Natural environment monitoring method and system based on big data analysis

Abstract

The invention relates to the technical field of big data processing, in particular to a natural environment monitoring method and system based on big data analysis, and the method comprises the following steps: collecting multi-dimensional data to generate a meteorological thermal feature set, comparing with a physical reference to construct a three-dimensional concentration distribution tensor, and generating an effective transmission load sequence in combination with a wind speed projection and an industrial load. The method comprises the following steps: constructing a pollutant transmission lag parameter based on cross-correlation analysis, synthesizing a transfer matrix by combining real-time multiplying power evolution, and outputting environment prediction data, and comprises the following steps: comparing meteorological thermal characteristics with a physical reference to generate a correlation weight, carrying out three-dimensional grid filling on pollutant data, and reconstructing a three-dimensional distribution form of an atmospheric environment. Calculating a wind speed vector projection, establishing a transmission weight, generating an effective transmission sequence in combination with an industrial load, quantifying a directional driving mechanism of an emission source to a monitoring point, locking a transmission lag parameter by utilizing cross-correlation analysis, and realizing dynamic prediction of an environment state in combination with a real-time transmission multiplying power evolution state transition matrix.

Description

Technical Field

[0001] This invention relates to the field of big data processing technology, and in particular to a method and system for monitoring the natural environment based on big data analysis. Background Technology

[0002] The field of big data processing technology encompasses a systematic technical framework for collecting, storing, managing, and analyzing massive and diverse datasets. Core aspects of this field include using distributed file systems to address data storage bottlenecks, cleaning heterogeneous data sources through extraction, transformation, and loading tools, and employing parallel computing frameworks to mine and process datasets. The aim is to extract valuable information from high-flow-rate data streams and support decision-making. Traditional natural environment monitoring methods refer to techniques for fixed-point observation of atmospheric, water, or soil indicators in target areas. This typically involves establishing fixed monitoring stations at designated geographical coordinates, using electrochemical or optical sensors to detect physicochemical quantities such as sulfur dioxide concentration, PM2.5 levels, or chemical oxygen demand. Analog-to-digital converters convert the sensor outputs from analog signals to digital signals, and the raw readings are transmitted to a central server using General Packet Radio Service (GPRS) or mobile communication networks. The server then writes the received values ​​into a relational database table in chronological order for manual review or simple linear regression calculations.

[0003] Traditional monitoring relies on discrete fixed stations to collect environmental indicators, resulting in spatial blind spots in non-sampling areas and failing to reflect the continuous distribution of pollutants in three-dimensional space. It ignores the complex nonlinear coupling relationship between meteorological thermodynamics and pollutant diffusion dynamics, and data processing lacks a correlation mechanism between industrial emission sources and environmental monitoring points, making it difficult to accurately pinpoint transmission lag time and effectively predict the dynamic evolution trend of environmental conditions under varying wind fields. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing a natural environment monitoring method and system based on big data analysis.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a natural environment monitoring method based on big data analysis, comprising the following steps: S1: Collect air pressure, temperature, horizontal wind speed component, wind direction angle, pollutant concentration sampling data and industrial power load data, calculate potential temperature vertical gradient parameters and atmospheric dynamic stability parameters, and generate meteorological and thermodynamic feature datasets. S2: Call the meteorological and thermodynamic feature dataset, compare the potential temperature vertical gradient parameter, atmospheric dynamic stability parameter and standard meteorological and physical benchmark record to generate correlation weight coefficients, fill the pollutant concentration sampling data with weights, and construct a three-dimensional concentration distribution tensor record; S3: Call the meteorological and thermal feature dataset, construct a spatial relative position vector, calculate the projection component of the horizontal wind speed component on the spatial relative position vector to generate a transmission weight coefficient, and perform point-by-point multiplication operation on the industrial electricity load data and the transmission weight coefficient to generate an effective transmission load sequence record. S4: Calculate the correlation parameter between the pollutant concentration time series data at the corresponding monitoring point locations of the effective transmission load sequence record and the three-dimensional concentration distribution tensor record, search for the peak position of the correlation parameter, and construct the pollutant transmission lag parameter. S5: Combine the pollutant transport lag parameter and geometric displacement vector to obtain the real-time power transport rate factor, use the real-time power transport rate factor to correct the time step of the basic state transition probability matrix and generate a synthetic transition matrix, map the three-dimensional concentration distribution tensor record, and output environmental state prediction data.

