Farmland carbon-water coupling process multi-parameter real-time detection method and system

Abstract

The invention provides a farmland carbon-water coupling process multi-parameter real-time detection method and system, and relates to the technical field of agricultural ecological monitoring. According to the method, a multi-parameter monitoring unit is arranged in a test area to obtain soil CO2 flux, soil moisture, soil temperature and micrometeorological data, a synchronous multi-source real-time sequence is constructed, and quality controlled data is formed through abnormity elimination and missing measurement supplementation. Carbon flux and moisture circulation characteristics are extracted based on a sliding time window, a carbon-water response coefficient and lag time are calculated, a carbon-water coupling index is constructed, and real-time identification of strong coupling, weak coupling and decoupling regions is realized. The method realizes continuous, quantitative and real-time monitoring of the farmland carbon-water process.

Description

Technical Field

[0001] This invention relates to the field of agricultural ecological monitoring technology, and in particular to a method and system for real-time detection of multiple parameters in the carbon-water coupling process in farmland. Background Technology

[0002] Against the backdrop of global climate change and the "dual carbon" goal, farmland ecosystems, as important terrestrial carbon sinks, are receiving increasing attention due to their soil carbon sequestration capacity and responses to water cycle factors such as precipitation patterns and evapotranspiration. Existing research utilizes static box methods and eddy covariance systems to monitor farmland CO2 flux, combined with soil moisture, temperature, and meteorological observations, to characterize farmland carbon budgets to some extent, providing support for assessing farmland carbon sink functions. Simultaneously, with the development of the Internet of Things (IoT) and low-power sensors, soil moisture and meteorological elements in farmland environments can be automatically collected at higher frequencies, providing a data foundation for analyzing farmland water cycle processes.

[0003] Current technological trends are based on multi-source monitoring data, expanding carbon and water cycles from single-factor observation to multi-factor collaborative observation. This involves employing data fusion and model analysis methods to reveal the coupling relationships between farmland carbon and water processes at both time-series and spatial scales. On one hand, multiple types of sensors, including those for soil moisture, temperature, and CO2 flux, are deployed at the field scale, combined with weather towers and precipitation monitoring to achieve continuous monitoring. On the other hand, methods such as sliding time window analysis, correlation analysis, and structural equation modeling are introduced to explore the impact mechanisms of precipitation events, soil wetting-drying cycles, and irrigation management on soil carbon sequestration processes, aiming to provide quantitative evidence for optimizing farmland water and fertilizer management and enhancing carbon sequestration.

[0004] However, existing technologies generally suffer from the following shortcomings: First, most monitoring systems still set up carbon flux observation and water parameter observation independently, only performing univariate statistics on carbon flux or water status. They lack a multi-parameter real-time detection framework for the "carbon-water coupling process" under a unified time benchmark and spatial unit, making it difficult to identify the dynamic response of carbon sequestration processes to instantaneous precipitation or short-term drought in a timely manner. Second, even with multi-source observation deployments, correlation analysis is often only performed on long-term statistical scales. Coupled characteristic parameters such as response coefficients and lag times within a sliding time window are not constructed for the non-stationarity of the carbon-water relationship in farmland, resulting in insufficient characterization of the changes in coupling intensity under different farming practices and extreme climate conditions. Third, existing systems place more emphasis on the real-time display and early warning of single variables, lacking the ability to construct and spatiotemporally diagnose the "carbon-water coupling index," and cannot directly reveal the sensitive areas, decoupling areas, and response mechanisms of soil carbon sequestration processes to water cycle disturbances. Therefore, it is necessary to propose a real-time multi-parameter detection method for carbon-water coupling processes in farmland by uniformly deploying a multi-parameter sensor network, constructing carbon-water coupling characteristic parameters, and realizing real-time detection and spatiotemporal diagnosis of carbon-water coupling processes, so as to make up for the shortcomings of existing technologies in quantitative characterization and real-time identification of coupling relationships. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, the purpose of this invention is to provide a method and system for real-time detection of multiple parameters in farmland carbon-water coupling processes. By constructing a data quality control system with multi-parameter collaborative monitoring and time-series unified monitoring, and a carbon-water coupling index based on response coefficient and hysteresis identification, the invention achieves continuous, quantitative, and real-time accurate detection of farmland carbon-water processes, which is impossible with existing technologies.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] A method for real-time detection of multiple parameters in a farmland carbon-water coupling process includes:

[0008] Multi-parameter monitoring units capable of simultaneously acquiring soil CO2 flux, soil moisture content, soil temperature, and canopy micrometeorological elements were set up in the farmland experimental area to continuously acquire raw monitoring data of carbon-water processes.

[0009] The original monitoring data were collected synchronously at the same time interval, and the soil CO2 flux monitoring data, soil moisture monitoring data and micrometeorological monitoring data were matched to the same time sequence to form a multi-source real-time data sequence.

[0010] Based on the multi-source real-time data sequence, abnormal data are removed and short-term missing data are supplemented to form a quality-controlled data sequence for carbon-water coupling analysis.

[0011] Within a sliding time window, the soil CO2 flux change rate and daily cumulative CO2 flux are calculated based on the quality-controlled data sequence to form carbon flux characteristic parameters.

[0012] Within the sliding time window, the change in soil moisture storage and evapotranspiration intensity are calculated based on the quality-controlled data sequence to form water cycle characteristic parameters.

[0013] Based on the carbon flux characteristic parameters and the water cycle characteristic parameters, the carbon-water response coefficient and response lag time are calculated, a carbon-water coupling index is constructed, and strong coupling region, weak coupling region and decoupling region are identified based on the changes of the carbon-water coupling index within a continuous time window, so as to realize the real-time detection of the carbon-water coupling process in farmland.

[0014] Preferably, a multi-parameter monitoring unit capable of simultaneously acquiring soil CO2 flux, soil moisture content, soil temperature, and canopy micrometeorological elements is set up in the farmland experimental area to continuously acquire raw monitoring data of carbon-water processes, including:

[0015] Based on the differences in soil physicochemical properties and the structure of the vegetation canopy, the deployment location and vertical sensing depth of the multi-parameter monitoring unit in the farmland experimental area are determined so that the multi-parameter monitoring unit can simultaneously cover the environmental gradient of the soil surface layer, the root zone layer and the canopy air layer.

[0016] The multi-parameter monitoring unit is equipped with multi-source sensing components for characterizing carbon flux, water dynamics and canopy micrometeorological elements, and a collaborative sensing system constrained by unified spatial positioning and continuous physical response is established.

[0017] Background correction and dimensional normalization are performed on the output signal of the multi-source sensing component to obtain the raw monitoring data that can reflect the processes of soil respiration, water migration and canopy exchange.

[0018] Preferably, the original monitoring data are collected synchronously at the same time intervals, and the soil CO2 flux monitoring data, soil moisture monitoring data, and micrometeorological monitoring data are correlated to the same time series to form a multi-source real-time data sequence, including:

[0019] The coordinated acquisition operation of the soil CO2 flux monitoring unit, soil moisture monitoring unit and micrometeorological monitoring unit in the multi-parameter monitoring unit is triggered according to the preset unified sampling time interval, so that different monitoring quantities generate corresponding data at the same acquisition time.

