[0231] in, This is the threshold range width coefficient, used to define the width of the threshold range and the tolerance range for secondary judgments. The larger the value, the more lenient the judgment criteria;

[0232] like If so, the secondary judgment parameters are normal;

[0233] like If the secondary judgment parameters are abnormal, then the error will occur.

[0234] This embodiment comprehensively considers multiple methods, including remote sensing monitoring, ground surveys, and laboratory analysis, to ensure the comprehensiveness and accuracy of carbon storage monitoring. Through preliminary and secondary judgments, combined with historical data and verification conclusions, the comparison threshold is dynamically adjusted to adapt to the carbon storage changes in different sub-regions. From regional division to overall analysis, a systematic monitoring process is formed, which can effectively identify anomalies in the target area. Through comparative verification and data collection verification, the dual verification mechanism improves the reliability of carbon storage assessment and provides reliable data support for subsequent monitoring and management.

Claims

1. A method for monitoring the dynamic of carbon storage in coastal wetlands based on time series remote sensing, characterized in that: include: Obtain a set of regions for the target area, wherein the set of regions is a collection of several sub-regions obtained by dividing the area according to the vegetation classification map; Parametric analysis is performed sequentially on each subregion within the region set to estimate the carbon storage of the subregion; For each sub-region within the region set, a preliminary assessment is made as to whether the parameter changes are normal, and a preliminary judgment conclusion is generated to determine whether the carbon storage of the sub-region is normal. For each sub-region within the region set, query whether there is a similar region. Extract the image data of the current sub-region. For each sub-region in the region set, calculate its similarity with the current sub-region. The similarity is quantified by comparing the brightness, contrast, and structural information of the two images. If the similarity of a sub-region in the region set is greater than or equal to the similarity threshold, then add the sub-region to the comparison sub-region set. If the comparison sub-region set is not empty, it is determined that there is a similar region. If the comparison sub-region set is empty, it is determined that there is no similar region. If similar regions exist, perform a comparison verification. If no similar region exists, perform data collection and verification. Based on the verification conclusions obtained from comparative verification or data collection verification, the comparison threshold is adjusted, and a second judgment is made on whether the parameter changes in the sub-region are normal, generating a second judgment conclusion. By combining the preliminary and secondary judgment results of each sub-region, an overall analysis of the target region is conducted to determine the overall carbon storage situation of the target region.

2. The method according to claim 1, wherein the method is characterized in that: To obtain the set of regions for the target region, specifically: collecting an image of a target region, denoted as ; using an image segmentation algorithm on the image performing vegetation classification to obtain a vegetation classification map ; According to the vegetation classification map The target region is divided into several sub-regions ; Check each sub-region Distribution of vegetation types within the area; By integrating all sub-regions, we obtain the final set of regions, denoted as . .

3. The method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing according to claim 1, characterized in that: Parametric analysis is performed sequentially on each sub-region within the region set, specifically as follows: Extract the image of the sub-region, denoted as ; Image segmentation algorithm is used to segment sub-regions of the image. The vegetation types were classified to obtain a vegetation classification map, denoted as... ; Using field survey data The classification results are corrected to obtain a corrected vegetation classification map, denoted as... ; Calculate vegetation indices to assess vegetation growth characteristics; Inverting vegetation cover To assess the growth status and carbon storage potential of vegetation; A remote sensing estimation model for carbon storage was constructed based on vegetation index and field sample data, denoted as [model name missing]. ; Use the constructed model Estimate the carbon storage of the sub-region.

4. The method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing according to claim 1, characterized in that: For each sub-region within the region set, a preliminary assessment is made as to whether the parameter changes are normal, generating a preliminary judgment conclusion, specifically: Extract current carbon storage assessment results for sub-regions ; Obtain historical carbon storage data for the sub-region, denoted as ; Calculate the average value of historical carbon storage data, denoted as ; The standard deviation of historical carbon storage data is calculated and denoted as . ; Based on the statistical characteristics of historical data, a comparison threshold range is determined, denoted as . ; Based on the preliminary judgment function A preliminary assessment was made to determine whether the carbon storage in this sub-region was normal. like If so, it can be preliminarily determined that the parameters are abnormal; like If so, the parameters are preliminarily determined to be normal.

