CN-121981869-A - Ecological system accounting system construction method and system based on multi-source data fusion

CN121981869ACN 121981869 ACN121981869 ACN 121981869ACN-121981869-A

Abstract

The invention discloses an ecological system accounting system construction method and system based on multi-source data fusion, and relates to the technical field of ecological environment monitoring, wherein the method comprises the steps of collecting and fusing multi-source ecological data of a target area to generate space-time aligned multi-source ecological data; the method comprises the steps of acquiring reference values of carbon fixation, air purification, soil maintenance and water purification of space-time grid units in a history reference period based on historical data in multi-source ecological data to form a history reference value data set, constructing a history ecological system service relation network according to ecological process space association rules, analyzing the structure of the history ecological system service relation network, identifying service nodes with key effects on network structure and functions to form a key service node set, and constructing a key node monitoring network based on the key service node set. The invention realizes the accounting of multiple services such as carbon fixation, air purification, soil preservation, water purification and the like in the whole area.

Inventors

SONG YOUTAO
LI JIAAO
ZHANG PENG
LIU XINGJING
WANG LI
XIAO FEIFEI
Zhao Zhaokai

Assignees

天津科技大学
天津市滨海新区生态环境局

Dates

Publication Date: 20260505
Application Date: 20260206

Claims (10)