[0006] As a further aspect of the present invention, the meteorological and thermodynamic feature dataset includes a potential temperature vertical distribution table, a dynamic stability feature vector, and a multidimensional meteorological element index. The three-dimensional concentration distribution tensor record specifically comprises a set of spatial grid concentration values, weighted filling confidence markers, and vertical barrier state matrices. The effective transmission load sequence record includes normalized wind field projection flux, industrial load time-series mapping values, and source point transmission correlation factors. The pollutant transmission lag parameters specifically refer to transmission time delay values, interrelated peak offsets, and time window search confidence. The environmental state prediction data includes a concentration distribution tensor for the next preset time step, a regional diffusion evolution trend map, and a synthetic state transition probability table.

[0007] As a further aspect of the present invention, the steps for obtaining the meteorological and thermodynamic feature dataset are as follows: S111: Collect air pressure parameters, temperature parameters, horizontal wind speed components, wind direction angle parameters, and pollutant concentration sampling data at multiple altitude levels; calculate the thermodynamic correlation between air pressure parameters and temperature parameters; perform adiabatic compression state conversion processing; determine the potential temperature parameters corresponding to the air mass conservation properties at each monitoring point; spatially sequence and arrange data at different altitude levels; and construct a basic meteorological potential temperature sequence. S112: Call the basic meteorological potential temperature sequence, analyze the vertical distribution structure and retrieve the potential temperature parameters corresponding to adjacent height layers, calculate the physical difference of potential temperature parameters between adjacent layers, perform gradient calculation in combination with the height difference between layers, determine the atmospheric stratification thermal stratification state and establish vertical stability characteristics, and construct a vertical potential temperature gradient vector. S113: Call the vertical potential temperature gradient vector, combine it with the horizontal wind speed component and vertical stability characteristics at the same location, calculate the atmospheric dynamic stability parameters, integrate the basic meteorological potential temperature sequence, wind direction angle and pollutant concentration data, perform multi-dimensional spatiotemporal correlation mapping, and generate a meteorological thermodynamic feature dataset.

[0008] As a further aspect of the present invention, the steps for obtaining the three-dimensional concentration distribution tensor record are specifically as follows: S211: Call the meteorological and thermodynamic feature dataset, call the standard meteorological reference record, extract the thermal stability reference and the shear turbulence reference, perform differential comparison operation on the potential temperature vertical gradient parameter and the thermal stability reference, perform interval determination operation on the atmospheric dynamic stability parameter and the shear turbulence reference, integrate the comparison results to quantify the degree of vertical flux obstruction, and generate a vertical obstruction status identifier. S212: Based on the vertical barrier status indicator, analyze the inhibition strength of atmospheric stratification on pollutant diffusion, perform numerical quantification calculation on the inhibition strength, establish a vertical diffusion attenuation factor, construct a spatial interpolation weight allocation relationship based on the vertical diffusion attenuation factor, calculate the normalized numerical index of the correlation degree between monitoring points, and generate correlation weight coefficients. S213: Call the associated weight coefficient, combine it with the meteorological and thermodynamic feature dataset, perform a three-dimensional spatial gridded weighted filling operation on the pollutant concentration sampling data according to the associated weight coefficient, calculate the estimated concentration value of the non-sampling area in space, perform structured encapsulation and tensor definition on the entire field concentration data according to the spatial dimension, and establish a three-dimensional concentration distribution tensor record.

[0009] As a further aspect of the present invention, the process of obtaining the standard meteorological reference record is specifically as follows: The system analyzes vertical stratified records of air pressure, temperature, and wind speed within the historical statistical period of the target monitoring area. It calculates the historical potential temperature vertical gradient set and performs Gaussian normal distribution fitting to obtain the expected mathematical parameters and standard deviation statistical parameters of the fitted distribution curve. These are set as the neutral stratification baseline and allowable fluctuation amplitude, respectively, to establish the thermal stability benchmark. The system calculates the historical vertical wind shear index set and performs bimodal clustering statistical analysis to identify cluster centers in low-shear and high-shear states. It calculates the arithmetic mean boundary value of the two cluster centers as the dynamic shear threshold to establish the shear turbulence benchmark. The system integrates the thermal stability benchmark and the shear turbulence benchmark to establish the standard meteorological benchmark record analysis. It identifies cluster centers in low-shear and high-shear states, calculates the arithmetic mean boundary value of the two cluster centers, and determines the arithmetic mean boundary value as the dynamic shear threshold for fluid state transitions to establish the shear turbulence benchmark. Finally, it integrates the thermal stability benchmark and the shear turbulence benchmark to establish the standard meteorological benchmark record.