[0020] The soil CO2 flux monitoring data, soil moisture monitoring data and micrometeorological monitoring data acquired at each collection time will be indexed and associated according to the collection time to generate synchronous time-series records corresponding to multiple parameters;

[0021] The multi-source real-time data sequence used to characterize the dynamic changes of the carbon-water process is constructed based on the synchronous time-series records.

[0022] Preferably, based on the multi-source real-time data sequence, outlier data is removed and short-term missing data is supplemented to form a quality-controlled data sequence for carbon-water coupling analysis, including:

[0023] Consistency checks and fluctuation constraint analyses were performed on the soil CO2 flux monitoring data, soil moisture monitoring data, and micrometeorological monitoring data in the multi-source real-time data sequence to identify abnormal data that deviated from the normal variation pattern.

[0024] Abnormal data that has been identified will be removed, and supplementary data will be generated based on the changing trends of adjacent valid data and the collaborative response relationship of similar monitoring quantities during the missing data period.

[0025] The monitoring data, after removing anomalies and filling in missing data, are reconstructed into a continuous quality-controlled data sequence.

[0026] Preferably, the formula for generating the supplementary data is:

[0027] ;

[0028] in, For the time during the missing test period Target monitoring quantity The generated padding data; Based on the target monitoring quantity The trend estimate is obtained by linear interpolation of observations at adjacent valid times before and after the missing period; For target monitoring quantity The long-term average during the training or benchmark statistics phase; For target monitoring quantity The standard deviation of the stage; In order to match the target monitoring quantity Similar monitoring quantities exhibiting a synergistic response relationship between carbon and water processes At any moment Observed values; For the same type of monitoring volume The long-term average value during the aforementioned period; For the same type of monitoring volume The standard deviation of the stage; In order to match the target monitoring quantity A set of similar monitoring quantities that exhibit significant synergistic response relationships; The target monitoring quantity is calculated based on historical quality-controlled data sequences during the training phase. Compared with similar monitoring volume The correlation coefficient between them; To revolve around time Target monitoring volume The length of the corresponding missing measurement interval; The time scale coefficient is used to control the relative contribution of the trend term and the collaborative response term.

[0029] Preferably, within the sliding time window, the change in soil moisture storage and evapotranspiration intensity are calculated based on the quality-controlled data sequence to form water cycle characteristic parameters, including:

[0030] Based on the aforementioned quality-controlled data sequence, the root zone soil volumetric water content is discretized and accumulated within a sliding time window, with the calculation time... Soil moisture storage And obtain the starting time of the sliding time window. Changes in soil moisture storage as a baseline The soil moisture storage satisfies: ;in, For a moment Soil moisture storage; The start time of the sliding time window Changes in water storage; For a moment No. The quality of each soil layer is controlled by its volumetric moisture content; For the first Each soil layer thickness; The number of discrete soil layers; The length of the sliding time window;

[0031] The average evapotranspiration intensity of the sliding time window is calculated based on the instantaneous evapotranspiration flux time series within the sliding time window. ,satisfy: ;in, The average evapotranspiration intensity corresponding to the sliding time window; For a moment Instantaneous evapotranspiration flux; The sampling time interval; This represents the number of sampling moments within the sliding time window;

[0032] The change in soil moisture storage With the evaporation intensity By combining these parameters, we can construct characteristic parameters of the water cycle that reflect the combined driving mechanism of root zone water consumption and atmospheric evapotranspiration.

[0033] Preferably, based on the carbon flux characteristic parameters and the water cycle characteristic parameters, the carbon-water response coefficient and response lag time are calculated to construct a carbon-water coupling index. Strong coupling regions, weak coupling regions, and decoupling regions are identified based on the changes in the carbon-water coupling index within a continuous time window, enabling real-time detection of the carbon-water coupling process in farmland, including:

[0034] Within the sliding time window, the carbon flux characteristic parameter sequence is time-series paired with the water cycle characteristic parameter sequence within the same sliding time window, and the starting time of each sliding time window is... Extraction length is carbon flux characteristic subsequence With water cycle characteristic subsequence We constructed a joint feature sequence to characterize the synergistic relationship between changes in carbon flux and water cycle status.

[0035] Based on the joint feature sequence, the carbon-water response coefficient and the carbon-water response lag time are calculated within each sliding time window, wherein the carbon-water response coefficient... satisfy: Furthermore, the carbon-water response lag time satisfy: ;in, The start time of the sliding time window The corresponding carbon-water response coefficient; The carbon-water response lag time within the sliding time window; For the first time window within the sliding time window The quality of the carbon flux characteristic parameters at each sampling time is controlled; For the first time window within the sliding time window The quality of the water cycle characteristic parameters at each sampling time is controlled; The carbon flux feature subsequence within the current sliding time window The average value; The water cycle feature subsequence within the current sliding time window The average value; The carbon flux feature subsequence within the current sliding time window Standard deviation; The water cycle feature subsequence within the current sliding time window Standard deviation; This represents the number of sampling moments included within the sliding time window; This is the discrete-time offset used to characterize carbon flux features relative to water cycle features; This represents the maximum number of steps that can be taken in a time-lapse search.

[0036] A carbon-water coupling index is constructed using the carbon-water response coefficient and the carbon-water response lag time as inputs. The carbon-water coupling process in farmland is then classified and identified based on the changes in the carbon-water coupling index within a continuous time window. satisfy: ;in, The start time of the sliding time window The corresponding carbon-water coupling index; The length of the sliding time window; The absolute value of the carbon-water response coefficient is used to characterize the correlation strength between carbon flux characteristics and water cycle characteristics. The absolute value of the carbon-water response lag time is used to characterize the degree of time mismatch in carbon-water interaction. Within a continuous time window, when the carbon-water coupling index remains at a stable high level, it is identified as a strong coupling region; when the carbon-water coupling index is at a medium level and fluctuates greatly, it is identified as a weak coupling region; and when the carbon-water coupling index is continuously lower than a preset threshold, it is identified as a decoupling region, thereby realizing real-time detection of the carbon-water coupling process in farmland.

[0037] A real-time multi-parameter monitoring system for carbon-water coupling processes in farmland includes:

[0038] The multi-parameter monitoring unit deployment module is used to set up multi-parameter monitoring units in farmland experimental areas that can simultaneously acquire soil CO2 flux, soil moisture content, soil temperature and canopy micrometeorological elements, so as to continuously acquire raw monitoring data of carbon-water processes.

[0039] The multi-source data synchronous acquisition module is used to synchronously acquire the original monitoring data at the same time interval, and to correspond the soil CO2 flux monitoring data, soil moisture monitoring data and micrometeorological monitoring data to the same time sequence to form a multi-source real-time data sequence.

[0040] The data quality control module is used to remove abnormal data and supplement short-term missing data based on the multi-source real-time data sequence to form a quality-controlled data sequence for carbon-water coupling analysis.

[0041] The carbon flux characteristic calculation module is used to calculate the soil CO2 flux change rate and daily cumulative CO2 flux based on the quality-controlled data sequence within a sliding time window, thereby forming carbon flux characteristic parameters.

