Method and system for rapidly evaluating carbon sequestration capability of biochar composite material
By applying labeled mineral composite biochar in the field and monitoring the carbon loss rate in real time, combined with comparison with a benchmark database, the carbon sequestration capacity of biochar composite materials can be quickly evaluated. This solves the problem of long evaluation cycles in existing technologies and improves the timeliness and reliability of the evaluation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-14
- Publication Date
- 2026-03-24
AI Technical Summary
In existing technologies, assessing the carbon sequestration capacity of biochar composite materials in neutral or alkaline soils requires a long time period, making it difficult to efficiently verify the stability of the material configuration and achieve rapid and sustainable monitoring.
By applying labeled mineral biochar in the field, the carbon loss rate is monitored in real time at high frequency, the loss deviation is calculated, and the loss pattern fitting quality is obtained by comparing with a benchmark database. This is used as an evaluation standard to assess carbon sequestration capacity.
This method enables rapid assessment of the carbon fixation capacity of biochar composites, improves the timeliness and reliability of the assessment, reduces the frequency and cost of traditional long-term buried sample analysis, and provides a reliable basis for long-term carbon fixation research.
Smart Images
Figure CN121725944A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of data analysis technology, specifically relating to a method and system for rapidly evaluating the carbon fixation capacity of biochar composite materials. Background Technology
[0002] While biochar possesses excellent carbon sequestration potential, its surface chemistry and carbon stabilization capabilities have certain limitations. Fresh biochar is rich in oxygen-containing functional groups, primarily carboxyl and phenolic hydroxyl groups, which can adsorb CO2 or organic carbon. However, these functional groups undergo oxidation and aging due to reduced microbial activity over time, often preventing them from effectively releasing carbon sequestration capacity for extended periods. In particular, biochar surface passivation weakens its adsorption capacity in neutral or alkaline soils because biochar surface functional groups readily form bridging bonds with divalent cations such as Ca²⁺ and Mg²⁺ in alkaline soils, occupying active sites and thus passivating the surface. High CEC minerals have good cation buffering capacity and can competitively adsorb free Ca²⁺ and Mg²⁺ ions in the soil, thereby reducing their chance of binding with biochar functional groups and slowing down the surface purification rate of biochar. In addition, the composite surface of soil minerals and biochar can gradually form a thin reaction film formed by the co-deposition of organic matter, metal ions and minerals, thereby reducing the chance of penetration by external oxidants and reducing the migration and desorption of free functional groups in biochar, thus slowing down the aging rate of biochar.
[0003] The 13C labeling method was used to periodically sample and analyze the changes in 13C content in the soil. The mass loss of biochar can be directly calculated by the reduction of 13C content, thereby assessing its decomposition rate. The stable carbon fraction method can be used to assess the proportion of recalcitrant carbon in composite biochar materials, which is an important indicator for predicting carbon sequestration capacity. However, the sustainable assessment of the carbon sequestration capacity of biochar and CEC soil-mineral composite materials often requires long-term performance inference, which cannot efficiently verify the stability of material configuration. Therefore, a rapid assessment method and system for the carbon sequestration capacity of biochar composite materials is needed to improve the efficiency of sustainability monitoring. Summary of the Invention
[0004] The purpose of this invention is to provide a rapid evaluation method and system for the carbon fixation capacity of biochar composite materials, so as to solve one or more technical problems existing in the prior art, and at least provide a beneficial option or create conditions.
[0005] To achieve the above objectives, according to one aspect of the present invention, a method for rapidly evaluating the carbon fixation capacity of biochar composite materials is provided, the method comprising the following steps: S100 involves the field application of marked mineral-based biochar. S200 is used to obtain the carbon content loss rate of composite biochar through high-frequency real-time monitoring. S300, obtains the loss deviation based on the carbon content loss rate between measurement intervals; S400, the churn pattern fitting quality is obtained by comparing the churn deviation with the benchmark database; The S500 uses the quality of the loss pattern fitting as an evaluation criterion to select reference samples for assessing carbon sequestration capacity and outputs evaluation data.
[0006] Further, in step S100, the method for field application of the labeled mineral composite biochar is as follows: biochar labeled with 13C stable isotope is mixed evenly with minerals of high cation exchange capacity in a preset ratio to obtain a mineral composite biochar sample; the sample is evenly spread on the surface of the target test plot at a unit area application rate of 2-10 t / ha, and covered with soil 2-5 cm to ensure full contact with the native soil; background soil samples are collected before and after application, and initial soil samples are collected after application to establish a 13C content baseline.
[0007] 13C refers to carbon-13, a stable isotope of carbon. Additionally, soil pH, CEC, organic matter content, temperature, humidity, and meteorological conditions must be recorded, and this information must be linked to the sample number for subsequent carbon sequestration rate correction and database comparison.
[0008] The application rate per unit area needs to be preset according to the soil environment. Generally, when the soil organic matter content is less than 1.0% or the pH is greater than 7.5, the upper limit of application rate is 6 to 10 t / ha to enhance the interface between cation buffer and carbon adsorption. When the soil organic matter content is greater than 2.0% or it is neutral soil with a pH range of 6.5±0.3, the upper limit of application rate is 2 to 5 t / ha to avoid excessive concentration of surface carbon and oxidation deviation.
[0009] Regarding the setting of the soil cover thickness, if the surface wind speed or surface evaporation rate in the test area is high, a soil cover thickness of 4-5 cm should be selected; otherwise, a middle value or a lower value should be selected.