[0010] As a further aspect of the present invention, the step of obtaining the effective transmission payload sequence record specifically comprises: S311: Call the meteorological and thermodynamic feature dataset, parse the horizontal wind speed component and wind direction angle data, retrieve the geospatial coordinate information of industrial emission sources and environmental monitoring points, construct a spatial relative position vector from the emission source location to the monitoring point location, perform vector projection decomposition on the horizontal wind speed component and wind direction angle parameters in the direction of the spatial relative position vector, calculate the effective component intensity of the wind speed vector on the transmission path, and generate wind direction projection component parameters. S312: Based on the wind direction projection component parameters, analyze the geometric consistency characteristics between the wind speed projection direction and the pollutant transport path, perform normalization mapping operation according to the positive and negative polarity and magnitude of the projection component values, determine the dynamic driving and blocking state of the wind field on pollutant diffusion under the current meteorological conditions, construct a quantitative weight index for the directional transport efficiency of pollutants, and generate normalized transport weight coefficients. S313: Obtain time-series data of industrial electricity load in the factory, and map it into a non-steady-state emission source strength sequence based on the transformation relationship between energy consumption and pollutant generation. Perform point-by-point multiplication of the non-steady-state emission source strength sequence with the normalized transmission weight coefficient in the time dimension, and generate an effective transmission load sequence record by taking into account the correction effect of meteorological transmission efficiency on source strength emissions.

[0011] As a further aspect of the present invention, the step of obtaining the pollutant transport hysteresis parameter specifically comprises: S411: Call the three-dimensional concentration distribution tensor record, extract pollutant concentration time series data according to the monitoring point spatial index, obtain the preset environmental background noise baseline, perform baseline difference processing on the pollutant concentration time series data, and generate net value concentration fluctuation sequence. S412: Call the effective transmission load sequence record, calculate the energy integral of the environmental background noise baseline, obtain the estimated value of the background noise energy, calculate the autocorrelation energy of the net value concentration fluctuation sequence, obtain the total signal energy value, set the candidate time displacement variable within the standard cross-correlation search window parameter range, calculate the sequence dynamic coupling degree between the net value concentration fluctuation sequence and the effective transmission load sequence record, and generate a time-shifted correlation strength distribution set; S413: Retrieve the global peak point in the time-shift correlation intensity distribution set, lock the time offset index corresponding to the global peak point, determine that the target index is a time delay feature, and convert the time delay feature into a physical time value in combination with the sampling frequency to construct pollutant transmission lag parameters for the target industrial emission source.

[0012] As a further aspect of the present invention, the step of obtaining the environmental state prediction data specifically includes: S511: Call the pollutant transport lag parameter, calculate the geographical distance scalar of the monitoring point relative to the industrial emission source, perform rate calculation based on the ratio of the geographical distance scalar to the pollutant transport lag parameter, establish a benchmark diffusion rate parameter, obtain the current average wind speed parameter of the monitoring area, perform a division comparison operation on the current average wind speed parameter of the monitoring area and the benchmark diffusion rate parameter, and generate a real-time power transmission rate factor. S512: Call the three-dimensional concentration distribution tensor record, perform global discrete statistical analysis on the concentration level fluctuation sequence, obtain the state change frequency characteristics between adjacent discrete moments, normalize and construct the basic state transition probability matrix, perform quantization and rounding on the real-time power transmission rate factor, obtain the integer step level, calculate the synthetic concentration state transition probability, and construct the synthetic transition matrix. S513: Extract the current state vector from the three-dimensional concentration distribution tensor record, map the current state vector to the synthetic transition matrix, perform matrix multiplication to deduce the state distribution probability, integrate the deduction results of the spatiotemporal dimensions, and output environmental state prediction data.

