A method and system for automatically measuring carbon flux in forest ecosystem

Abstract

The present invention relates to the technical field of carbon flux determination, and specifically to a method and system for automatically determining carbon flux in a forest ecosystem, comprising the following steps: deploying a sensor collection network at different locations in the forest ecosystem to measure the CO₂ concentration and vertical wind speed in the atmosphere in real time; performing multiple preprocessing on the collected raw data; inputting the preprocessed data into a carbon flux calculation model, calculating the instantaneous carbon flux between the forest ecosystem and the atmosphere based on the principle of eddy covariance method, introducing high-order spectrum analysis technology to capture nonlinear interactions and high-order statistical characteristics, and generating the final instantaneous carbon flux value; calculating the influence coefficient according to a high-dimensional coupling model, and correcting the calculation result of the carbon flux in real time. The present invention improves the sensitivity and responsiveness of carbon flux determination to environmental changes, and the stability and accuracy of the carbon flux determination results.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon flux determination, and in particular to a method and system for automatically determining carbon flux of a forest ecosystem. Background Art

[0002] As the problem of global climate change becomes increasingly serious, the measurement and monitoring of carbon flux plays a vital role in the study of forest ecosystems. Forest ecosystems exchange a large amount of carbon with the atmosphere through processes such as photosynthesis, respiration and soil respiration, which directly affects the global carbon cycle. However, traditional carbon flux measurement methods mainly rely on eddy covariance method and soil respiration measurement method. Although these methods can provide basic carbon flux data, they still have many limitations in measurement accuracy, real-time performance and adaptability to complex environmental changes.

[0003] First, although the traditional eddy covariance method can measure the carbon flux between forest ecosystems and the atmosphere in real time, its measurement accuracy is often affected by high-frequency noise, environmental interference and data inconsistency.

[0004] Secondly, the dynamic impact of microclimate factors on carbon flux is often ignored during carbon flux measurement. Existing measurement methods are usually unable to effectively capture the complex nonlinear relationship between these microclimate factors and carbon flux, resulting in significant deviations in measurement results under different seasons, times and weather conditions. In traditional models, microclimate factors are mostly processed linearly or simply nonlinearly, lacking in-depth exploration of dynamic coupling relationships and the ability to respond to environmental changes in real time.

[0005] Finally, most of the carbon flux calculation models in the prior art are static models, lacking a dynamic adjustment mechanism for model parameters. When environmental conditions change, these static models are difficult to adjust in time to reflect actual carbon flux changes, resulting in a large gap between the measurement results and the actual situation. Summary of the invention

[0006] Based on the above objectives, the present invention provides a method and system for automatically measuring carbon flux in a forest ecosystem.

[0007] A method for automatically determining carbon flux in a forest ecosystem comprises the following steps:

[0008] S1, deploying a sensor collection network at different locations in the forest ecosystem. The sensor collection network includes an infrared gas analyzer (IRGA) and an ultrasonic anemometer to measure the CO2 concentration and vertical wind speed in the atmosphere in real time;

[0009] S2, multiple pre-processing of the collected raw data, including data cleaning and time synchronization, and smoothing of wind speed and CO2 concentration data using Kalman filter to eliminate high-frequency noise and environmental interference to ensure data accuracy;

[0010] S3, input the preprocessed data into the carbon flux calculation model, and calculate the instantaneous carbon flux between the forest ecosystem and the atmosphere based on the principle of eddy covariance method, which improves the accuracy of carbon flux calculation. The carbon flux calculation model is expressed as: ,in, represents the instantaneous carbon flux, represents the time-averaged covariance of vertical wind speed disturbance and CO2 concentration disturbance, represents the vertical wind speed, represents the CO2 concentration, w(t) is the vertical wind speed value measured at time t, c(t) is the CO2 concentration value measured at time t, and high-order spectrum analysis technology is introduced to capture nonlinear interactions and high-order statistical characteristics. High-order spectrum analysis is used to identify the nonlinear coupling and high-frequency components between wind speed and CO2 concentration. The instantaneous carbon flux is comprehensively calculated by combining the covariance results calculated by the eddy correlation method with the high-frequency components obtained by high-order spectrum analysis. The linear and nonlinear components are integrated through multivariate regression methods to generate the final instantaneous carbon flux value. ;

[0011] S4, analyze the dynamic coupling relationship between carbon flux and microclimate factors in forest ecosystems. Microclimate factors include temperature, humidity, atmospheric pressure and light. Atmospheric pressure affects gas diffusion and gas exchange processes, thereby affecting carbon flux. A high-dimensional coupling model is constructed using the generalized additive analysis method. The high-dimensional coupling model is used to describe the dynamic relationship between carbon flux and microclimate factors. The real-time collected microclimate factor data is input into the high-dimensional coupling model, and the influence coefficient of the change of microclimate factors on the carbon flux is output. The influence value reflects the deviation or change of carbon flux under specific environmental conditions. According to the influence coefficient calculated by the high-dimensional coupling model, key parameters are introduced to adjust the carbon flux calculation model in combination with the influence coefficient, and the calculation result of carbon flux is corrected in real time.