[0042] The water cycle characteristic calculation module is used to calculate the change in soil moisture storage and evapotranspiration intensity based on the quality-controlled data sequence within the sliding time window, thereby forming water cycle characteristic parameters.

[0043] The carbon-water coupling analysis and segment identification module is used to calculate the carbon-water response coefficient and response lag time based on the carbon flux characteristic parameters and the water cycle characteristic parameters, construct the carbon-water coupling index, and identify strong coupling zone, weak coupling zone and decoupling zone based on the change of the carbon-water coupling index within a continuous time window, so as to realize the real-time detection of the carbon-water coupling process in farmland.

[0044] The present invention discloses the following technical effects:

[0045] This invention overcomes the spatial inconsistencies and temporal asynchrony caused by the dispersed monitoring elements and independent instrument deployment in existing technologies by deploying multi-parameter monitoring units in farmland experimental areas that can simultaneously sense soil CO2 flux, soil moisture content, soil temperature, and canopy microclimate. This allows carbon flux and water cycle processes to be continuously collected in the same observation scenario and time frame, significantly improving the integrity, reliability, and cross-scale applicability of farmland carbon-water process monitoring data.

[0046] This invention constructs a multi-source real-time data sequence by unifying the sampling time interval and the same time-series framework, and further forms a quality-controlled data sequence by eliminating abnormal data and supplementing short-term missing measurements. This effectively solves the problems of inconsistent sampling frequencies of multi-source sensors, data discontinuity, and unusable missing measurements in the background technology, and ensures that the basic data on which the subsequent carbon-water coupling analysis depends has high continuity, high integrity and traceability.

[0047] This invention constructs a dynamic characteristic parameter system that reflects carbon release, carbon absorption, water consumption, and water replenishment by calculating the rate of change of soil CO2 flux, daily cumulative CO2 flux, soil moisture storage changes, and evapotranspiration intensity within a sliding time window. This system overcomes the shortcomings of existing technologies that can only perform static measurements or low-frequency analyses of carbon and water processes, and realizes the capture of short-timescale dynamic responses of carbon-water processes in farmland, providing a key quantitative basis for studying the carbon-water coupling mechanism.

[0048] Based on the aforementioned carbon flux characteristic parameters and water cycle characteristic parameters, this invention proposes a method for calculating the carbon-water response coefficient and response lag time. This method overcomes the shortcomings of traditional research, such as the lack of unified response indicators, difficulty in quantifying the strength of coupling and time lag, and enables the interaction between carbon and water processes to be expressed in a quantifiable and comparable manner. This reveals the spatiotemporal linkage mechanism between soil respiration, water supply and atmospheric exchange.

[0049] This invention overcomes the limitations of existing technologies that rely on human experience and lack real-time diagnostic capabilities by constructing a carbon-water coupling index and automatically identifying strong coupling, weak coupling, and decoupling regions based on index changes within a continuous time window. This enables the accurate, real-time, dynamic, and automated detection of the carbon-water coupling process in farmland, providing timely and valuable decision-making basis for farmland ecological management, precision irrigation control, and carbon sequestration and enhancement strategies. Attached Figure Description

[0050] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 A flowchart of the method provided in an embodiment of the present invention;

[0052] Figure 2 This is a schematic diagram of the system structure provided in an embodiment of the present invention. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] The purpose of this invention is to provide a method and system for real-time detection of multiple parameters in the carbon-water coupling process in farmland. By integrating multi-source monitoring, dynamic feature extraction and automatic identification of coupling sections, the accuracy, completeness and real-time performance of the characterization of the carbon-water process in farmland are significantly improved, overcoming the technical bottleneck of traditional methods that rely on a single parameter and cannot quantify the strength of coupling.

[0055] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] Figure 1 The method flowchart provided in the embodiments of the present invention is as follows: Figure 1 As shown, this invention provides a method for real-time detection of multiple parameters in a farmland carbon-water coupling process, comprising:

[0057] Step 100: Set up a multi-parameter monitoring unit in the farmland experimental area that can simultaneously acquire soil CO2 flux, soil moisture content, soil temperature and canopy micrometeorological elements, so as to continuously acquire raw monitoring data of carbon-water processes.

[0058] Step 200: Collect raw monitoring data synchronously at the same time interval, and match the soil CO2 flux monitoring data, soil moisture monitoring data and micrometeorological monitoring data to the same time series to form a multi-source real-time data sequence;

[0059] Step 300: Based on the multi-source real-time data sequence, remove abnormal data and fill in short-term missing data to form a quality-controlled data sequence for carbon-water coupling analysis;

[0060] Step 400: Within the sliding time window, calculate the soil CO2 flux change rate and daily cumulative CO2 flux based on the quality-controlled data sequence to form carbon flux characteristic parameters;

[0061] Step 500: Within the sliding time window, calculate the change in soil moisture storage and evapotranspiration intensity based on the quality-controlled data sequence to form water cycle characteristic parameters;

[0062] Step 600: Based on the carbon flux characteristic parameters and water cycle characteristic parameters, calculate the carbon-water response coefficient and response lag time, construct the carbon-water coupling index, and identify the strong coupling region, weak coupling region and decoupling region based on the change of the carbon-water coupling index within a continuous time window, so as to realize the real-time detection of the carbon-water coupling process in farmland.

[0063] Specifically, in step 100 of this embodiment, representative experimental plots are selected in a typical farmland experimental area. Based on the physicochemical properties of soil texture, organic matter content, bulk density, etc., and canopy structure characteristics such as crop row spacing, plant height, and leaf area index, the planar layout location and vertical sensing depth of the multi-parameter monitoring unit are determined. The "multi-parameter monitoring unit" mentioned in this embodiment refers to a comprehensive monitoring unit that integrates soil CO2 flux sensing components, soil moisture sensing components, soil temperature sensing components, and canopy micro-meteorological sensing components in a similar spatial location. This monitoring unit covers the environmental gradient of the soil surface layer, root zone layer, and canopy air layer in the vertical direction. For example, in a test plot with a long side of 100m and a short side of 50m, a monitoring unit is set up every 20m to 30m along the crop row. The soil monitoring depth is preferably set at three levels: 5cm, 20cm and 40cm. The canopy micro-meteorological sensing height is preferably set in the range of 1.5m to 2m to ensure that the monitoring unit can simultaneously capture the carbon-water process characteristics of the ground surface, root zone and above the canopy.

[0064] This embodiment configures multi-source sensing components in the multi-parameter monitoring unit to characterize carbon flux, water dynamics, and canopy micrometeorological elements, and establishes a collaborative sensing system constrained by unified spatial positioning and continuous physical response. The "multi-source sensing components" in this embodiment refer to a collection of multiple sensing components set up within the same monitoring unit for different physical quantities, including flux measurement components for acquiring soil CO2 flux, water sensing components for acquiring soil moisture content, temperature sensing components for acquiring soil temperature, and meteorological sensing components for acquiring canopy micrometeorological elements such as wind speed, air temperature, relative humidity, and net radiation. The "collaborative sensing system" in this embodiment refers to unifying the spatial identifier of each monitoring unit, unifying the vertical depth calibration method, and unifying the sampling triggering benchmark, so that different sensing components generate comparable physical response signals at the same location and under the same monitoring rhythm. For example, the spatial location of each monitoring unit is uniquely marked with plot coordinates and row / column numbers, the depth stratification scheme is fixed at 3 or 4 layers, and the sampling time interval is preferably set to 10 minutes, so that various sensing components generate a set of corresponding monitoring values ​​every 10 minutes, thereby constructing a clearly structured multi-source raw data architecture.