[0010] The goal of this step is to construct an initial experimental state with 13C labeling, uniform spatial distribution, and clear environmental parameters. This will provide a stable assessment quality for subsequent high-frequency monitoring to accurately calculate the carbon loss rate and compare it with the database, thus enabling rapid evaluation of the carbon fixation capacity of biochar composites.
[0011] Further, in step S200, the method for obtaining the carbon content loss rate by high-frequency real-time monitoring of composite biochar is as follows: the preset measurement interval TG value range is 5-60 minutes; a flux chamber is set up above the release area, and gas samples are continuously collected at sampling intervals of 5-60 minutes; the CO2 volume fraction and 13C:12C isotope ratio of the gas sample are measured in real time using an isotope analyzer, and the 13CO2 flux generated by the composite biochar is calculated by inputting the isotope source model, and the 13CO2 flux change rate per unit time is used as the carbon content loss rate; the carbon content loss rate signal is corrected for drift due to temperature, humidity, air pressure and wind speed, and the carbon content loss rate sequence is output.
[0012] Before or during the online isotope monitoring in step S200, daily calibration and drift correction are performed using zero gas and two-point standard gas; the flux chamber is connected to the gas analysis unit, which includes an isotope analyzer that is capable of distinguishing between 13CO2 and 12CO2.
[0013] During the process of correcting the carbon loss rate signal for temperature, humidity, air pressure, and wind speed drift, the environmental parameters are obtained from a meteorological module deployed near the flux chamber, including temperature and humidity probes, a barometer, and an ultrasonic anemometer. Drift correction employs a joint compensation algorithm combining multiple regression and LOESS or a Kalman filter algorithm based on a state-space model to obtain the environmentally normalized carbon loss rate. The unit time for calculating the carbon loss rate is the TG value. The output carbon loss rate sequence is obtained once at each measurement interval. Using multiple flux chambers can improve the numerical accuracy of carbon loss rate and enhance the quality of subsequent assessments. When multiple flux chambers are used, the rates at each measurement point are spatially weighted and averaged to obtain the carbon loss rate in a representative area.
[0014] Further, in step S300, the method for obtaining the loss deviation based on the carbon content loss rate between measurement intervals is as follows: the content loss rate sequence obtained at the current time is denoted as clrLst, and clrLst(n1) represents its n1th element; the content loss rate sequence obtained at the previous time is denoted as la_clrLst, and la_clrLst(n2) represents its n2th element; the first loss rate is denoted as clrLst(1) - clrLst(2), and the second loss rate is denoted as clrLst(1) - la_clrLst(1); the loss deviation is obtained by dynamically weighting the first loss rate and the second loss rate through a state-space model based on Kalman filtering.
[0015] The first element of the content loss rate sequence is denoted as clrLst(1), which refers to the content loss rate measured at the current time. The second element of the content loss rate sequence is denoted as clrLst(2), which refers to the content loss rate measured at the previous time. The content loss rate sequence obtained at the previous time is denoted as la_clrLst, and the first element of the content loss rate measured at the previous time is la_clrLst(1). The implementation process of the dynamic weighted fusion is as follows: based on the currently obtained content loss rate sequence, a state-space model is established with drift and environmental disturbance as observation inputs. The system state and measurement error variance are estimated in real time through Kalman filtering, and the first loss rate and the second loss rate are dynamically weighted to generate loss deviation. The drift indicates the baseline shift of the instrument due to changes in temperature, light source or detector response during continuous monitoring. The environmental disturbance includes instantaneous interference caused by wind speed, air pressure, temperature and humidity on CO2 flux measurement.
[0016] Due to the updating of the content loss rate sequence, discrepancies arise at different calculation times for the same point in time. These discrepancies are caused by drift correction. The reason for drift correction is that the flux chamber and isotope analyzer are affected by minute drifts in dynamic external factors such as temperature, air pressure, infrared source intensity, and detector gain during continuous measurement. This causes the baseline of the original CO2 volume fraction and the 13CO2 / 12CO2 ratio to slowly shift over time. Once the new drift parameters are obtained through zero gas and two-point standard gas calibration, the system backcalculates the original signals from previous times based on the latest calibration curve, thus forming a posterior recalculated value for the same time point. This difference caused by drift correction is not noise, but rather a manifestation of the self-correcting mechanism established in this method. Drift correction, by introducing the latest reference gas information, ensures that previous rate values are consistently calibrated under the new drift model, thereby compensating for the slow accumulation of errors caused by the instrument and the environment. Therefore, the loss deviation obtained by weighted averaging the two deviations can dynamically reflect the true carbon loss trend after drift correction, and realize the time self-consistency and drift self-repair of the monitoring results, which is the key basis for the rapid assessment of carbon sequestration capacity of this invention.
[0017] Further, in step S400, the method for obtaining the fitting quality of the churn pattern by comparing the churn deviation with the benchmark database is as follows: all churn deviations obtained before the current time are constructed in chronological order to form a churn deviation sequence with a length of len; the churn deviation sequences of the reference samples stored in the benchmark database are each truncated with a prefix subsequence of length len as a reference deviation sequence; the reference deviation sequence and the churn deviation sequence are denoted as the fitting sequence; a time period DW is preset as an early window; the shape parameter obtained by performing Weibull least squares fitting on the first DW elements of the fitting sequence is denoted as the early decay degree. The shape parameters are the result parameters from the Weibull least squares fitting algorithm; The early window ranges from 15 to 30 calendar days. The principle behind constructing the early decay characteristic is that the carbon release from the biochar complex follows a two-stage decay pattern. In the early stage, the fast-moving component, which is easily decomposed organic carbon, dominates and has a high release rate. In the later stage, the slow-moving component, which is mainly composed of stable carbon framework, dominates and its release rate tends to be slower. Therefore, the early decay value actually fits the kinetic rate of the conversion of easily decomposed carbon to stable carbon in the complex system. This early decay value, which can be obtained early, can significantly improve the efficiency of assessment conclusions and screen out samples that are still in the unstable stage.