[0012] Optionally, the sampling points of the sensor collection network are arranged at different altitudes, canopy levels, open areas and dense vegetation areas. The selection of each sampling point is based on the spatial heterogeneity and diversity of the forest ecosystem to ensure the broad representativeness of the collected data. An infrared gas analyzer and an ultrasonic anemometer are arranged at each sampling point. The infrared gas analyzer monitors the CO2 concentration in the atmosphere in real time, and the ultrasonic anemometer is used to measure the vertical wind speed.

[0013] Optionally, the multiple preprocessing in S2 specifically includes:

[0014] S21, data cleaning: clean the raw data obtained from the acquisition network, fill in missing values ​​through interpolation, and filter noise points by setting a threshold range to ensure the integrity and reliability of the data;

[0015] S22, time synchronization: To ensure data consistency, all instrument timestamps are corrected and synchronized, and a distributed time synchronization protocol is used to unify the time references of different sampling points to ensure that the timestamps of all data are accurately aligned;

[0016] S23, Kalman filter: The Kalman filter is applied to the cleaned and synchronized wind speed and CO2 concentration data for smoothing. The Kalman filter recursively estimates the system state by combining real-time measurements with predicted values, eliminating high-frequency noise and environmental interference. The Kalman filter uses noise as an uncertain factor in state estimation, and generates smoothed wind speed and CO2 concentration data by continuously updating the covariance matrix of the system state and measurements, ensuring the data accuracy of the input carbon flux calculation model.

[0017] Optionally, the high-order spectrum analysis technique in S3 is expressed as:

[0018] ,in, represents the bispectrum, is the wind speed and The second-order autocorrelation function of the concentration disturbance signal, and is a frequency variable.

[0019] Optionally, the multivariate regression method is expressed as:

[0020] ,in, and is a model parameter reflecting the weights of linear and nonlinear components in carbon flux calculation.

[0021] Optionally, the S4 specifically includes:

[0022] Microclimate factor data collection: Real-time collection of microclimate factor data in forest ecosystems, including temperature ,humidity , atmospheric pressure And light intensity ,These data are collected synchronously with the carbon flux data to ensure data consistency during the analysis;

[0023] Based on the generalized additive analysis method, a high-dimensional coupling model (GAM) is constructed. The high-dimensional coupling model is expressed as:

[0024] ,

[0025] in, are nonlinear smooth functions, used to capture the nonlinear effects of each factor. is the interaction term between temperature and humidity, capturing the joint effect of the two;

[0026] Real-time data input and dynamic coupling relationship calculation: Input the real-time collected microclimate factor data into the high-dimensional coupling model to calculate the influence coefficient of microclimate factor changes on carbon flux , influence coefficient The calculation of is based on the difference between the model prediction and the actual measured value, and the formula is: ,in, is the carbon flux value predicted by the model, is the actual measured carbon flux value.

[0027] Optionally, the nonlinear smooth function further includes adaptive adjustment, the nonlinear smooth function Designed as an adaptive function, include , whose shape and complexity are dynamically adjusted according to real-time environmental data, introducing adaptive weight factors , which is determined by data-driven in order to be more flexible to seasonal and environmental changes, is expressed as: ,in, It is the traditional spline smoothing function form, represents the instantaneous value of a microclimate factor at time 𝑡, It is an adaptive weight factor determined by the historical data trend and real-time data volatility.

[0028] Optionally, the adaptive weight factor Based on the coefficient of variation, the coefficient of variation of the microclimate factor is calculated at each time step. : ,in, Microclimate factor The standard deviation of is its mean, and dynamically adjusts the adaptive weight factor according to the size of the coefficient of variation The weight of is used to enhance the model's ability to respond to high variability data.