[0065] This embodiment performs background correction and dimensional normalization on the output signal of the multi-source sensing component to obtain raw monitoring data that reflects soil respiration, water migration, and canopy exchange processes. "Background correction" in this embodiment refers to addressing the background shift of the sensing component under zero flux, still water, or windless conditions by combining field blank measurement results and historical stable period observation results, subtracting instrument drift and environmental background, and eliminating systematic biases introduced by non-target processes. "Dimensional normalization" refers to converting the observation results of different physical quantities according to a unified unit system and a unified time reference, so that soil CO2 flux, soil volumetric water content, soil temperature, and canopy micrometeorological elements can be compared and combined under the same time series and the same dimensional system. For example, CO2 flux is unified to a flux unit based on area and time (e.g., g·m³). -2 ·h -1 Soil moisture was standardized to volumetric water content, temperature to °C, and radiation flux to W·m. -2 All monitoring data are mapped onto a unified time axis of 10 min or 60 min, and a continuous raw monitoring dataset for carbon-water coupling analysis is formed, provided that the daily missing observation ratio is no higher than 5% and the abnormal data removal ratio is no higher than 3%.

[0066] Specifically, in step 200 of this embodiment, to ensure that the monitoring results of different physical quantities have a consistent time reference, it is preferable to trigger the collaborative acquisition operation of each monitoring component within the multi-parameter monitoring unit according to a unified sampling time interval. The "unified sampling time interval" mentioned in this embodiment refers to using the same sampling rhythm for all monitored quantities, so that they synchronously generate observation values ​​within each sampling cycle. For example, it is preferable to set the sampling time interval to 10 minutes, so that soil CO2 flux, soil moisture content, and canopy micrometeorological elements acquire data simultaneously at the trigger time every 10 minutes. This embodiment, through this synchronous triggering mechanism, ensures that the raw data of different monitored quantities have a one-to-one correspondence at the same acquisition time, avoiding problems such as time misalignment, inconsistent frequency, or acquisition drift under traditional independent acquisition methods, thereby providing a stable and reliable data foundation for subsequent carbon-water coupling analysis.

[0067] This embodiment associates soil CO2 flux monitoring data, soil moisture monitoring data, and micrometeorological monitoring data acquired at the unified trigger time using a collection time index to form a multi-parameter synchronous time-series record containing timestamps. The "collection time index association" in this embodiment refers to using the sampling trigger time as the primary key to structurally integrate multiple types of observations generated by different monitoring components at the same trigger time according to the time index. This ensures that each time point corresponds to a complete set of carbon, water, and meteorological monitoring data. For example, under a 10-minute sampling rhythm, 144 sets of structurally consistent synchronous time-series records can be formed daily. This structuring process ensures the alignment consistency of multi-source monitoring data in the time dimension, providing a complete input matrix for subsequent feature extraction, correlation analysis, and coupling identification.

[0068] This embodiment constructs a multi-source real-time data sequence to characterize the dynamic changes of carbon-water processes based on the aforementioned synchronous time-series records. The "multi-source real-time data sequence" mentioned in this embodiment refers to a serialized data structure composed of continuous timestamps and their corresponding multi-parameter monitoring values, capable of reflecting fine-scale temporal processes such as changes in soil respiration intensity, water migration dynamics, and canopy evapotranspiration exchange. For example, by arranging 144 sets of synchronous records daily in chronological order with a 10-minute sampling interval, a real-time data curve covering diurnal variations, flux peaks, and evapotranspiration fluctuations can be formed. The multi-source real-time data sequence constructed in this embodiment can further provide continuous, high-resolution data support for anomaly detection, missing data completion, feature parameter extraction, and coupling index construction.

[0069] As an optional implementation, in step 300 of this embodiment, the soil CO2 flux monitoring data, soil moisture monitoring data, and micrometeorological monitoring data are first subjected to quality checks based on multi-source real-time data sequences. This embodiment identifies data points that significantly deviate from normal variation patterns by jointly analyzing the amplitude, direction, and diurnal periodicity characteristics of continuous time series changes. Examples include unreasonable jumps, sign reversals, or abnormal fluctuations that violate the diurnal cycle within several adjacent sampling intervals. Preferably, this embodiment uses complete growing season or at least 30 days of quality-controlled historical data as a benchmark to calculate the normal variation range of each monitoring quantity under typical sunny, cloudy, and precipitation processes. The current observations are compared using these ranges, and data exceeding reasonable ranges are marked as abnormal data. Simultaneously, continuous gaps caused by sensor disconnection, communication interruption, etc., are marked as short-term missing measurement intervals for subsequent differential processing.

[0070] In this embodiment, when generating supplementary data for a target monitoring quantity within a short-term missing measurement interval, the estimated value at the missing measurement time is decomposed into two parts: a "self-trend term" and a "cooperative response term," which are then weighted and combined. The self-trend term originates from the effective observations of the target monitoring quantity adjacent to those before and after the missing measurement interval, and a trend estimate that smoothly changes over time is obtained through linear interpolation or piecewise linear interpolation. The cooperative response term originates from a set of similar monitoring quantities that exhibit significant cooperative changes with the target monitoring quantity in the carbon-water process. For example, supplementing soil CO2 flux can introduce soil moisture content, soil temperature, and net radiation flux at the same depth as similar monitoring quantities. In this embodiment, during the training phase, a segment of historical data with controlled quality is selected. The linear synergistic relationship between the target monitoring quantity and each candidate similar monitoring quantity is statistically analyzed. The correlation coefficient, which reflects the strength and direction of the correlation, is calculated, and variables with absolute values ​​greater than a preset threshold (such as 0.6) are selected to form a synergistic response set. At the same time, during the training phase, the long-term average value and standard deviation of the target monitoring quantity and each similar monitoring quantity are statistically analyzed so that the standardized fluctuation of the similar monitoring quantity can be converted into the synergistic response correction quantity under the dimension of the target monitoring quantity through the inverse vector dimension. This allows the supplemented value to inherit the time trend of the target monitoring quantity itself and reflect the synergistic change structure in the carbon-water process.

[0071] This embodiment controls the relative contributions of the self-trend term and the collaborative response term in the completion result by jointly controlling the length of the missing interval and the time scale coefficient, so as to avoid distortion of single linear interpolation caused by long-term missing data. When the missing interval only spans a small number of sampling intervals (e.g., no more than 3 to 6 sampling steps), this embodiment makes the self-trend term dominate the completion result, with only the collaborative response term used for minor correction. When the missing interval is long, the weight of the collaborative response term is gradually increased, and the completion result is constrained by the statistical synergistic relationship with variables such as soil moisture, soil temperature, and radiation, avoiding excessive smoothing of extrapolation or deviation from the actual process. The time scale coefficient can be preferably set within the range of 0.2 to 1.0, based on typical working conditions and cross-validation results, to ensure that as the length of the missing interval increases from 1 sampling step to 10 sampling steps, the proportion of the self-trend term gradually decreases and the proportion of the collaborative response term gradually increases. Finally, this embodiment rearranges the monitoring data after removing anomalies and filling in missing data in chronological order to form a continuous quality-controlled data sequence with a missing data ratio of less than 5% on a daily scale, controlled length of artificial interpolation segments, and maintained physical rationality, providing sufficient and reliable input for subsequent carbon-water coupling feature extraction and index construction.