[0018] The characteristics of the samples that are still in the unstable period in terms of early decay are that the shape parameter of the Weibull fit is less than 1 and the scale parameter is small, which makes the early decay value significantly higher than that of the stable samples. It can be deduced that the carbon release of the sample is a sharp drop with a large variance, so it is still in the stage dominated by easily decomposable carbon.
[0019] A time period is defined as the dynamic window MW, with the dynamic window value ranging from 5 to 10 natural days. The exponential average of each element within the MW period before any time in the fitted sequence is taken as the dynamic mean. The difference between any element in the fitted sequence and its corresponding dynamic mean is taken as the dynamic mean deviation. All negative values in the dynamic mean deviation are updated to 0. Any number of consecutive and non-zero dynamic mean deviations corresponding to the time periods constitute a high mean deviation segment. The standard deviation of each element in the fitted sequence corresponding to the high mean deviation segment is denoted as the fast sub-deviation. The weighted average of each fast sub-deviation is calculated using the length of the high mean deviation segment as the weight and is denoted as the fast group deviation. The fast group deviation ratio of the reference deviation sequence and the churn deviation sequence, the early decay ratio, and the dtw similarity are multiplied to obtain the churn pattern fitting quality.
[0020] The DTW similarity is calculated using the `tslearn.metrics.dtw()` function in Python, taking the reference bias sequence and the churn bias sequence as inputs. The DTW distance is then monotonically mapped to a similarity value within the range [0,1].
[0021] The principle behind obtaining the fitting quality of the loss mode in this method is actually a comprehensive comparison of early loss modes, the consistency of fast component fluctuation intensity and instability, and time-regular similarity. The calculation principle of the consistency of fast component fluctuation intensity and instability is that biochar composites are driven by rapid oxidation of surface oxygen-containing functional groups and transient microbial metabolism in the early stage of release, such as carboxyl and phenolic hydroxyl groups. At the same time, they are prone to super-average uprush clusters and local peaks due to diurnal temperature and humidity and air pressure fluctuations. If the material composition and mineral-char interface structure of two samples are similar, their fast component fluctuation intensity, duration and cluster morphology have comparable statistical regularities. Therefore, the consistency of fast component deviation in characterizing this instability effectively reflects the commonality of the two in early kinetics. The calculation principle of time-regular similarity is to quantify the matching degree of the two sequences in overall morphology and phase alignment, thereby fitting the natural law of the two-stage decay kinetic trajectory caused by the parallel reaction of fast and slow phases. Therefore, it can effectively refine the fitting degree of carbon fixation behavior mode and provide stable and accurate mathematical support for further anchoring the rapid assessment of the carbon fixation capacity of biochar composites.
[0022] Further, in step S400, the method for obtaining the fitting quality of the churn pattern by comparing the churn deviation with the benchmark database is as follows: all churn deviations obtained before the current time are constructed in chronological order to form a churn deviation sequence with a length of len; the churn deviation sequences of the reference samples stored in the benchmark database are each truncated with a prefix subsequence of length len as a reference deviation sequence; the reference deviation sequence and the churn deviation sequence are denoted as the fitting sequence. The fitted sequence is sliced into units of natural days, with each natural day serving as a recording point. The mean and range of the fitted sequence within the recording point are used to form state pairs. The state pairs corresponding to each recording point are used to form the fitted state sequence. Unsupervised clustering analysis is performed on all reference deviation sequences to obtain several clusters, which are denoted as state clusters. The Euclidean distance between the cluster center of the state cluster and the state pairs of the lost deviation sequence is denoted as the cluster distance. Unsupervised clustering analysis employs either the K-means algorithm or the Gaussian mixture model algorithm. Before clustering analysis, all mean data and all range data are normalized. The subsequent Euclidean distance must be calculated based on the normalized values.
[0023] The average fitted state sequence of state clusters is calculated, and its average value sequence is taken as the cluster reference sequence. The average value of each record point of the loss deviation sequence is used to construct a sequence as the cluster control sequence. The dtw similarity between the cluster reference sequence and the cluster control sequence is used as the cluster normalization degree. The calculation of the average fitted state sequence for state clusters refers to forming a set of fitted state sequences for each cluster. The average value of each sequence in this set is the average fitted state sequence. This process is the same as the calculation of the average vector. The average fitted state sequence includes two dimensions: the average value and the range. Therefore, only the average value dimension needs to be extracted to obtain the cluster baseline sequence. The cluster control sequence is constructed by using the average value of each record point in the churn deviation sequence as the cluster control sequence. This means that for the churn deviation sequence to be evaluated, the average value of the churn deviation at each record point is used to construct a sequence with record points as the unit. This ensures that the cluster control sequence and the cluster baseline sequence are effectively aligned in terms of record points and data classes.
[0024] Savitzky-Golay filtering is applied to the fitted sequence to obtain the filter curve. The first minimum value in the history of the filter curve is recorded as the early filter vortex point, and the average value of the fitted sequence is recorded as the filter mean. The first maximum value after the early filter vortex point in the history of the filter curve that is greater than the filter mean is recorded as the early filter peak point. The number of recorded points between the early filter vortex point and the early filter peak point is the vortex peak duration. The difference between the maximum and minimum values of each element in the fitted sequence between the early filter vortex point and the early filter peak point is the vortex peak difference. The ratio of the vortex peak duration to the vortex peak difference is the early overflow ratio. The lapse pattern fitting quality (LRSP) was calculated based on cluster spacing CDS, cluster centralization CDTW, and early overflow ratio REO. ; Where exp() is an exponential function with the natural constant e as the base, LOD is the length of the fitted state sequence, and err is a minimal positive number 10. -8 To prevent division by zero.