[0029] Optionally, determining the parameter adjustment value of the carbon flux calculation model according to the influence coefficient calculated by the high-dimensional coupling model, adjusting the parameters in the carbon flux calculation model, and correcting the calculation result of the carbon flux in real time specifically includes:

[0030] The relationship between influence coefficient calculation and parameter adjustment: influence coefficients of high-dimensional coupling model output Indicates at time At this moment, due to the deviation of carbon flux prediction results caused by changes in microclimate factors, in order to ensure the accuracy of the carbon flux calculation model, it is necessary to Introducing key parameters to adjust the carbon flux calculation model, key parameters include covariance coefficient , time lag parameter ;

[0031] The covariance coefficient The adjustment is made by correcting the covariance coefficient in the carbon flux calculation model , to reflect the impact of microclimate factors on carbon flux prediction, when the influence coefficient is calculated by the high-dimensional coupling model When it is positive, it means that the current carbon flux model underestimates the actual carbon flux, so the increase On the contrary, when When it is negative, it means that the model overestimates the carbon flux, so reduce The value of this adjustment is determined by a step size coefficient Control the magnitude of the adjustment to ensure that the model can dynamically respond to changes in microclimate factors, thereby improving prediction accuracy;

[0032] The time lag parameter The adjustment is used to capture the time delay of carbon flux response to changes in microclimate factors, and the influence coefficient The rate of change over time to determine if adjustment is needed If the value of If the carbon flux is sensitive to changes over time (large rate of change), it means that there is a time lag in the response of the carbon flux to the microclimate factor, and it needs to be adjusted. To accurately reflect the hysteresis relationship;

[0033] Real-time correction of carbon flux calculation results: according to and , recalculate the instantaneous carbon flux , expressed as:

[0034] ,in, is the corrected carbon flux value, is the error term after time lag correction.

[0035] A system for automatically measuring carbon flux of a forest ecosystem, used to implement the above-mentioned method for automatically measuring carbon flux of a forest ecosystem, comprising:

[0036] Multi-level sensor network: It consists of infrared gas analyzers and ultrasonic anemometers distributed in different locations of the forest ecosystem, which are used to collect real-time data on CO2 concentration and vertical wind speed in the atmosphere and transmit the data to the central control unit through a wireless network;

[0037] Data preprocessing module: used to perform multiple preprocessing on the collected raw data, including data cleaning and time synchronization;

[0038] Carbon flux calculation module: Based on the eddy covariance method and high-order spectrum analysis technology, combined with generalized additive analysis technology, it calculates the instantaneous carbon flux between the forest ecosystem and the atmosphere in real time, and analyzes the dynamic coupling relationship between carbon flux and microclimate factors;

[0039] Parameter dynamic adjustment module: According to the influence coefficient calculated by the high-dimensional coupling model, the covariance coefficient and time lag parameter in the carbon flux calculation model are dynamically adjusted to correct the calculation results of the carbon flux in real time to ensure the accuracy and long-term stability of the measurement results.

[0040] Beneficial effects of the present invention:

[0041] The present invention improves the accuracy of carbon flux measurement by deploying a high-precision sensor network in the forest ecosystem, collecting wind speed and CO2 concentration data in real time, and processing the data using a carbon flux calculation model. The instantaneous carbon flux calculation method combining eddy covariance method and high-order spectral analysis technology can capture the dynamic changes of carbon flux in real time and provide measurement results with high time resolution.

[0042] The present invention analyzes the dynamic coupling relationship between carbon flux and microclimate factors (temperature, humidity, atmospheric pressure, and light) in forest ecosystems, and uses an improved generalized additive model (GAM) to construct a high-dimensional coupling model. The model not only captures the complex nonlinear relationship between microclimate factors and carbon flux, but also introduces an adaptive smoothing function to dynamically adjust model parameters according to real-time environmental data, thereby improving the sensitivity and responsiveness of carbon flux measurement to environmental changes. The adaptive adjustment mechanism ensures the stability and accuracy of carbon flux measurement results under different seasons, times, and weather conditions.