[0072] In this embodiment, the parameters of the supplementary data are preferably calibrated and constrained using historical quality-controlled data sequences: the supplementary data itself is an estimate of the target monitoring quantities such as soil CO2 flux, soil moisture content, or related micrometeorological quantities during the missing period, and its numerical range is constrained by the minimum and maximum values ​​of the monitoring quantity observed in the historical observations in this experimental area; the trend estimate is obtained from 2 to 4 adjacent effective observation points before and after the missing period through time interpolation, preferably using linear interpolation or piecewise linear interpolation when the missing interval does not exceed 6 sampling steps, its function is to maintain the temporal smoothness of the target monitoring quantity itself; the long-term mean and standard deviation of the target monitoring quantity are both within the training... The stage or baseline statistical stage is calculated using continuous quality-controlled data for at least 30 days, preferably more than 60 days. It characterizes the central level and fluctuation range of the monitored quantity under normal operating conditions and provides a benchmark for the standardization and inverse dimension recovery of subsequent similar monitored quantities. The long-term average and standard deviation of similar monitored quantities in the same stage are obtained in the same way, with their numerical ranges corresponding to the normal value ranges of their respective physical quantities. For example, the average soil moisture content can fall between 0.15 and 0.35 (volume fraction), and the standard deviation can fall between 0.02 and 0.10. The set of similar monitored quantities is obtained by training the target monitored quantity and candidate variables (e.g., soil moisture content and soil temperature at the same depth) during the training phase. The correlation coefficients of variables such as net radiation and temperature are calculated and thresholds are set to filter them. Variables with an absolute correlation coefficient of not less than 0.6 and preferably not less than 0.7 are included in the set to ensure that the synergistic response relationship is sufficiently significant. The correlation coefficient itself is derived from the point-by-point correspondence statistics of the quality-controlled data sequence during the training phase, and is used to characterize the strength and direction of the synergistic response of similar monitoring quantities to the target monitoring quantity. The length of the missing interval can be calculated by the difference between the start and end sampling times of the missing interval, and its unit is the number of sampling steps or the time length. When the length of the missing interval does not exceed 3 to 6 sampling steps, this embodiment focuses more on using trend estimation to maintain time continuity. When the length of the missing interval is longer... As the degree gradually approaches 10 sampling steps, the weight of the co-response correction is gradually increased. The time scale coefficient is pre-calibrated according to different monitoring objects and typical climatic conditions. Under the premise of ensuring that the supplementation results do not overly rely on a single linear trend or over-amplify the noise of co-variables, the optimal value is selected in the range of 0.2 to 1.0 through cross-validation. For example, 0.3 to 0.5 can be selected in soil CO2 flux supplementation, and 0.5 to 0.8 can be selected in soil moisture supplementation. This achieves continuous adjustment of the contribution ratio of trend term and co-response term, so that the supplementation data not only falls within the reasonable range of historical observations in terms of value, but also maintains the change characteristics consistent with typical carbon-water processes in terms of morphology.

[0073] Specifically, in step 400 of this embodiment, the soil CO2 flux in the quality-controlled data sequence is structured within a sliding time window to extract key features that reflect the dynamic changes in carbon emissions. First, based on continuous soil CO2 flux observation records, this embodiment arranges the flux values ​​within the window in chronological order according to the start and end range of the sliding time window. When the sampling interval is 10 minutes and the window length is 2 hours, one window contains 12 flux observation points. This embodiment calculates the quotient of the CO2 flux difference between two adjacent sampling times within the window and the sampling interval to obtain the CO2 flux change rate at the window scale, thereby reflecting the growth, decay, or fluctuation trend of soil respiration intensity within a short time scale. For example, when the CO2 flux at the beginning of a window is 0.40 g·m³... -2 ·h -1 The end of the window is 0.55 g·m -2 ·h -1 If the flux change rate within the window shows an upward trend, it can be used to identify rapid flux response processes caused by increased temperature, humidity, or enhanced microbial activity.

[0074] This embodiment further accumulates CO2 flux within a sliding time window to construct a CO2 flux accumulation characteristic consistent across day-scale. This embodiment sums the instantaneous flux values ​​at all sampling times within the window according to the sampling time interval, thereby obtaining the CO2 accumulation within the window period. Then, starting from midnight 00:00 on the day-scale, the accumulation values ​​of all windows within the same natural day are superimposed in chronological order to form the daily cumulative CO2 flux process curve. For example, under clear weather conditions, the daily cumulative CO2 flux for the entire natural day may reach 5.0 g·m³. -2 However, under moist conditions following rainfall, it can increase to 8.0 g·m³. -2 This embodiment can fully reflect this diurnal carbon flux enhancement process, and can be used to identify seasonal peaks, rainfall pulse effects and diurnal variation patterns.

[0075] This embodiment ultimately constructs a carbon flux characteristic parameter by combining the soil CO2 flux change rate and the daily cumulative CO2 flux to achieve a dual characterization of the dynamic fluctuations and cumulative effects of carbon emissions. In this parameter combination, the flux change rate reflects rapid dynamic processes, such as diurnal respiration rhythms, temperature impulse responses, and humidity-induced changes; while the daily cumulative CO2 flux is used to characterize the overall level and net effect of carbon emissions, and to identify periods of high soil respiration load, emission increases caused by farmland management measures, and emission reduction trends during ecological compensation phases. For example, when the flux change rate shows an upward trend and the daily cumulative flux is significantly higher than the daily average (e.g., exceeding 6.0 g·m³), the carbon flux characteristic parameter is determined. -2 ·d -1This indicates that the system is in a state of high carbon emission activity; however, when the flux change rate fluctuation weakens and the daily cumulative flux decreases to 3.0 g·m³, the system is considered to be in a state of high carbon emission activity. -2 ·d -1 The following indicates that carbon emissions are in a suppression or recovery phase. Through this structured feature extraction process, this embodiment can ensure that the carbon flux characteristic parameters have sufficient physical interpretability, temporal continuity, and index stability, providing a solid data foundation for the subsequent construction of the carbon-water coupling index.

[0076] Further, in step 500 of this embodiment, to obtain the dynamic changes in root zone moisture within a sliding time window, the soil volumetric water content at different depths is first discretely accumulated to calculate the soil moisture storage at any given moment within the window. In this embodiment, the root zone is typically divided into several discrete soil layers of fixed thickness. For example, 5cm, 10cm, and 20cm can be selected as representative layer thicknesses. The volumetric water content value after quality control is multiplied by the thickness of each layer, and then all soil layers are summed to obtain the soil moisture storage at that moment. Subsequently, using the starting time of the sliding time window as a reference, the moisture storage at the end of the window is subtracted from the moisture storage at the reference time to obtain the change in moisture storage within the window. For example, if a 2-hour sliding time window starts at 08:00, and the soil moisture storage at 08:00 is 35mm and at 10:00 is 30mm, then the change in moisture storage within the window is a decrease of 5mm, thus reflecting the actual water consumption of the root zone within that window.