[0025] The smaller the cluster distance term in this model, the closer the diurnal mean and range are to the cluster center. This term is used to fit the macroscopic steady-state trend of carbon release from biochar complexes under diurnal temperature and humidity fluctuations, reflecting the system's stable response capability under diurnal perturbations. A smaller value indicates higher consistency between the sample's behavior and that of typical steady-state clusters, thus improving the steady-state fitting effect of carbon fixation pattern recognition. A higher time-warped similarity indicates better overall morphology and phase alignment, used to fit the temporal dynamics of the transition from fast component decay to slow component stability, reflecting the phase consistency of different samples during the fast and slow dual-stage decay process. Therefore, a larger value suggests that the sample's carbon fixation process is more consistent with the reference pattern in terms of temporal evolution path, thereby improving the dynamic matching accuracy of carbon fixation capacity assessment. A larger early spillover ratio indicates stronger early fast component peaks, used to fit the non-steady-state pattern of short-term carbon loss peaks caused by oxidation and microbial action of easily decomposable carbon in biochar complexes. A larger value indicates that the system is still in an unstable stage with high fluctuation amplitude, thus playing a role in identifying unstable samples and suppressing weights in carbon fixation capacity assessment.
[0026] In step S400, the calculation of the bleed pattern fitting quality can be based on the difference in data volume and signal stability. The fast group deviation method or the clustering method can be selected. When the monitoring period is short, usually less than 60 calendar days and the sample carbon release curve is still in the early fast group dominant stage, the sample size of the bleed deviation sequence is limited but the resolution of a single measurement is high. The fitting method based on early decay degree and fast group deviation is preferred because this method emphasizes the early identification of the two-stage decay dynamic characteristics, can quickly converge the conclusion within a limited observation period, and improve the timeliness of the evaluation.
[0027] When the monitoring period is long, usually more than 60 calendar days, or when there are many reference samples in the database and there are complex multi-peak fluctuation characteristics among the samples, it is preferable to use a fitting method based on state clustering and daily-scale clustering. This is because the method can separate environmental disturbances from the actual loss pattern in long-term monitoring through daily-scale normalization and unsupervised clustering, thereby improving the robustness of the fitting in the steady-state stage and the accuracy of long-term prediction.
[0028] Further, in step S500, the loss mode fitting quality is used as an evaluation criterion to select reference samples for evaluating carbon sequestration capacity. The method for outputting evaluation data is to sort the loss mode fitting quality of each reference sample from high to low according to the score; select samples with scores in the top several percentiles from the sorting results as a set of high-fit samples; the several percentiles are preferably the top 10%-20%; take the average value of the carbon stability score or long-term carbon retention rate in the set of high-fit samples to obtain the predicted value of carbon sequestration capacity of the current sample to be evaluated.
[0029] If the score distribution exhibits multiple peaks—meaning that the set of scores obtained after calculating the fitting quality of the loss patterns for all reference samples is not a single-peaked distribution—it is assumed that not all samples are clustered around an optimal interval, but rather that two or more local maxima exist. This suggests the existence of multiple sample groups with high scores but different characteristic patterns. In such cases, a Gaussian-weighted clustering weighted average method is used to calculate the predicted carbon sequestration capacity to avoid single-peak bias. Alternatively, in the clustering method, the cluster normalization degree can be weighted by adding a confidence level based on the number of samples within the cluster.
[0030] Preferably, all undefined variables in this invention, if not explicitly defined, can be manually set thresholds.
[0031] This invention also provides a rapid evaluation system for the carbon fixation capacity of biochar composite materials. The rapid evaluation system for the carbon fixation capacity of biochar composite materials includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps in the rapid evaluation method for the carbon fixation capacity of biochar composite materials. The rapid evaluation system for the carbon fixation capacity of biochar composite materials can run on computing devices such as desktop computers, laptops, handheld computers, and cloud data centers. The runnable system may include, but is not limited to, processors, memory, and server clusters. The processor executes the computer program within the following system units: The sample placement unit is used to apply the labeled mineral composite biochar in the field. The data acquisition unit is used to obtain the carbon content loss rate of composite biochar through high-frequency real-time monitoring. The loss calculation unit is used to obtain the loss deviation based on the carbon content loss rate between measurement intervals; The pattern fitting unit is used to obtain the churn pattern fitting quality by comparing it with the benchmark database based on the churn bias. The sample evaluation unit is used to select reference samples for evaluating carbon sequestration capacity by using the quality of the loss pattern fitting as an evaluation criterion, and outputs evaluation data.
[0032] The beneficial effects of this invention are as follows: This invention provides a rapid assessment method and system for the carbon sequestration capacity of biochar composite materials. By periodically acquiring the carbon content loss rate of the composite biochar and calculating the loss deviation, a quality data fitting of the loss pattern for each sample in the time series dimension is constructed. Using this quality data fitting, the relative stability of carbon sequestration capacity in biochar composite systems can be rapidly quantified under the premise of limited monitoring data flow, even when carbon release kinetics differ. This allows for a quantitative assessment of the greenhouse gas emission suppression effect of the currently studied composite samples, greatly improving the timeliness and reliability of carbon sequestration effect assessment. Furthermore, it provides a reliable observational basis for long-term carbon sequestration research of composite biochar materials under different soil and environmental conditions, reducing the sampling frequency and experimental costs required for traditional long-term buried sample analysis. Attached Figure Description
[0033] The above and other features of the present invention will become more apparent from the detailed description of the embodiments shown in conjunction with the accompanying drawings. In the accompanying drawings, the same reference numerals denote the same or similar elements. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without any creative effort. In the drawings: Figure 1The flowchart shown is a rapid evaluation method for the carbon fixation capacity of a biochar composite material. Figure 2 The diagram shows a structural diagram of a rapid carbon fixation capacity assessment system for a biochar composite material. Detailed Implementation
[0034] The following will provide a clear and complete description of the concept, specific structure, and technical effects of the present invention in conjunction with the embodiments and accompanying drawings, so as to fully understand the purpose, solution, and effects of the present invention. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.