[0043] The present invention dynamically adjusts the key parameters in the carbon flux calculation model according to the influence coefficient calculated by the high-dimensional coupling model. This adjustment mechanism makes real-time corrections to the model parameters so that the carbon flux calculation results can accurately reflect the current forest carbon exchange status, ensuring that the carbon flux measurement system can still provide high-precision measurement results when facing a complex and ever-changing microclimate environment, and has significant practical application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 Schematic diagram of a determination method according to an embodiment of the present invention;

[0046] Figure 2 A schematic diagram of a method for correcting calculation results of carbon flux according to an embodiment of the present invention;

[0047] Figure 3 Schematic diagram of the functional modules of the measurement system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0049] It should be noted that the references to "one embodiment", "embodiment", "exemplary embodiments", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0050] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0051] like Figure 1-Figure 2 As shown, a method for automatically determining carbon flux in a forest ecosystem comprises the following steps:

[0052] S1, deploying a sensor collection network at different locations in the forest ecosystem. The sensor collection network includes an infrared gas analyzer (IRGA) and an ultrasonic anemometer to measure the CO2 concentration and vertical wind speed in the atmosphere in real time;

[0053] S2, multiple pre-processing of the collected raw data, including data cleaning and time synchronization, and smoothing of wind speed and CO2 concentration data using Kalman filter to eliminate high-frequency noise and environmental interference to ensure data accuracy;

[0054] S3, input the preprocessed data into the carbon flux calculation model, and calculate the instantaneous carbon flux between the forest ecosystem and the atmosphere based on the principle of eddy covariance method, which improves the accuracy of carbon flux calculation. The carbon flux calculation model is expressed as: ,in, represents the instantaneous carbon flux, represents the time-averaged covariance of vertical wind speed disturbance and CO2 concentration disturbance, represents the vertical wind speed, represents the CO2 concentration, w(t) is the vertical wind speed value measured at time t, c(t) is the CO2 concentration value measured at time t, and the disturbance is calculated as: ,in, and The wind speed and Time average of concentration; In order to improve the accuracy of carbon flux calculation, high-order spectrum analysis technology is introduced to capture nonlinear interactions and high-order statistical characteristics. High-order spectrum analysis is used to identify nonlinear coupling and high-frequency components between wind speed and CO2 concentration. The instantaneous carbon flux is comprehensively calculated by combining the covariance results calculated by the eddy correlation method with the high-frequency components obtained by high-order spectrum analysis. The linear and nonlinear components are integrated through multivariate regression methods to generate the final instantaneous carbon flux value. ;

[0055] In forest ecosystems, the relationship between carbon flux and environmental variables (such as wind speed and CO2 concentration) is not only linear, but also involves complex nonlinear interactions. Through comprehensive calculations, including eddy covariance method (capturing linear interactions) and high-order spectral analysis (identifying nonlinear interactions), these complex relationships can be more comprehensively described and quantified, thereby more accurately estimating carbon flux. Instantaneous carbon flux is the measurement result at a certain moment, but instantaneous data can be accumulated to form long-term series data, providing a basis for further trend analysis and carbon balance research. By integrating a large amount of instantaneous carbon flux data, the carbon balance of forest ecosystems in the long term can be more accurately estimated.

[0056] S4, analyze the dynamic coupling relationship between carbon flux and microclimate factors in forest ecosystems. Microclimate factors include temperature, humidity, atmospheric pressure and light. Atmospheric pressure affects gas diffusion and gas exchange processes, thereby affecting carbon flux. A high-dimensional coupling model is constructed using the generalized additive analysis method. The high-dimensional coupling model is used to describe the dynamic relationship between carbon flux and microclimate factors. The real-time collected microclimate factor data is input into the high-dimensional coupling model, and the influence coefficient of the change of microclimate factors on the carbon flux is output. The influence value reflects the deviation or change of carbon flux under specific environmental conditions. According to the influence coefficient calculated by the high-dimensional coupling model, key parameters are introduced in combination with the influence coefficient to adjust the carbon flux calculation model, and the calculation results of carbon flux are corrected in real time to ensure that the measurement results can accurately reflect the current forest carbon flux status and maintain stability and accuracy under different seasons, times and weather conditions.

[0057] The sampling points of the sensor collection network are arranged at different altitudes, canopy levels, open areas and dense vegetation areas. The selection of each sampling point is based on the spatial heterogeneity and diversity of the forest ecosystem to ensure the broad representativeness of the collected data. Infrared gas analyzers and ultrasonic anemometers are deployed at each sampling point. The infrared gas analyzer monitors the CO2 concentration in the atmosphere in real time, and the ultrasonic anemometer is used to measure the vertical wind speed. The instrument should be installed in a place to avoid obstacles to ensure the accuracy and consistency of data measurement. The installation height of the instrument is adjusted according to the height of the canopy and the density of vegetation to capture carbon flux data at different vertical levels.