[0077] This embodiment also requires summarizing the evapotranspiration intensity within the sliding time window to characterize the overall moisture output rate of the surface and canopy systems to the atmosphere during the window period. To this end, this embodiment sums the instantaneous evapotranspiration fluxes at each sampling moment within the window and divides this sum by the window duration to obtain the average evapotranspiration intensity of the window. For example, when the sampling interval is 10 minutes and the window length is 2 hours, there are 12 instantaneous evapotranspiration flux values ​​within the window. If their average level is approximately 0.25 mm per hour, then the evapotranspiration intensity of the window can be characterized as a moisture output of 0.25 mm per hour. This time smoothing method effectively reduces interference from short-term fluctuations and more accurately reflects the moisture consumption process during the window period.

[0078] This embodiment ultimately constructs a combined model of water storage change and evapotranspiration intensity to obtain water cycle characteristic parameters that simultaneously reflect root zone water consumption and atmospheric evapotranspiration. Water storage change reflects soil water consumption or replenishment, while evapotranspiration intensity reflects the rate at which the farmland ecosystem releases water into the atmosphere; together, they constitute the complete water cycle process. For example, when water storage decreases by 5 mm and evapotranspiration intensity reaches 0.25 mm per hour within a certain window, it can be inferred that the window is in a state dominated by significant water deficit; conversely, when water storage increases by 3 mm and evapotranspiration intensity remains below 0.10 mm per hour, it reflects the water recovery process after rainfall replenishment. Through this structured combination, this embodiment can more accurately reveal the driving mechanisms of water transport in farmland ecosystems, providing stable, sufficient, and high-resolution water process input for subsequent identification of carbon-water coupling processes.

[0079] In this embodiment, the soil moisture storage parameter is determined by the volumetric water content of each discrete soil layer and the corresponding soil layer thickness. The volumetric water content is derived from soil moisture observations at different depths in the aforementioned quality-controlled data sequence, and the unit can be cubic meters per cubic meter, with a typical value range between 0.10 and 0.45. The soil layer thickness can be preset according to the profile structure of the test area. For example, the root zone is divided into several layers, each with a thickness of 5 cm, 10 cm, or 20 cm. The volumetric water content of each layer is multiplied by the thickness and then summed to obtain the soil moisture storage of the root zone at a certain moment. The unit of soil moisture storage can be converted to millimeters. The sliding time window start time parameter marks the beginning of each calculation window. For example, it can slide forward sequentially at 00:00, 00:10, and 00:20 each day, each time covering one sampling interval. The sliding time window length parameter limits the duration of each window's coverage. For example, it can be selected as 1 hour, 2 hours, or 24 hours. When the window length is 2 hours and the sampling interval is 10 minutes, one window contains 12 sampling times. The water storage change parameter is obtained by subtracting the soil moisture storage at the beginning of the window from the soil moisture storage at the end of the window. If the beginning is 35 mm and the end is 30 mm, the water storage change is a decrease of 5 mm. Its function is to reflect the net consumption or net recharge of root zone water within the window. The instantaneous evapotranspiration flux parameter is obtained from canopy micrometeorological monitoring results or energy balance calculations, and can be measured in millimeters per hour or energy flux. After quality control, these values ​​are arranged at fixed time intervals. The sampling interval parameter, for example, is 10 minutes or 30 minutes, corresponding to 144 or 48 evapotranspiration flux values ​​per day. The number of sampling moments within a sliding time window is determined by both the window length and the sampling interval; for example, when the window length is 2 hours and the sampling interval is 10 minutes, the number of sampling moments within the window is 12. The average evapotranspiration intensity parameter is obtained by summing all instantaneous evapotranspiration fluxes within the window and dividing by the total window duration. It characterizes the average rate at which surface and canopy water is released into the atmosphere within that time window. Ultimately, the change in water storage and the average evapotranspiration intensity together constitute the characteristic parameters of the water cycle. One primarily reflects the water surplus or deficit within the soil, while the other reflects the intensity of water output into the atmosphere. The combination of their values ​​and signs can be used to identify different water processes such as rainfall replenishment, evaporation depletion, and soil drought development.

[0080] Furthermore, in step 600 of this embodiment, to characterize the coordinated changes in carbon flux and water cycle processes over time, the carbon flux characteristic parameter sequence and the water cycle characteristic parameter sequence are first paired one-to-one within a sliding time window. Specifically, using the start time of each sliding time window as an index, the carbon flux characteristic values ​​and the corresponding water cycle characteristic values ​​at all sampling times within that window are arranged chronologically to form two sub-sequences of equal length. For example, when the sliding time window length is 2 hours and the sampling interval is 10 minutes, each window contains 12 sampling points, corresponding to 12 carbon flux characteristic values ​​and 12 water cycle characteristic values. This embodiment, through this temporal pairing method, organizes the synchronous changes of carbon flux and water cycle within the same window into a joint characteristic sequence, providing an input basis for subsequent calculations of response intensity and response hysteresis.

[0081] This embodiment, after obtaining the joint feature sequence, further calculates the carbon-water response coefficient to quantitatively characterize the correlation strength and direction between carbon flux characteristics and water cycle characteristics within the current sliding time window. Specifically, the average and standard deviation of the carbon flux feature subsequence within the window are first calculated, followed by the average and standard deviation of the water cycle feature subsequence. Then, the degree to which the carbon flux deviates from its average value and the degree to which the water cycle deviates from its average value at each pair of synchronous sampling points are multiplied and summed, and normalized using the standard deviation to obtain a dimensionless response coefficient between -1 and 1. When the coefficient is close to 1, it indicates a significant synergistic relationship of mutual enhancement or weakening between carbon flux and water cycle within the time window; when it is close to -1, it indicates a significant inverse relationship between the two within the time window; when it is close to 0, it indicates that the linear response relationship between carbon and water is not obvious within the time window. For example, for a period of time after irrigation during the peak growing season, the carbon-water response coefficient may stabilize above 0.7, reflecting the high carbon release characteristics driven by sufficient water.

[0082] This embodiment also reveals the temporal delay or advance effect of carbon processes on water processes by calculating the carbon-water response lag time. The basic idea is to perform correlation analysis on the carbon flux characteristic sequence and the water cycle characteristic sequence at different time offsets within a given sliding time window, searching for the offset that maximizes the correlation between the two within the maximum allowed offset step range. This offset is then converted into a time length and used as the carbon-water response lag time. When the lag time is positive, it indicates that the change in carbon flux lags behind the change in water cycle; for example, after rainfall or irrigation, soil respiration intensity peaks several hours later. When the lag time is negative, it indicates that the change in carbon flux precedes the change in water cycle characteristics; for example, a rapid increase in soil temperature leads to enhanced carbon release, which then triggers a significant increase in water consumption. This embodiment avoids identifying excessively long lags that do not conform to the actual physical process by setting an upper limit on the maximum number of steps for the lag search, for example, limiting it to several sampling step lengths corresponding to 1h to 6h.