[0035] like Figure 1 The diagram shows a flowchart of a rapid assessment method for the carbon fixation capacity of biochar composite materials. The following section will combine... Figure 1 This invention describes a method for rapidly evaluating the carbon fixation capacity of a biochar composite material according to an embodiment of the present invention. The method includes the following steps: S100 involves the field application of marked mineral-based biochar. S200 is used to obtain the carbon content loss rate of composite biochar through high-frequency real-time monitoring. S300, obtains the loss deviation based on the carbon content loss rate between measurement intervals; S400, the churn pattern fitting quality is obtained by comparing the churn deviation with the benchmark database; The S500 uses the quality of the loss pattern fitting as an evaluation criterion to select reference samples for assessing carbon sequestration capacity and outputs evaluation data.
[0036] Further, in step S100, the method for field application of the labeled mineral composite biochar is as follows: biochar labeled with 13C stable isotope is mixed evenly with minerals of high cation exchange capacity in a preset ratio to obtain a mineral composite biochar sample; the sample is evenly spread on the surface of the target test plot at a unit area application rate of 2-10 t / ha, and covered with soil 2-5 cm to ensure full contact with the native soil; background soil samples are collected before and after application, and initial soil samples are collected after application to establish a 13C content baseline.
[0037] 13C refers to carbon-13, a stable isotope of carbon. Additionally, soil pH, CEC, organic matter content, temperature, humidity, and meteorological conditions must be recorded, and this information must be linked to the sample number for subsequent carbon sequestration rate correction and database comparison.
[0038] The application rate per unit area needs to be preset according to the soil environment. Generally, when the soil organic matter content is less than 1.0% or the pH is greater than 7.5, the upper limit of application rate is 6 to 10 t / ha to enhance the interface between cation buffer and carbon adsorption. When the soil organic matter content is greater than 2.0% or it is neutral soil with a pH range of 6.5±0.3, the upper limit of application rate is 2 to 5 t / ha to avoid excessive concentration of surface carbon and oxidation deviation.
[0039] Regarding the setting of the soil cover thickness, if the surface wind speed or surface evaporation rate in the test area is high, a soil cover thickness of 4-5 cm should be selected; otherwise, a middle value or a lower value should be selected.
[0040] Further, in step S200, the method for obtaining the carbon content loss rate by high-frequency real-time monitoring of composite biochar is as follows: the preset measurement interval TG value range is 5-60 minutes; a flux chamber is set up above the release area, and gas samples are continuously collected at sampling intervals of 5-60 minutes; the CO2 volume fraction and 13C:12C isotope ratio of the gas sample are measured in real time using an isotope analyzer, and the 13CO2 flux generated by the composite biochar is calculated by inputting the isotope source model, and the 13CO2 flux change rate per unit time is used as the carbon content loss rate; the carbon content loss rate signal is corrected for drift due to temperature, humidity, air pressure and wind speed, and the carbon content loss rate sequence is output.
[0041] Before or during the online isotope monitoring in step S200, daily calibration and drift correction are performed using zero gas and two-point standard gas; the flux chamber is connected to the gas analysis unit, which includes an isotope analyzer that is capable of distinguishing between 13CO2 and 12CO2.
[0042] During the process of correcting the carbon loss rate signal for temperature, humidity, air pressure, and wind speed drift, the environmental parameters are obtained from a meteorological module deployed near the flux chamber, including temperature and humidity probes, a barometer, and an ultrasonic anemometer. Drift correction employs a joint compensation algorithm combining multiple regression and LOESS or a Kalman filter algorithm based on a state-space model to obtain the environmentally normalized carbon loss rate. The unit time for calculating the carbon loss rate is the TG value. The output carbon loss rate sequence is obtained once at each measurement interval. Using multiple flux chambers can improve the numerical accuracy of carbon loss rate and enhance the quality of subsequent assessments. When multiple flux chambers are used, the rates at each measurement point are spatially weighted and averaged to obtain the carbon loss rate in a representative area.
[0043] Further, in step S300, the method for obtaining the loss deviation based on the carbon content loss rate between measurement intervals is as follows: the content loss rate sequence obtained at the current time is denoted as clrLst, and clrLst(n1) represents its n1th element; the content loss rate sequence obtained at the previous time is denoted as la_clrLst, and la_clrLst(n2) represents its n2th element; the first loss rate is denoted as clrLst(1) - clrLst(2), and the second loss rate is denoted as clrLst(1) - la_clrLst(1); the loss deviation is obtained by dynamically weighting the first loss rate and the second loss rate through a state-space model based on Kalman filtering.