[0058] The sensors at all sampling points are connected to the central control unit via a wireless network to form an integrated acquisition network. The acquisition network is equipped with a time synchronization mechanism to ensure that the timestamps of all sensor data are consistent, which facilitates subsequent data analysis and processing.

[0059] The local climate conditions and terrain characteristics were taken into consideration when laying out the instruments. Weather-resistant equipment was used and reinforced with fixed brackets and protective casings to ensure stable operation even in extreme weather conditions.

[0060] The multiple preprocessing in S2 specifically includes:

[0061] S21, data cleaning: clean the raw data obtained from the acquisition network, fill in missing values ​​through interpolation, and filter noise points by setting a threshold range to ensure the integrity and reliability of the data;

[0062] S22, time synchronization: To ensure data consistency, all instrument timestamps are corrected and synchronized, and a distributed time synchronization protocol (Network Time Protocol NTP or IEEE 1588 protocol) is used to unify the time reference of different sampling points to ensure that the timestamps of all data are accurately aligned;

[0063] S23, Kalman filter: The Kalman filter is applied to the cleaned and synchronized wind speed and CO2 concentration data for smoothing. The Kalman filter recursively estimates the system state by combining real-time measurements with predicted values, eliminating high-frequency noise and environmental interference. The Kalman filter uses noise as an uncertain factor in state estimation, and generates smoothed wind speed and CO2 concentration data by continuously updating the covariance matrix of the system state and measurements, ensuring the data accuracy of the input carbon flux calculation model.

[0064] Recursive update equation of Kalman filter: For the cleaned and synchronized wind speed and CO2 concentration data, Kalman filter is applied for smoothing, and the expression is as follows:

[0065] Status prediction: ,in, It is at the moment The predicted state estimate, is the state transition matrix, is the control input matrix, is the control input.

[0066] Covariance prediction: ,in, is the predicted covariance matrix, is the process noise covariance matrix.

[0067] Update the Kalman gain: ,in, is the Kalman gain, is the observation matrix, is the observation noise covariance matrix.

[0068] Status Update: ,in, It's time The observed value of is the updated state estimate.

[0069] Covariance update: ,in, is the updated covariance matrix, is the identity matrix.

[0070] Through the above processing steps of the Kalman filter, the wind speed and CO2 concentration data can be recursively corrected during the real-time measurement process, high-frequency noise and environmental interference can be eliminated, and smoother and more accurate data results can be output, thereby ensuring the accuracy of carbon flux calculation.

[0071] The high-order spectral analysis technique in S3 is expressed as:

[0072] ,in, represents the bispectrum, is the wind speed and The second-order autocorrelation function of the concentration disturbance signal, and is a frequency variable.

[0073] The multivariate regression method is expressed as:

[0074] ,in, and It is a model parameter that reflects the weight of linear and nonlinear components in carbon flux calculation. Through training and verification of historical data, the parameters are optimized to improve the accuracy of carbon flux calculation. The instantaneous carbon flux finally calculated is Will be used as the output of carbon flux determination.

[0075] S4 specifically includes:

[0076] Microclimate factor data collection: Real-time collection of microclimate factor data in forest ecosystems, including temperature ,humidity , atmospheric pressure And light intensity ,These data are collected synchronously with the carbon flux data to ensure data consistency during the analysis;

[0077] Construction of a high-dimensional coupling model (GAM) based on the generalized additive analysis method: Based on the standard generalized additive analysis method and combined with the ecological characteristics of forest ecosystems, the high-dimensional coupling model is expressed as:

[0078] ,

[0079] in, are nonlinear smooth functions, used to capture the nonlinear effects of each factor. is the interaction term between temperature and humidity, capturing the joint effect of the two;

[0080] Real-time data input and dynamic coupling relationship calculation: Input the real-time collected microclimate factor data into the high-dimensional coupling model to calculate the influence coefficient of microclimate factor changes on carbon flux , which reflects the expected deviation of carbon flux under current environmental conditions. The calculation of is based on the difference between the model prediction and the actual measured value, and the formula is: ,in, is the carbon flux value predicted by the model, is the actual measured carbon flux value, and the influence coefficient reflects the variation deviation of carbon flux under specific environmental conditions.