[0083] After obtaining the carbon-water response coefficient and response lag time, this embodiment constructs a carbon-water coupling index using these two as inputs to comprehensively characterize the degree of coupling between carbon flux and water cycle processes in terms of both intensity and timing. This coupling index increases with the absolute value of the response coefficient, reflecting a stronger correlation and higher coupling degree between carbon and water; conversely, it decreases with the increase of the response lag time, reflecting a greater temporal mismatch between the carbon and water processes and a weaker coupling degree. For example, when the absolute value of the carbon-water response coefficient is close to 1 and the lag time is less than 30 minutes within a certain time window, the carbon-water coupling index can approach its theoretical upper limit, indicating that carbon and water are almost synchronous and highly coordinated within this window. However, when the absolute value of the response coefficient is less than 0.2 and the lag time is close to several hours, the coupling index will decrease significantly, indicating that carbon and water are essentially decoupled within this time window.

[0084] This embodiment dynamically tracks the carbon-water coupling index over a continuous time window and divides the farmland carbon-water process into strongly coupled, weakly coupled, and decoupled zones based on its numerical level and temporal stability, thereby achieving real-time detection of the carbon-water coupling process. In specific implementation, a set of empirical thresholds can be set based on years of observation data from the experimental area. For example, periods when the coupling index remains above 0.6 for multiple consecutive windows with small fluctuations are identified as strongly coupled zones; periods when the coupling index is mostly between 0.3 and 0.6 with large fluctuations are identified as weakly coupled zones; and periods when the coupling index is consistently below 0.3 are identified as decoupled zones. By comparing with irrigation events, rainfall processes, drought processes, and farmland management measures such as fertilization and tillage, this embodiment can correlate strongly coupled zones with key water regulation periods, thereby providing real-time and quantitative decision-making basis for irrigation scheduling optimization, carbon sink management, and ecological compensation assessment.

[0085] Corresponding to the above methods, such as Figure 2 As shown in the figure, this embodiment also provides a real-time multi-parameter detection system for farmland carbon-water coupling processes, including:

[0086] The multi-parameter monitoring unit deployment module is used to set up multi-parameter monitoring units in farmland experimental areas that can simultaneously acquire soil CO2 flux, soil moisture content, soil temperature and canopy micrometeorological elements, so as to continuously acquire raw monitoring data of carbon-water processes.

[0087] The multi-source data synchronous acquisition module is used to synchronously acquire the original monitoring data at the same time interval, and to correspond the soil CO2 flux monitoring data, soil moisture monitoring data and micrometeorological monitoring data to the same time sequence to form a multi-source real-time data sequence.

[0088] The data quality control module is used to remove abnormal data and supplement short-term missing data based on the multi-source real-time data sequence to form a quality-controlled data sequence for carbon-water coupling analysis.

[0089] The carbon flux characteristic calculation module is used to calculate the soil CO2 flux change rate and daily cumulative CO2 flux based on the quality-controlled data sequence within a sliding time window, thereby forming carbon flux characteristic parameters.

[0090] The water cycle characteristic calculation module is used to calculate the change in soil moisture storage and evapotranspiration intensity based on the quality-controlled data sequence within the sliding time window, thereby forming water cycle characteristic parameters.

[0091] The carbon-water coupling analysis and segment identification module is used to calculate the carbon-water response coefficient and response lag time based on the carbon flux characteristic parameters and the water cycle characteristic parameters, construct the carbon-water coupling index, and identify strong coupling zone, weak coupling zone and decoupling zone based on the change of the carbon-water coupling index within a continuous time window, so as to realize the real-time detection of the carbon-water coupling process in farmland.

[0092] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0093] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for real-time detection of multiple parameters in a carbon-water coupling process in farmland, characterized in that, include: Multi-parameter monitoring units capable of simultaneously acquiring soil CO2 flux, soil moisture content, soil temperature, and canopy micrometeorological elements were set up in the farmland experimental area to continuously acquire raw monitoring data of carbon-water processes. The original monitoring data were collected synchronously at the same time interval, and the soil CO2 flux monitoring data, soil moisture monitoring data and micrometeorological monitoring data were matched to the same time sequence to form a multi-source real-time data sequence. Based on the multi-source real-time data sequence, abnormal data are removed and short-term missing data are supplemented to form a quality-controlled data sequence for carbon-water coupling analysis. Within a sliding time window, the soil CO2 flux change rate and daily cumulative CO2 flux are calculated based on the quality-controlled data sequence to form carbon flux characteristic parameters. Within the sliding time window, the change in soil moisture storage and evapotranspiration intensity are calculated based on the quality-controlled data sequence to form water cycle characteristic parameters. Based on the carbon flux characteristic parameters and the water cycle characteristic parameters, the carbon-water response coefficient and response lag time are calculated, a carbon-water coupling index is constructed, and strong coupling region, weak coupling region and decoupling region are identified based on the changes of the carbon-water coupling index within a continuous time window, so as to realize the real-time detection of the carbon-water coupling process in farmland.

2. The method for real-time detection of multiple parameters in farmland carbon-water coupling processes according to claim 1, characterized in that, Multi-parameter monitoring units capable of simultaneously acquiring soil CO2 flux, soil moisture content, soil temperature, and canopy microclimate elements were set up in the farmland experimental area to continuously obtain raw monitoring data on carbon-water processes, including: Based on the differences in soil physicochemical properties and the structure of the vegetation canopy, the deployment location and vertical sensing depth of the multi-parameter monitoring unit in the farmland experimental area are determined so that the multi-parameter monitoring unit can simultaneously cover the environmental gradient of the soil surface layer, the root zone layer and the canopy air layer. The multi-parameter monitoring unit is equipped with multi-source sensing components for characterizing carbon flux, water dynamics and canopy micrometeorological elements, and a collaborative sensing system constrained by unified spatial positioning and continuous physical response is established. Background correction and dimensional normalization are performed on the output signal of the multi-source sensing component to obtain the raw monitoring data that can reflect the processes of soil respiration, water migration and canopy exchange.

3. The method for real-time detection of multiple parameters in farmland carbon-water coupling processes according to claim 1, characterized in that, The original monitoring data were collected synchronously at the same time intervals, and the soil CO2 flux monitoring data, soil moisture monitoring data, and micrometeorological monitoring data were mapped to the same time series to form a multi-source real-time data sequence, including: The coordinated acquisition operation of the soil CO2 flux monitoring unit, soil moisture monitoring unit and micrometeorological monitoring unit in the multi-parameter monitoring unit is triggered according to the preset unified sampling time interval, so that different monitoring quantities generate corresponding data at the same acquisition time. The soil CO2 flux monitoring data, soil moisture monitoring data and micrometeorological monitoring data acquired at each collection time will be indexed and associated according to the collection time to generate synchronous time-series records corresponding to multiple parameters; The multi-source real-time data sequence used to characterize the dynamic changes of the carbon-water process is constructed based on the synchronous time-series records.