[0013] A natural environment monitoring system based on big data analysis, wherein the natural environment monitoring system based on big data analysis is used to execute the aforementioned natural environment monitoring method based on big data analysis, the system comprising: The data acquisition and calculation module collects air pressure, temperature, horizontal wind speed components, wind direction angle, pollutant concentration sampling data and industrial power load data, calculates potential temperature vertical gradient parameters and atmospheric dynamic stability parameters, and generates meteorological and thermodynamic characteristic datasets. The concentration-weighted construction module calls the meteorological and thermodynamic feature dataset, compares the potential temperature vertical gradient parameter, atmospheric dynamic stability parameter and standard meteorological and physical benchmark record to generate correlation weight coefficients, and fills the pollutant concentration sampling data with weights to construct a three-dimensional concentration distribution tensor record. The transmission load calculation module calls the meteorological and thermal feature dataset, constructs a spatial relative position vector, calculates the projection component of the horizontal wind speed component on the spatial relative position vector to generate a transmission weight coefficient, and performs point-by-point multiplication operation on the industrial electricity load data and the transmission weight coefficient to generate an effective transmission load sequence record. The lag parameter extraction module calculates the correlation parameters between the pollutant concentration time series data at the corresponding monitoring point locations of the effective transmission load sequence record and the three-dimensional concentration distribution tensor record, searches for the peak position of the correlation parameters, and constructs pollutant transmission lag parameters. The prediction data generation module combines the pollutant transport lag parameter and geometric displacement vector to obtain the real-time power transport rate factor, uses the real-time power transport rate factor to correct the time step of the basic state transition probability matrix and generate a synthetic transition matrix, maps the three-dimensional concentration distribution tensor record, and outputs environmental state prediction data.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by comparing meteorological and thermodynamic characteristics with physical benchmarks to generate correlation weights, pollutant data is filled into a three-dimensional grid to reconstruct the three-dimensional distribution of the atmospheric environment. Wind speed vector projection is calculated, transmission weights are established and combined with industrial load to generate effective transmission sequences. The directional driving mechanism of emission sources on monitoring points is quantified. Cross-correlation analysis is used to lock transmission lag parameters and combined with the real-time transmission rate evolution state transition matrix to achieve dynamic prediction of environmental conditions. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the workflow of the present invention; Figure 2 This is a flowchart of the meteorological and thermodynamic feature dataset acquisition process of the present invention; Figure 3 This is a flowchart of the process for obtaining the three-dimensional concentration distribution tensor record according to the present invention; Figure 4 This is a flowchart of the effective transmission load sequence record acquisition process of the present invention; Figure 5 This is a flowchart of the pollutant transport hysteresis parameter acquisition process of the present invention; Figure 6 This is a flowchart of the environmental state prediction data acquisition process of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0017] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0018] Please see Figure 1 This invention provides a technical solution for natural environment monitoring based on big data analysis, comprising the following steps: S1: Collect air pressure, temperature, horizontal wind speed component, wind direction angle, pollutant concentration sampling data and industrial power load data, calculate potential temperature vertical gradient parameters and atmospheric dynamic stability parameters, and generate meteorological and thermodynamic feature datasets. S2: Call the meteorological and thermodynamic feature dataset, compare the potential temperature vertical gradient parameter, atmospheric dynamic stability parameter and standard meteorological and physical benchmark record to generate correlation weight coefficients, fill the pollutant concentration sampling data with weights, and construct a three-dimensional concentration distribution tensor record; S3: Call the meteorological and thermal feature dataset, construct the spatial relative position vector, calculate the projection component of the horizontal wind speed component on the spatial relative position vector to generate the transmission weight coefficient, perform point-by-point multiplication operation on the industrial electricity load data and the transmission weight coefficient, and generate an effective transmission load sequence record. S4: Calculate the correlation parameter between the effective transport load sequence record and the pollutant concentration time series data at the corresponding monitoring point locations in the three-dimensional concentration distribution tensor record, search for the peak position of the correlation parameter, and construct the pollutant transport lag parameter. S5: Combine pollutant transport lag parameters and geometric displacement vectors to obtain real-time power transport rate factor, use the real-time power transport rate factor to correct the time step of the basic state transition probability matrix and generate a synthetic transition matrix, map the three-dimensional concentration distribution tensor record, and output environmental state prediction data.

[0019] The meteorological and thermodynamic feature dataset includes a potential temperature vertical distribution table, a dynamic stability feature vector, and a multi-dimensional meteorological element index. The three-dimensional concentration distribution tensor record specifically includes spatial grid concentration values, weighted filling confidence markers, and a set of vertical barrier state matrices. The effective transport load sequence record includes normalized wind field projection flux, industrial load time-series mapping values, and source point transport correlation factors. The pollutant transport lag parameters specifically refer to transport time delay values, interrelated peak offsets, and time window search confidence. The environmental state prediction data includes the concentration distribution tensor for the next preset time step, a regional diffusion evolution trend map, and a synthetic state transition probability table.