[0081] The nonlinear smoothing function also includes adaptive adjustment. According to the environmental characteristics of forest ecosystems, the nonlinear smoothing function Designed as an adaptive function, include , whose shape and complexity are dynamically adjusted according to real-time environmental data, introducing adaptive weight factors , which is determined by data-driven in order to be more flexible to seasonal and environmental changes, is expressed as: ,in, It is the traditional spline smoothing function form, represents the instantaneous value of a microclimate factor (temperature, humidity, atmospheric pressure or light intensity) at time 𝑡, It is an adaptive weight factor determined by the historical data trend and real-time data volatility. , the model can automatically adjust when environmental conditions change rapidly to avoid overfitting or underfitting. In the high-dimensional coupling model, joint training is performed based on historical data and real-time collected data, and the model coefficients are dynamically updated using the online gradient descent method. , joint training is used to gradually adapt to long-term changes in forest ecosystems while maintaining sensitivity to short-term fluctuations.

[0082] Adaptive weight factor Based on the coefficient of variation, the coefficient of variation of the microclimate factor is calculated at each time step. : ,in, Microclimate factor The standard deviation of is its mean, and dynamically adjusts the adaptive weight factor according to the size of the coefficient of variation The weight of is used to enhance the model's ability to respond to high variability data.

[0083] According to the influence coefficient calculated by the high-dimensional coupling model, the parameter adjustment value of the carbon flux calculation model is determined, the parameters in the carbon flux calculation model are adjusted, and the calculation results of the carbon flux are corrected in real time. Specifically, the following are included:

[0084] The relationship between influence coefficient calculation and parameter adjustment: influence coefficients of high-dimensional coupling model output Indicates at time At this moment, due to the deviation of carbon flux prediction results caused by changes in microclimate factors, in order to ensure the accuracy of the carbon flux calculation model, it is necessary to Introducing key parameters to adjust the carbon flux calculation model, key parameters include covariance coefficient , time lag parameter ,in;

[0085] Covariance coefficient : Used to describe vertical wind speed and Concentration disturbance The relationship between them directly affects the accuracy of carbon flux calculation;

[0086] Time lag parameter : Reflects the time delay of carbon flux in response to changes in microclimate factors.

[0087] Covariance coefficient The adjustment is made by correcting the covariance coefficient in the carbon flux calculation model , to reflect the impact of microclimate factors on carbon flux prediction, when the influence coefficient is calculated by the high-dimensional coupling model When it is positive, it means that the current carbon flux model underestimates the actual carbon flux, so the increase On the contrary, when When it is negative, it means that the model overestimates the carbon flux, so reduce The value of this adjustment is determined by a step size coefficient Control the magnitude of the adjustment to ensure that the model can dynamically respond to changes in microclimate factors, thereby improving prediction accuracy;

[0088] Time lag parameter The adjustment is used to capture the time delay of carbon flux response to changes in microclimate factors, and the influence coefficient The rate of change over time to determine if adjustment is needed If the value of If the carbon flux is sensitive to changes over time (large rate of change), it means that there is a time lag in the response of the carbon flux to the microclimate factor, and it needs to be adjusted. To accurately reflect the lag relationship, the adjustment range is determined by the coefficient Control, so that the model can better capture dynamic changes and improve the accuracy of the measurement;

[0089] The specific calculation expression is as follows:

[0090] Covariance coefficient The adjustment is expressed as: ,in, is the adjustment step size coefficient, according to Size and orientation adjustment Value: When When it is positive, it means that the current model underestimates the carbon flux and needs to be increased. value; when When it is negative, it means that the carbon flux is overestimated and needs to be reduced. value.

[0091] Time lag parameter The adjustment is expressed as: ,in, is the adjustment factor, is the rate of change of the influence coefficient over time, if Sensitive to time changes, that is, the rate of change is large, so adjustment is required To better capture the dynamic response between carbon flux and microclimate factors.

[0092] Real-time correction of carbon flux calculation results: according to and , recalculate the instantaneous carbon flux , expressed as:

[0093] ,in, is the corrected carbon flux value, This is the error term after time lag correction. By adjusting the parameter value, it is ensured that the carbon flux calculation results can more accurately reflect the carbon exchange status in the real-time environment.

[0094] like Figure 3 As shown, a system for automatically measuring carbon flux of a forest ecosystem is used to implement the above-mentioned method for automatically measuring carbon flux of a forest ecosystem, comprising:

[0095] Multi-level sensor network: It consists of infrared gas analyzers and ultrasonic anemometers distributed in different locations of the forest ecosystem, which are used to collect real-time data on CO2 concentration and vertical wind speed in the atmosphere and transmit the data to the central control unit through a wireless network;

[0096] Data preprocessing module: used to perform multiple preprocessing on the collected raw data, including data cleaning and time synchronization;

[0097] Carbon flux calculation module: Based on the eddy covariance method and high-order spectrum analysis technology, combined with generalized additive analysis technology, it calculates the instantaneous carbon flux between the forest ecosystem and the atmosphere in real time, and analyzes the dynamic coupling relationship between carbon flux and microclimate factors;

[0098] Parameter dynamic adjustment module: According to the influence coefficient calculated by the high-dimensional coupling model, the covariance coefficient and time lag parameter in the carbon flux calculation model are dynamically adjusted to correct the calculation results of the carbon flux in real time to ensure the accuracy and long-term stability of the measurement results.