[0044] The first element of the content loss rate sequence is denoted as clrLst(1), which refers to the content loss rate measured at the current time. The second element of the content loss rate sequence is denoted as clrLst(2), which refers to the content loss rate measured at the previous time. The content loss rate sequence obtained at the previous time is denoted as la_clrLst, and the first element of the content loss rate measured at the previous time is la_clrLst(1). The implementation process of the dynamic weighted fusion is as follows: based on the currently obtained content loss rate sequence, a state-space model is established with drift and environmental disturbance as observation inputs. The system state and measurement error variance are estimated in real time through Kalman filtering, and the first loss rate and the second loss rate are dynamically weighted to generate loss deviation. The drift indicates the baseline shift of the instrument due to changes in temperature, light source or detector response during continuous monitoring. The environmental disturbance includes instantaneous interference caused by wind speed, air pressure, temperature and humidity on CO2 flux measurement.
[0045] Further, in step S400, the method for obtaining the fitting quality of the churn pattern by comparing the churn deviation with the benchmark database is as follows: all churn deviations obtained before the current time are constructed in chronological order to form a churn deviation sequence with a length of len; the churn deviation sequences of the reference samples stored in the benchmark database are each truncated with a prefix subsequence of length len as a reference deviation sequence; the reference deviation sequence and the churn deviation sequence are denoted as the fitting sequence; a time period DW is preset as an early window; the shape parameter obtained by performing Weibull least squares fitting on the first DW elements of the fitting sequence is denoted as the early decay degree. The shape parameters are the result parameters from the Weibull least squares fitting algorithm; A time period is defined as the dynamic window MW, with the dynamic window value ranging from 5 to 10 natural days. The exponential average of each element within the MW period before any time in the fitted sequence is taken as the dynamic mean. The difference between any element in the fitted sequence and its corresponding dynamic mean is taken as the dynamic mean deviation. All negative values in the dynamic mean deviation are updated to 0. Any number of consecutive and non-zero dynamic mean deviations corresponding to the time periods constitute a high mean deviation segment. The standard deviation of each element in the fitted sequence corresponding to the high mean deviation segment is denoted as the fast sub-deviation. The weighted average of each fast sub-deviation is calculated using the length of the high mean deviation segment as the weight and is denoted as the fast group deviation. The fast group deviation ratio of the reference deviation sequence and the churn deviation sequence, the early decay ratio, and the dtw similarity are multiplied to obtain the churn pattern fitting quality.
[0046] The DTW similarity is calculated using the `tslearn.metrics.dtw()` function in Python, taking the reference bias sequence and the churn bias sequence as inputs. The DTW distance is then monotonically mapped to a similarity value within the range [0,1].
[0047] Further, in step S400, the method for obtaining the fitting quality of the churn pattern by comparing the churn deviation with the benchmark database is as follows: all churn deviations obtained before the current time are constructed in chronological order to form a churn deviation sequence with a length of len; the churn deviation sequences of the reference samples stored in the benchmark database are each truncated with a prefix subsequence of length len as a reference deviation sequence; the reference deviation sequence and the churn deviation sequence are denoted as the fitting sequence. The fitted sequence is sliced into units of natural days, with each natural day serving as a recording point. The mean and range of the fitted sequence within the recording point are used to form state pairs. The state pairs corresponding to each recording point are used to form the fitted state sequence. Unsupervised clustering analysis is performed on all reference deviation sequences to obtain several clusters, which are denoted as state clusters. The Euclidean distance between the cluster center of the state cluster and the state pairs of the lost deviation sequence is denoted as the cluster distance. Unsupervised clustering analysis employs either the K-means algorithm or the Gaussian mixture model algorithm. Before clustering analysis, all mean data and all range data are normalized. The subsequent Euclidean distance must be calculated based on the normalized values.
[0048] The average fitted state sequence of state clusters is calculated, and its average value sequence is taken as the cluster reference sequence. The average value of each record point of the loss deviation sequence is used to construct a sequence as the cluster control sequence. The dtw similarity between the cluster reference sequence and the cluster control sequence is used as the cluster normalization degree. The calculation of the average fitted state sequence for state clusters refers to forming a set of fitted state sequences for each cluster. The average value of each sequence in this set is the average fitted state sequence. This process is the same as the calculation of the average vector. The average fitted state sequence includes two dimensions: the average value and the range. Therefore, only the average value dimension needs to be extracted to obtain the cluster baseline sequence. The cluster control sequence is constructed by using the average value of each record point in the churn deviation sequence as the cluster control sequence. This means that for the churn deviation sequence to be evaluated, the average value of the churn deviation at each record point is used to construct a sequence with record points as the unit. This ensures that the cluster control sequence and the cluster baseline sequence are effectively aligned in terms of record points and data classes.
[0049] Savitzky-Golay filtering is applied to the fitted sequence to obtain the filter curve. The first minimum value in the history of the filter curve is recorded as the early filter vortex point, and the average value of the fitted sequence is recorded as the filter mean. The first maximum value after the early filter vortex point in the history of the filter curve that is greater than the filter mean is recorded as the early filter peak point. The number of recorded points between the early filter vortex point and the early filter peak point is the vortex peak duration. The difference between the maximum and minimum values of each element in the fitted sequence between the early filter vortex point and the early filter peak point is the vortex peak difference. The ratio of the vortex peak duration to the vortex peak difference is the early overflow ratio. The lapse pattern fitting quality (LRSP) was calculated based on cluster spacing CDS, cluster centralization CDTW, and early overflow ratio REO. ; Where exp() is an exponential function with the natural constant e as the base, LOD is the length of the fitted state sequence, and err is a minimal positive number 10. -8 To prevent division by zero.
[0050] Further, in step S500, the loss mode fitting quality is used as an evaluation criterion to select reference samples for evaluating carbon sequestration capacity. The method for outputting evaluation data is to sort the loss mode fitting quality of each reference sample from high to low according to the score; select samples with scores in the top several percentiles from the sorting results as a set of high-fit samples; the several percentiles are preferably the top 10%-20%; take the average value of the carbon stability score or long-term carbon retention rate in the set of high-fit samples to obtain the predicted value of carbon sequestration capacity of the current sample to be evaluated.