4. The method for real-time detection of multiple parameters in farmland carbon-water coupling processes according to claim 1, characterized in that, Based on the multi-source real-time data sequence, outlier data is removed and short-term missing data is supplemented to form a quality-controlled data sequence for carbon-water coupling analysis, including: Consistency checks and fluctuation constraint analyses were performed on the soil CO2 flux monitoring data, soil moisture monitoring data, and micrometeorological monitoring data in the multi-source real-time data sequence to identify abnormal data that deviated from the normal variation pattern. Abnormal data that has been identified will be removed, and supplementary data will be generated based on the changing trends of adjacent valid data and the collaborative response relationship of similar monitoring quantities during the missing data period. The monitoring data, after removing anomalies and filling in missing data, are reconstructed into a continuous quality-controlled data sequence.

5. The method for real-time detection of multiple parameters in farmland carbon-water coupling process according to claim 4, characterized in that, The formula for generating the supplementary data is: ； in, For the time during the missing test period Target monitoring quantity The generated completion data; Based on the target monitoring quantity The trend estimate is obtained by linear interpolation of observations at adjacent valid times before and after the missing period; For target monitoring quantity The long-term average during the training or benchmark statistics phase; For target monitoring quantity The standard deviation of the stage; In order to match the target monitoring quantity Similar monitoring quantities exhibiting a synergistic response relationship between carbon and water processes At any moment Observed values; For the same type of monitoring volume The long-term average value during the aforementioned period; For the same type of monitoring volume The standard deviation of the stage; In order to match the target monitoring quantity A set of similar monitoring quantities that exhibit significant synergistic response relationships; The target monitoring quantity is calculated based on historical quality-controlled data sequences during the training phase. Compared with similar monitoring volume The correlation coefficient between them; To revolve around time Target monitoring volume The length of the corresponding missing measurement interval; The time scale coefficient is used to control the relative contribution of the trend term and the collaborative response term.

6. The method for real-time detection of multiple parameters in farmland carbon-water coupling processes according to claim 1, characterized in that, Within the sliding time window, soil moisture storage changes and evapotranspiration intensity are calculated based on the quality-controlled data sequence to form water cycle characteristic parameters, including: Based on the aforementioned quality-controlled data sequence, the root zone soil volumetric water content is discretized and accumulated within a sliding time window, with the calculation time... Soil moisture storage And obtain the starting time of the sliding time window. Changes in soil moisture storage as a baseline The soil moisture storage satisfies: ;in, For a moment Soil moisture reserves; The start time of the sliding time window Changes in water storage; For a moment No. The quality of each soil layer is controlled by its volumetric moisture content; For the first Each soil layer thickness; The number of discrete soil layers; The length of the sliding time window; The average evapotranspiration intensity of the sliding time window is calculated based on the instantaneous evapotranspiration flux time series within the sliding time window. ,satisfy: ;in, The average evapotranspiration intensity corresponding to the sliding time window; For a moment Instantaneous evapotranspiration flux; The sampling time interval; This represents the number of sampling moments within the sliding time window; The change in soil moisture storage With the evaporation intensity By combining these parameters, we can construct characteristic parameters of the water cycle that reflect the combined driving mechanism of root zone water consumption and atmospheric evapotranspiration.

7. The method for real-time detection of multiple parameters in farmland carbon-water coupling processes according to claim 1, characterized in that, Based on the carbon flux characteristic parameters and the water cycle characteristic parameters, the carbon-water response coefficient and response lag time are calculated to construct a carbon-water coupling index. Strong coupling, weak coupling, and decoupling regions are identified based on the changes in the carbon-water coupling index within a continuous time window, enabling real-time detection of the carbon-water coupling process in farmland, including: Within the sliding time window, the carbon flux characteristic parameter sequence is time-series paired with the water cycle characteristic parameter sequence within the same sliding time window, and the starting time of each sliding time window is... Extraction length is carbon flux characteristic subsequence With water cycle characteristic subsequence We constructed a joint feature sequence to characterize the synergistic relationship between changes in carbon flux and water cycle status. Based on the joint feature sequence, the carbon-water response coefficient and the carbon-water response lag time are calculated within each sliding time window, wherein the carbon-water response coefficient... satisfy: Furthermore, the carbon-water response lag time satisfy: ;in, The start time of the sliding time window The corresponding carbon-water response coefficient; The carbon-water response lag time within the sliding time window; For the first time window within the sliding time window The quality of the carbon flux characteristic parameters at each sampling time is controlled; For the first time window within the sliding time window The quality of the water cycle characteristic parameters at each sampling time is controlled; The carbon flux feature subsequence within the current sliding time window The average value; The water cycle feature subsequence within the current sliding time window The average value; The carbon flux feature subsequence within the current sliding time window Standard deviation; The water cycle feature subsequence within the current sliding time window Standard deviation; This represents the number of sampling moments included within the sliding time window; This is the discrete-time offset used to characterize carbon flux features relative to water cycle features; This represents the maximum number of steps that can be taken in a time-lapse search. A carbon-water coupling index is constructed using the carbon-water response coefficient and the carbon-water response lag time as inputs. The carbon-water coupling process in farmland is then classified and identified based on the changes in the carbon-water coupling index within a continuous time window. satisfy: ;in, The start time of the sliding time window The corresponding carbon-water coupling index; The length of the sliding time window; The absolute value of the carbon-water response coefficient is used to characterize the correlation strength between carbon flux characteristics and water cycle characteristics. The absolute value of the carbon-water response lag time is used to characterize the degree of time mismatch in carbon-water interaction. Within a continuous time window, when the carbon-water coupling index remains at a stable high level, it is identified as a strong coupling region; when the carbon-water coupling index is at a medium level and fluctuates greatly, it is identified as a weak coupling region; and when the carbon-water coupling index is continuously lower than a preset threshold, it is identified as a decoupling region, thereby realizing real-time detection of the carbon-water coupling process in farmland.

8. A real-time multi-parameter detection system for a carbon-water coupling process in farmland, characterized in that, include: The multi-parameter monitoring unit deployment module is used to set up multi-parameter monitoring units in farmland experimental areas that can simultaneously acquire soil CO2 flux, soil moisture content, soil temperature and canopy micrometeorological elements, so as to continuously acquire raw monitoring data of carbon-water processes. The multi-source data synchronous acquisition module is used to synchronously acquire the original monitoring data at the same time interval, and to correspond the soil CO2 flux monitoring data, soil moisture monitoring data and micrometeorological monitoring data to the same time sequence to form a multi-source real-time data sequence. The data quality control module is used to remove abnormal data and supplement short-term missing data based on the multi-source real-time data sequence to form a quality-controlled data sequence for carbon-water coupling analysis. The carbon flux characteristic calculation module is used to calculate the soil CO2 flux change rate and daily cumulative CO2 flux based on the quality-controlled data sequence within a sliding time window, thereby forming carbon flux characteristic parameters. The water cycle characteristic calculation module is used to calculate the change in soil moisture storage and evapotranspiration intensity based on the quality-controlled data sequence within the sliding time window, thereby forming water cycle characteristic parameters. The carbon-water coupling analysis and segment identification module is used to calculate the carbon-water response coefficient and response lag time based on the carbon flux characteristic parameters and the water cycle characteristic parameters, construct the carbon-water coupling index, and identify strong coupling zone, weak coupling zone and decoupling zone based on the change of the carbon-water coupling index within a continuous time window, so as to realize the real-time detection of the carbon-water coupling process in farmland.

Images