[0099] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0100] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for automatically determining carbon flux in a forest ecosystem, characterized in that: The following steps are involved: S1, deploying a sensor collection network at different locations in the forest ecosystem. The sensor collection network includes infrared gas analyzers and ultrasonic anemometers to measure the CO2 concentration and vertical wind speed in the atmosphere in real time; S2, multiple preprocessing of the collected raw data, including data cleaning and time synchronization, and smoothing of wind speed and CO2 concentration data using Kalman filter to eliminate high-frequency noise and environmental interference; S3, input the preprocessed data into the carbon flux calculation model, and calculate the instantaneous carbon flux between the forest ecosystem and the atmosphere based on the principle of eddy covariance method. The carbon flux calculation model is expressed as: ,in, represents the instantaneous carbon flux, represents the time-averaged covariance of vertical wind speed disturbance and CO2 concentration disturbance, represents the vertical wind speed, represents the CO2 concentration, w(t) is the vertical wind speed value measured at time t, c(t) is the CO2 concentration value measured at time t, and high-order spectrum analysis technology is introduced to capture nonlinear interactions and high-order statistical characteristics. High-order spectrum analysis is used to identify the nonlinear coupling and high-frequency components between wind speed and CO2 concentration. The instantaneous carbon flux is comprehensively calculated by combining the covariance results calculated by the eddy correlation method with the high-frequency components obtained by high-order spectrum analysis. The linear and nonlinear components are integrated through multivariate regression methods to generate the final instantaneous carbon flux value. ; The high-order spectrum analysis technique is expressed as: ,in, represents the bispectrum, is the wind speed and The second-order autocorrelation function of the concentration disturbance signal, and is a frequency variable; The multivariate regression method is expressed as: ,in, and is a model parameter reflecting the weights of linear and nonlinear components in carbon flux calculation; S4, analyzing the dynamic coupling relationship between carbon flux and microclimate factors in the forest ecosystem, microclimate factors include temperature, humidity, atmospheric pressure and light, using the generalized additive analysis method to build a high-dimensional coupling model, the high-dimensional coupling model is used to describe the dynamic relationship between carbon flux and microclimate factors, the real-time collected microclimate factor data is input into the high-dimensional coupling model, the influence coefficient of the change of microclimate factors on the carbon flux is output, according to the influence coefficient calculated by the high-dimensional coupling model, the key parameters are introduced in combination with the influence coefficient to adjust the carbon flux calculation model, and the calculation result of the carbon flux is corrected in real time; The S4 specifically includes: S41, Microclimate factor data collection: Real-time collection of microclimate factor data in forest ecosystems, including temperature ,humidity , atmospheric pressure And light intensity ; The high-dimensional coupling model is constructed based on the generalized additive analysis method. The high-dimensional coupling model is expressed as: ， in, are nonlinear smooth functions, used to capture the nonlinear effects of each factor. is the interaction term between temperature and humidity, capturing the joint effect of the two; S42, real-time data input and dynamic coupling relationship calculation: input the real-time collected microclimate factor data into the high-dimensional coupling model to calculate the influence coefficient of microclimate factor changes on carbon flux , influence coefficient The calculation of is based on the difference between the model prediction and the actual measured value, and the formula is: ,in, is the carbon flux value predicted by the model, is the actual measured carbon flux value; The nonlinear smooth function also includes adaptive adjustment, the nonlinear smooth function Designed as an adaptive function, include , whose shape and complexity are dynamically adjusted according to real-time environmental data, introducing adaptive weight factors , expressed as: ,in, It is the traditional spline smoothing function form, represents the instantaneous value of a microclimate factor at time 𝑡, It is an adaptive weight factor determined by the trend of historical data and the volatility of real-time data; The adaptive weight factor Based on the coefficient of variation, the coefficient of variation of the microclimate factor is calculated at each time step. : ,in, Microclimate factor The standard deviation of is its mean, and dynamically adjusts the adaptive weight factor according to the size of the coefficient of variation The weight of In S4, according to the influence coefficient calculated by the high-dimensional coupling model, the parameter adjustment value of the carbon flux calculation model is determined, the parameters in the carbon flux calculation model are adjusted, and the calculation result of the carbon flux is corrected in real time, specifically including: S43, Correlation between influence coefficient calculation and parameter adjustment: Influence coefficients output by high-dimensional coupling models Indicates at time At this moment, due to the change of microclimate factors, the deviation of carbon flux prediction results is Introducing key parameters to adjust the carbon flux calculation model, key parameters include covariance coefficient , time lag parameter ; Covariance coefficient The adjustment is expressed as: ,in, is the adjustment step size coefficient, according to Size and orientation adjustment Value, when When it is positive, it means that the current model underestimates the carbon flux and needs to be increased. value; when When it is negative, it means that the carbon flux is overestimated and needs to be reduced. value; Time lag parameter The adjustment is expressed as: ,in, is the adjustment factor, is the rate of change of the influence coefficient over time; The covariance coefficient The adjustment is made by correcting the covariance coefficient in the carbon flux calculation model , to reflect the impact of microclimate factors on carbon flux prediction, when the influence coefficient is calculated by the high-dimensional coupling model When it is positive, it means that the current carbon flux model underestimates the actual carbon flux, so the increase On the contrary, when When it is negative, it means that the model overestimates the carbon flux, so reduce The value of The time lag parameter The adjustment is used to capture the time delay of carbon flux response to changes in microclimate factors, and the influence coefficient The rate of change over time to determine if adjustment is needed If the value of If the carbon flux is sensitive to changes over time, it means that there is a time lag in the response of the carbon flux to the microclimate factors, which needs to be adjusted. To accurately reflect the hysteresis relationship; S44, real-time correction of carbon flux calculation results: according to and , recalculate the instantaneous carbon flux , expressed as: ,in, is the corrected carbon flux value, is the error term after time lag correction.