[0051] If the score distribution exhibits multiple peaks—meaning that the set of scores obtained after calculating the fitting quality of the loss patterns for all reference samples is not a single-peaked distribution—it is assumed that not all samples are clustered around an optimal interval, but rather that two or more local maxima exist. This suggests the existence of multiple sample groups with high scores but different characteristic patterns. In such cases, a Gaussian-weighted clustering weighted average method is used to calculate the predicted carbon sequestration capacity to avoid single-peak bias. Alternatively, in the clustering method, the cluster normalization degree can be weighted by adding a confidence level based on the number of samples within the cluster.
[0052] An embodiment of the present invention provides a rapid evaluation system for the carbon fixation capacity of biochar composite materials, such as... Figure 2 The diagram shows a structural diagram of a rapid carbon fixation capacity assessment system for biochar composite materials according to the present invention. This embodiment of the rapid carbon fixation capacity assessment system for biochar composite materials includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps in the above embodiment of the rapid carbon fixation capacity assessment method for biochar composite materials.
[0053] The system includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program in units of the following system: The sample placement unit is used to apply the labeled mineral composite biochar in the field. The data acquisition unit is used to obtain the carbon content loss rate of composite biochar through high-frequency real-time monitoring. The loss calculation unit is used to obtain the loss deviation based on the carbon content loss rate between measurement intervals; The pattern fitting unit is used to obtain the churn pattern fitting quality by comparing it with the benchmark database based on the churn bias. The sample evaluation unit is used to select reference samples for evaluating carbon sequestration capacity by using the quality of the loss pattern fitting as an evaluation criterion, and outputs evaluation data.
[0054] The rapid carbon fixation capacity assessment system for biochar composite materials described above can run on computing devices such as desktop computers, laptops, handheld computers, and cloud servers. The system that can run on the rapid carbon fixation capacity assessment system for biochar composite materials may include, but is not limited to, processors and memory. Those skilled in the art will understand that the above example is merely an illustration of a rapid carbon fixation capacity assessment system for biochar composite materials and does not constitute a limitation on such a system. It may include more or fewer components, or a combination of certain components, or different components. For example, the rapid carbon fixation capacity assessment system for biochar composite materials may also include input / output devices, network access devices, buses, etc.
[0055] The processor referred to can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the rapid carbon fixation capacity assessment system for biochar composite materials, connecting various parts of the system via various interfaces and lines.
[0056] The memory can be used to store the computer program and / or modules. The processor, by running or executing the computer program and / or modules stored in the memory and calling the data stored in the memory, realizes various functions of the rapid carbon fixation capacity assessment system for the biochar composite material. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created based on the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.
[0057] Although the invention has been described in considerable detail and particularly with regard to several of the described embodiments, it is not intended to limit itself to any of these details or embodiments or any particular embodiment, thereby effectively covering the intended scope of the invention. Furthermore, the invention has been described above with respect to embodiments foreseeable by the inventors in order to provide a useful description, and non-substantial modifications to the invention that have not yet been foreseen may still represent equivalent modifications.
Claims
1. A rapid evaluation method for the carbon fixation capacity of a biochar composite material, characterized in that, The method includes the following steps: S100 involves the field application of marked mineral-based biochar. S200 is used to obtain the carbon content loss rate of composite biochar through high-frequency real-time monitoring. S300, obtains the loss deviation based on the carbon content loss rate between measurement intervals; S400, the churn pattern fitting quality is obtained by comparing the churn deviation with the benchmark database; S500 uses the quality of the loss pattern fitting as an evaluation criterion to select reference samples for evaluating carbon sequestration capacity and outputs evaluation data. The S400 process involves: acquiring the loss deviation sequence and constructing a reference deviation sequence to form a fitted sequence for comparison; calculating the early decay rate in the fitted sequence; obtaining the dynamic mean by calculating the exponential average based on the fitted sequence; calculating the dynamic mean difference by the difference between each element in the fitted sequence and its corresponding dynamic mean; identifying high mean difference segments based on the non-zero continuous segments of the dynamic mean difference; calculating the fast group deviation by the standard deviation of each element in the high mean difference segment; and weighting each fast group deviation with the length of the high mean difference segment as the weight to obtain the fast group deviation. The fitting quality of the churn pattern is obtained by multiplying the fast group deviation, early decay degree, and dtw similarity.
2. The method for rapidly evaluating the carbon fixation capacity of a biochar composite material according to claim 1, characterized in that, In step S100, the method for field application of the labeled mineral composite biochar is as follows: biochar labeled with 13C stable isotope is mixed evenly with minerals with high cation exchange capacity in a preset ratio to obtain a mineral composite biochar sample; the sample is evenly spread on the surface of the target test plot at a unit area application rate of 2-10 t / ha, and covered with soil 2-5 cm to ensure full contact with the native soil; background soil samples are collected before and after application, and initial soil samples are collected after application to establish a baseline of 13C content.
3. The method for rapidly evaluating the carbon fixation capacity of a biochar composite material according to claim 1, characterized in that, In step S200, the method for obtaining the carbon loss rate of composite biochar through high-frequency real-time monitoring is as follows: the preset measurement interval TG value range is 5-60 minutes; a flux chamber is set up above the release area, and gas samples are continuously collected at sampling intervals of 5-60 minutes; the CO2 volume fraction and 13C:12C isotope ratio of the gas samples are measured in real time using an isotope analyzer, and the 13CO2 flux generated by the composite biochar is calculated by inputting the isotope source model, and the 13CO2 flux change rate per unit time is used as the carbon loss rate; the carbon loss rate signal is corrected for drift due to temperature, humidity, air pressure and wind speed, and the carbon loss rate sequence is output.