1. The construction method of the ecological system accounting system based on the multi-source data fusion is characterized by comprising the steps of collecting and fusing multi-source ecological data of a target area to generate space-time aligned multi-source ecological data; Acquiring reference values of carbon fixation, air purification, soil maintenance and water purification of space-time grid units in a historical reference period based on historical data in the multi-source ecological data, and forming a historical reference value data set; according to the space association rule of the ecological process, a historical ecological system service relationship network is constructed, the structure of the historical ecological system service relationship network is analyzed, service nodes with key effects on the network structure and functions are identified, and a key service node set is formed; Constructing a key node monitoring network based on the key service node set, and training a global service deduction model based on the historical reference value data set and the historical ecosystem service relation network; and acquiring a real-time state vector of the key node through the key node monitoring network, inputting the real-time state vector of the key node into a global service deduction model, and deducting and outputting a global dynamic accounting result.
2. The method for constructing an ecosystem accounting system based on multi-source data fusion of claim 1, wherein the steps of collecting and fusing multi-source ecological data of a target area to generate space-time aligned multi-source ecological data comprise the following steps: Acquiring space-sky remote sensing image data, ground internet of things sensor monitoring data, basic geographic data and socioeconomic activity data, and performing time reference unification and space coordinate system first registration; And resampling the space-day remote sensing image data, the ground internet of things sensor monitoring data, the basic geographic data and the socioeconomic activity data which are subjected to the registration of the time base unification and the space coordinate system unification to a preset space-time grid coordinate system to generate space-time aligned multisource ecological data.
3. The method for constructing an ecosystem accounting system based on multi-source data fusion according to claim 2, wherein based on the historical data in the multi-source ecological data, reference values of carbon fixation, air purification, soil maintenance and water purification of space-time grid units in a historical reference period are obtained to form a historical reference value data set, and the method comprises the following steps: Aiming at the ecological process of vegetation photosynthesis and carbon circulation, integrating the space-day remote sensing image data and the ground internet of things sensor monitoring data in the historical reference period to generate a carbon fixation reference value of each space-time grid unit; aiming at the ecological process of the sedimentation of the atmospheric pollutants, integrating the space-day remote sensing image data and the ground internet of things sensor monitoring data in the historical reference period to generate an air purification reference value of each space-time grid unit; Aiming at the ecological process of water and soil loss control, integrating the space-sky remote sensing image data, the ground internet of things sensor monitoring data and the basic geographic data in the historical reference period to generate a soil retention reference value of each space-time grid unit; Aiming at the ecological process of purifying the water pollutants, integrating the space-sky remote sensing image data, the ground internet of things sensor monitoring data, the basic geographic data and the socioeconomic activity data in the historical reference period to generate a water quality purifying reference value of each space-time grid unit; And collecting the carbon fixation reference value, the air purification reference value, the soil maintenance reference value and the water quality purification reference value of the space-time grid unit to form a historical reference value data set of four ecological system service reference values.
4. The method for constructing an ecosystem accounting system based on multi-source data fusion of claim 3, wherein constructing a historical ecosystem service relationship network according to the ecological process space association rule comprises the following steps: Based on the hydrologic process principle, establishing downstream directed connection representing water purification and soil conservation service space transfer among the space-time grid units according to the water flow motion rule; Based on the atmospheric diffusion process principle, establishing downwind directional connection representing air purification service space transmission among space-time grid cells according to the dominant wind direction determined by long-term meteorological data; Based on the ecological proximity effect principle, establishing bidirectional connection representing the mutual influence of ecological processes between adjacent space-time grid cells according to the adjacent relation of the space-time grid cells; The method comprises the steps of taking a carbon fixation reference value, an air purification reference value, a soil maintenance reference value and a water quality purification reference value of each space-time grid unit in a historical reference value data set as node attributes, taking downstream directed connection established based on a hydrologic process principle, downwind directed connection established based on an atmospheric diffusion process principle and bidirectional connection established based on an ecological proximity effect principle as edges, and constructing a historical ecological system service relation network.
5. The method for constructing an ecosystem accounting system based on multi-source data fusion of claim 4, wherein analyzing the structure of a historical ecosystem service relationship network, identifying service nodes having a key effect on the network structure and function, and forming a key service node set comprises the following steps: Analyzing the pivot attribute of the nodes in the global circulation path of the network based on the spatial association topology contained in the service relationship network of the historical ecosystem, quantifying the frequency and the weight of each node as a channel for transmitting different service flows, and generating a medium centrality value; In a historical ecosystem service relation network, evaluating derived importance of each node obtained by establishing association with other important nodes in the network according to a connection influence transfer principle of the node, and further obtaining a feature vector centrality value reflecting a core position in a network influence pattern; Performing k kernel decomposition on the historical ecosystem service relation network, and identifying core structures of different layers in the network through recursively removing nodes with the degree value smaller than k to obtain a k kernel value; Combining the median centrality value, the feature vector centrality value and the k kernel value, and obtaining the comprehensive importance score of each node through weighted summation; and sequencing all the nodes according to the sequence from high to low of the comprehensive importance score to form a key service node set.
6. The method for constructing an ecosystem accounting system based on multi-source data fusion of claim 5, wherein the constructing a key node monitoring network based on the key service node set comprises the following steps: based on the key service node set, a flux monitoring station, an air quality monitoring station, a soil moisture sensor and a water quality multi-parameter analyzer are configured on the geographic position marked by the key service node set to form a key node monitoring network.
7. The method for constructing an ecosystem accounting system based on multi-source data fusion of claim 6, wherein training a global service deduction model based on a historical benchmark dataset and a historical ecosystem service relationship network comprises the steps of: Constructing a training sample set based on the historical reference value data set and the historical ecosystem service relation network, wherein the training sample set comprises historical state characteristics of a key service node set and corresponding carbon fixation reference values, air purification reference values, soil maintenance reference values and water quality purification reference values; Initializing a global service deduction model by using a graph neural network architecture, wherein the input dimension of the graph neural network architecture is matched with the historical state characteristic dimension of the key service node set, and the output layer dimension is matched with the vector dimension formed by the carbon fixation reference value, the air purification reference value, the soil maintenance reference value and the water quality purification reference value; Inputting the training sample set into a global service deduction model, minimizing the prediction output of the global service deduction model through a back propagation algorithm, and updating the parameters of the global service deduction model according to the difference among the real carbon fixation reference value, the air purification reference value, the soil maintenance reference value and the water quality purification reference value; and iteratively executing forward propagation and backward propagation updating, wherein the prediction error of the global service deduction model is lower than a preset threshold value, and obtaining the trained global service deduction model.
8. The method for constructing an ecosystem accounting system based on multi-source data fusion of claim 7, wherein the method for acquiring the real-time state vector of the key node through the key node monitoring network comprises the following steps: at the triggering moment of the accounting period, acquiring real-time monitoring data of a key service node set through a key node monitoring network to obtain real-time monitoring original data of the key service node set; Performing outlier detection and elimination processing on the real-time monitoring original data of the key service node set to obtain cleaned real-time monitoring data of the key service node set; and arranging and combining the cleaned real-time monitoring data of the key service node set according to a predefined sequence and format to form a key node real-time state vector.
9. The method for constructing an ecosystem accounting system based on multi-source data fusion of claim 8, wherein the step of inputting the real-time state vector of the key node into the global service deduction model to deduct and output the global dynamic accounting result comprises the following steps: injecting the real-time state vector of the key node into the trained global service deduction model, and activating a coupling deduction process based on graph structure information transmission; The global service deduction model performs space collaborative calculation on the real-time state vector of the key node according to the internally learned nonlinear mapping relation from the key node to the global service to generate the coupling characteristic expression of the global service state; The coupling characteristic expression is subjected to decoupling and reconstruction at the output end of the global service deduction model, and the carbon fixation quantity dynamic value, the air purification quantity dynamic value, the soil holding quantity dynamic value and the water quality purification quantity dynamic value of each space-time grid unit are analyzed; and combining the carbon fixation amount dynamic value, the air purification amount dynamic value, the soil holding amount dynamic value and the water quality purification amount dynamic value of the space-time grid unit, and assembling to form a global dynamic accounting result.
10. The ecological system accounting system construction system based on the multi-source data fusion is characterized by comprising a processing module, a processing module and a processing module, wherein the processing module is used for acquiring and fusing multi-source ecological data of a target area to generate time-space aligned multi-source ecological data; the reference value module is used for acquiring reference values of carbon fixation, air purification, soil maintenance and water purification of the space-time grid unit in a history reference period based on the history data in the multi-source ecological data to form a history reference value data set; The identification module is used for constructing a historical ecological system service relationship network according to the ecological process space association rule, analyzing the structure of the historical ecological system service relationship network, identifying service nodes with key effects on the network structure and functions, and forming a key service node set; The construction module is used for constructing a key node monitoring network based on the key service node set and training a global service deduction model based on the historical benchmark value data set and the historical ecosystem service relation network; the deduction module is used for acquiring the real-time state vector of the key node through the key node monitoring network, inputting the real-time state vector of the key node into the global service deduction model, and deducting and outputting a global dynamic accounting result.