2. The method for automatically determining carbon flux of a forest ecosystem according to claim 1, characterized in that: The sampling points of the sensor collection network are arranged at different altitudes, tree canopy levels, open areas and dense vegetation areas. An infrared gas analyzer and an ultrasonic anemometer are arranged at each sampling point. The infrared gas analyzer monitors the CO2 concentration in the atmosphere in real time, and the ultrasonic anemometer is used to measure the vertical wind speed.

3. The method for automatically determining carbon flux of a forest ecosystem according to claim 1, characterized in that: The multiple preprocessing in S2 specifically includes: S21, data cleaning: clean the raw data obtained from the acquisition network, fill in missing values ​​by interpolation, and filter noise points by setting a threshold range; S22, time synchronization: calibrate and synchronize the timestamps of all instruments, use a distributed time synchronization protocol, unify the time references of different sampling points, and ensure that the timestamps of all data are accurately aligned; S23, Kalman filter: The Kalman filter is applied to the cleaned and synchronized wind speed and CO2 concentration data for smoothing. The Kalman filter recursively estimates the system state by combining real-time measurement values ​​with predicted values, eliminating high-frequency noise and environmental interference. The Kalman filter regards noise as an uncertain factor in state estimation, and generates smoothed wind speed and CO2 concentration data by continuously updating the covariance matrix of the system state and measurement values.

4. A system for automatically measuring carbon flux in a forest ecosystem, used to implement a method for automatically measuring carbon flux in a forest ecosystem as claimed in any one of claims 1 to 3, characterized in that: include: Multi-level sensor network: It consists of infrared gas analyzers and ultrasonic anemometers distributed in different locations of the forest ecosystem, which are used to collect real-time data on CO2 concentration and vertical wind speed in the atmosphere and transmit the data to the central control unit through a wireless network; Data preprocessing module: used to perform multiple preprocessing on the collected raw data, including data cleaning and time synchronization; Carbon flux calculation module: Based on the eddy covariance method and high-order spectrum analysis technology, combined with generalized additive analysis technology, it calculates the instantaneous carbon flux between the forest ecosystem and the atmosphere in real time, and analyzes the dynamic coupling relationship between carbon flux and microclimate factors; Parameter dynamic adjustment module: According to the influence coefficient calculated by the high-dimensional coupling model, the covariance coefficient and time lag parameter in the carbon flux calculation model are dynamically adjusted to correct the calculation results of the carbon flux in real time to ensure the accuracy and long-term stability of the measurement results.

Images