5. The method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing according to claim 1, characterized in that: Specifically, for each sub-region within the region set, query whether there are similar regions. Extract the image data of the current sub-region, denoted as ; Compare this sub-region with the image of each sub-region within the region set; For each subregion in the region set, calculate its relationship with the current subregion. The similarity is denoted as ; For the current sub-region Generate its set of contrasting subregions, denoted as ; For each sub-region in the region set, a similarity comparison function is used. Determine its relationship with the current sub-region. Are there similar regions? like If so, then it is determined that similar regions exist; like If the condition is met, then it is determined that no similar regions exist.

6. The method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing according to claim 1, characterized in that: Perform the comparison verification as follows: Obtain the historical carbon storage data for the current sub-region, denoted as ; Extract image data of the current sub-region and comparison sub-region set ; For the set of contrasting subregions For each comparison sub-region, calculate the similarity between its historical carbon storage data and the historical carbon storage data of the current sub-region, denoted as . ; By comparing historical data To determine whether the historical data of the comparison sub-region is highly similar to that of the current sub-region; like If so, the historical data similarity is low; Then, perform data collection and verification on the current sub-region; like If so, it is determined that the historical data has a high degree of similarity; Then, carbon storage assessments are performed on the current sub-region and the comparable sub-regions with high similarity, respectively, to obtain carbon storage assessment results, which are denoted as follows: and ; Calculate the similarity of carbon storage changes between the current sub-region and each similar comparison sub-region, denoted as . ; By changing the comparison function To determine whether the carbon storage changes of the comparison sub-region are highly similar to those of the current sub-region; like If the similarity of the changes is high, then the changes are judged to be of high similarity. If the data error is small, then the verification conclusion is determined. ; like If the similarity of the changes is low, then the change is judged to be low. If the data error is large, then data collection and verification will be performed on the current sub-region.

7. The method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing according to claim 6, characterized in that: Perform data collection and verification, specifically as follows: Get the image data of the current sub-region and historical carbon storage data ; Sampling points were set up in the coastal wetlands to collect vegetation and soil samples; The collected vegetation samples were subjected to blanching, drying, weighing, crushing and carbon content analysis to calculate the vegetation carbon storage per unit area. Soil samples were analyzed for carbon content and carbon density to assess soil carbon storage. The role of microorganisms in carbon sequestration was assessed by monitoring microbial residual carbon in soil using amino sugar biomarkers. Classify the vegetation type of the current sub-region and extract its spatial distribution information; Vegetation growth characteristics are assessed using vegetation indices. Vegetation cover was retrieved by combining vegetation indices with methods such as pixel binarization, and the growth status and carbon storage potential of vegetation were assessed. A remote sensing estimation model for carbon storage was constructed based on vegetation index and field sample data, and the carbon storage of coastal wetland vegetation was estimated through the model. Identify the main carbon pools in coastal wetlands; Assess greenhouse gases produced by soil microbial metabolism and consider their impact on carbon storage; Based on the current overall situation of carbon storage in the sub-region and historical carbon storage data Assess changes in carbon storage. ; Output the overall carbon storage situation of the current sub-region and changes in carbon reserves .

8. The method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing according to claim 1, characterized in that: Based on the verification conclusions obtained from comparative verification or data collection verification, a secondary judgment is made as to whether the parameter changes in the sub-region are normal, and a secondary judgment conclusion is generated, specifically as follows: Obtain the overall carbon storage situation of the current sub-region Historical carbon storage data and changes in carbon reserves and initial comparison threshold ; Calculate the average value of historical carbon storage data ; The standard deviation of historical carbon storage data is calculated and denoted as . ; Based on the overall situation of carbon storage and historical data, the comparison threshold was adjusted and denoted as... ; Through the quadratic decision function A second check is performed to determine whether the parameter changes in the current sub-region are normal. like If so, the secondary judgment parameters are normal; like If the secondary judgment parameters are abnormal, then the error will occur.

9. The method for dynamic monitoring of carbon storage in coastal wetlands based on time-series remote sensing according to claim 1, characterized in that: A comprehensive analysis of the target area is conducted, specifically as follows: Get region set ; For each sub-region, combining the preliminary judgment conclusion and the secondary judgment conclusion, a comprehensive judgment conclusion is calculated, denoted as... ; The comprehensive judgment results of all sub-regions are traversed, and the number of abnormal sub-regions is counted, denoted as . ; Based on the overall decision function To determine the overall carbon storage situation in the target area; like If so, the target area parameters are determined to be abnormal; Then, manual monitoring will be conducted; like If so, the target area parameters are determined to be normal; Then, monitoring will continue.

Images