Description

Ecological system accounting system construction method and system based on multi-source data fusion Technical Field The invention relates to the technical field of ecological environment monitoring, in particular to a construction method and a construction system of an ecological system accounting system based on multi-source data fusion. Background The multi-source data fusion technology is a mainstream method of current large-scale and high-precision ecological accounting, a unified space-time data base is constructed by integrating multi-source data such as remote sensing, ground monitoring, weather, socioeconomic and the like, and an ecological process mechanism model or a statistical model is applied to carry out accounting on services such as carbon fixation, air purification, soil maintenance, water quality purification and the like in a specific space unit, and the spatial precision and index integrity of the accounting are improved by the full data-full model, so that reliable basis is provided for biological asset management and space planning. The method has the limitation in coping with the high-frequency and near-real-time dynamic monitoring demands, relies on full-area data collection and complex model calculation, and has the advantages of high calculation cost, long period and result lag, so that the conventional method is difficult to provide timely and continuous calculation results for management scenes needing immediate response, such as environmental emergency influence evaluation, ecological engineering effect rapid tracking and the like, and the application of the method in dynamic decision support is restricted. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the invention provides an ecological system accounting system construction method based on multi-source data fusion, which solves the problems of high cost, long period and result hysteresis caused by full-area data re-acquisition and full-model re-calculation when the high-frequency and near-real-time dynamic accounting requirements are met in the prior art. In order to solve the technical problems, the invention provides the following technical scheme: In a first aspect, the present invention provides a method for constructing an ecosystem accounting system based on multi-source data fusion, which includes collecting and fusing multi-source ecological data of a target area, and generating space-time aligned multi-source ecological data; Acquiring reference values of carbon fixation, air purification, soil maintenance and water purification of space-time grid units in a historical reference period based on historical data in the multi-source ecological data, and forming a historical reference value data set; according to the space association rule of the ecological process, a historical ecological system service relationship network is constructed, the structure of the historical ecological system service relationship network is analyzed, service nodes with key effects on the network structure and functions are identified, and a key service node set is formed; Constructing a key node monitoring network based on the key service node set, and training a global service deduction model based on the historical reference value data set and the historical ecosystem service relation network; and acquiring a real-time state vector of the key node through the key node monitoring network, inputting the real-time state vector of the key node into a global service deduction model, and deducting and outputting a global dynamic accounting result. The invention relates to an optimal scheme of an ecological system accounting system construction method based on multi-source data fusion, which comprises the following steps of collecting and fusing multi-source ecological data of a target area to generate time-space aligned multi-source ecological data: Acquiring space-sky remote sensing image data, ground internet of things sensor monitoring data, basic geographic data and socioeconomic activity data, and performing time reference unification and space coordinate system first registration; And resampling the space-day remote sensing image data, the ground internet of things sensor monitoring data, the basic geographic data and the socioeconomic activity data which are subjected to the registration of the time base unification and the space coordinate system unification to a preset space-time grid coordinate system to generate space-time aligned multisource ecological data. The invention relates to a construction method of an ecological system accounting system based on multi-source data fusion, which comprises the following steps of acquiring reference values of carbon fixation, air purification, soil maintenance and water purification of space-time grid units in a history reference period based on history data in the multi-source ecological data to form a