4. The method for rapidly evaluating the carbon fixation capacity of a biochar composite material according to claim 1, characterized in that, In step S300, the method for obtaining the loss deviation based on the carbon content loss rate between measurement intervals is as follows: the content loss rate sequence obtained at the current time is denoted as clrLst, and clrLst(n1) represents its n1th element; The content loss rate sequence obtained at the previous moment is denoted as la_clrLst, and la_clrLst(n2) represents its n2th element; Let the first churn rate be clrLst(1) - clrLst(2) and the second churn rate be clrLst(1) - la_clrLst(1). The churn deviation is obtained by dynamically weighting the first churn rate and the second churn rate using a state-space model based on Kalman filtering.
5. The method for rapidly evaluating the carbon fixation capacity of a biochar composite material according to claim 1, characterized in that, In step S400, the method for obtaining the fitting quality of the bleed pattern by comparing the bleed deviation with the benchmark database is as follows: All bleed deviations obtained before the current time are constructed in chronological order to form a bleed deviation sequence with a length of len; the bleed deviation sequences of the reference samples stored in the benchmark database are each truncated with a prefix subsequence of equal length len as a reference deviation sequence; the reference deviation sequence and the bleed deviation sequence are denoted as the fitting sequence; a time period DW is preset as an early window; the shape parameter obtained by performing Weibull least squares fitting on the first DW elements of the fitting sequence is denoted as the early decay rate. A time period is defined as the dynamic window MW. The exponential average of each element within the MW period before any time in the fitted sequence is taken as the dynamic mean. The difference between any element in the fitted sequence and its corresponding dynamic mean is taken as the dynamic mean deviation. All negative values in the dynamic mean deviation are updated to 0. Any number of consecutive and non-zero dynamic mean deviations corresponding to the time periods constitute a high mean deviation segment. The standard deviation of each element in the fitted sequence corresponding to the high mean deviation segment is denoted as the fast sub-deviation. The weighted average of each fast sub-deviation is calculated using the length of the high mean deviation segment as the weight and is denoted as the fast group deviation. The fast group deviation ratio of the reference deviation sequence and the churn deviation sequence, the early decay ratio, and the dtw similarity are multiplied to obtain the churn pattern fitting quality.
6. The method for rapidly evaluating the carbon fixation capacity of a biochar composite material according to claim 1, characterized in that, In step S400, the method for obtaining the fitting quality of the churn pattern by comparing the churn deviation with the benchmark database is as follows: all churn deviations obtained before the current time are constructed in chronological order to form a churn deviation sequence with a length of len; the churn deviation sequences of the reference samples stored in the benchmark database are each truncated with a prefix subsequence of length len as a reference deviation sequence; the reference deviation sequence and the churn deviation sequence are denoted as the fitting sequence. The fitted sequence is sliced into units of natural days, with each natural day serving as a recording point. The mean and range of the fitted sequence within the recording point are used to form state pairs. The state pairs corresponding to each recording point are used to form the fitted state sequence. Unsupervised clustering analysis is performed on all reference deviation sequences to obtain several clusters, which are denoted as state clusters. The Euclidean distance between the cluster center of the state cluster and the state pairs of the lost deviation sequence is denoted as the cluster distance. The average fitted state sequence of state clusters is calculated, and its average value sequence is taken as the cluster reference sequence. The average value of each record point of the loss deviation sequence is used to construct a sequence as the cluster control sequence. The dtw similarity between the cluster reference sequence and the cluster control sequence is used as the cluster normalization degree. Savitzky-Golay filtering is applied to the fitted sequence to obtain the filtered curve. The first minimum value in the history of the filtered curve is identified and recorded as the early filtered vortex point. The average value of the fitted sequence is recorded as the filtered mean. The first maximum value after the early filtered vortex point in the history of the filtered curve that is greater than the filtered mean is identified and recorded as the early filtered peak point. The number of recorded points between the early filtered vortex point and the early filtered peak point is the vortex peak duration. The difference between the maximum and minimum values of each element in the fitted sequence between the early filtered vortex point and the early filtered peak point is the vortex peak difference. The ratio of the vortex peak duration to the vortex peak difference is the early overflow ratio. The fitting quality of the bleed pattern is calculated based on the cluster spacing, cluster normalization, and early overflow ratio.
7. The method for rapidly evaluating the carbon fixation capacity of a biochar composite material according to claim 1, characterized in that, In step S500, the loss pattern fitting quality is used as the evaluation criterion to select reference samples for evaluating carbon sequestration capacity. The method for outputting evaluation data is to sort the loss pattern fitting quality of each reference sample from high to low according to the score; select samples with scores in the top several percentiles from the sorting results as a set of high-fit samples; the top several percentiles are preferably the top 10%-20%; take the average value of the carbon stability score or long-term carbon retention rate in the set of high-fit samples to obtain the predicted value of carbon sequestration capacity of the current sample to be evaluated.
8. A rapid evaluation system for the carbon fixation capacity of biochar composite materials, characterized in that, The rapid carbon fixation capacity assessment system for a biochar composite material includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps in the rapid carbon fixation capacity assessment method for a biochar composite material according to any one of claims 1-7. The rapid carbon fixation capacity assessment system for a biochar composite material runs on a desktop computer, a laptop computer, a handheld computer, or a computing device in a cloud data center.