[0020] Please see Figure 2 The specific steps for obtaining the meteorological and thermal feature dataset are as follows: S111: Collect air pressure parameters, temperature parameters, horizontal wind speed components, wind direction angle parameters, and pollutant concentration sampling data at multiple altitude levels; calculate the thermodynamic correlation between air pressure parameters and temperature parameters; perform adiabatic compression state conversion processing; determine the potential temperature parameters corresponding to the air mass conservation properties at each monitoring point; spatially sequence and arrange data at different altitude levels; and construct a basic meteorological potential temperature sequence. For multi-dimensional environmental monitoring scenarios in complex environments such as chemical industrial parks, a Doppler lidar and phased array ultrasonic anemometer array were first deployed within the monitoring area with a horizontal grid spacing of 50 meters and a vertical layer spacing of 10 meters. Data acquisition was performed every 5 minutes, acquiring air pressure (hPa), air temperature (°C), horizontal wind speed (m / s), and wind direction (degrees) at 50 altitude levels from the ground to 500 meters. After acquiring the raw data, the Poisson equation was used to perform thermodynamic correlation calculations on the collected air pressure and temperature parameters. The air temperature at each altitude level was converted to potential temperature parameters at a standard reference pressure of 1000 hPa. Specifically, the ratio of the gas constant to the specific heat at isobaric pressure, 0.286, was introduced as an exponential term in the calculation to eliminate thermodynamic interference caused by the decrease in air pressure with altitude, thereby determining the conservation properties of the air mass at each monitoring point under the adiabatic movement assumption. Subsequently, the potential temperature values ​​calculated from each layer are spatially sequenced and arranged according to the height index (z=1 to z=50), and the wind speed vector and pollutant concentration (such as PM2.5 or SO2) sampling values ​​of each layer are simultaneously aligned. After data cleaning to remove null values ​​caused by equipment failure or outliers that deviate significantly from physical laws (such as negative concentration values), a basic meteorological potential temperature sequence containing timestamps, three-dimensional coordinates and thermal states is constructed.

[0021] S112: Call the basic meteorological potential temperature sequence, analyze the vertical distribution structure and retrieve the potential temperature parameters corresponding to adjacent height layers, calculate the physical difference of potential temperature parameters between adjacent layers, perform gradient calculation in combination with the height difference between layers, determine the thermal stratification state of atmospheric stratification and establish vertical stability characteristics, and construct the vertical potential temperature gradient vector. Using the aforementioned basic meteorological potential temperature sequence, the potential temperature parameters of adjacent vertical layers (such as layer z and layer z+1) are retrieved through a sliding window algorithm. The finite difference method is then employed to calculate the physical difference in potential temperature between adjacent layers. Specifically, the potential temperature of the upper layer is subtracted from the potential temperature of the lower layer and divided by the vertical distance between the layers (10 meters) to obtain the vertical gradient value of potential temperature. The thermal stratification state of the atmosphere is determined based on the sign and magnitude of the gradient value: if the gradient value is greater than 0.1℃ / 100m, it is determined to be a stable layer (inversion or isothermal state); if the gradient value is between -0.1℃ / 100m and 0.1℃ / 100m, it is determined to be a neutral layer; if the gradient value is less than -0.1℃ / 100m, it is determined to be an unstable layer (convective state). Based on this judgment logic, a vertical stability feature label is established for each pair of adjacent layers (e.g., 1 represents extremely stable, 0 represents neutral, and -1 represents extremely unstable). The gradient values ​​of all layers and the feature labels are encapsulated in vertical order to construct a vertical potential temperature gradient vector. This vector directly reflects the ability of the thermal structure above the monitoring area at the current moment to inhibit or promote vertical diffusion.

[0022] S113: Call the vertical potential temperature gradient vector, combine the horizontal wind speed component and vertical stability characteristics at the same location, calculate the atmospheric dynamic stability parameters, integrate the basic meteorological potential temperature sequence, wind direction angle and pollutant concentration data, perform multi-dimensional spatiotemporal correlation mapping, and generate a meteorological thermodynamic feature dataset. The atmospheric dynamic stability parameter is quantified by incorporating the vertical potential temperature gradient vector, combined with the horizontal wind speed component at the same altitude layer, and using the Richardson number (Ri) calculation logic. Specifically, the vertical potential temperature gradient is used as the numerator to represent the thermo-buoyancy term, and the square of the vertical wind speed shear is used as the denominator to represent the shear generation term. When the calculated Ri value is greater than 0.25, the flow field is considered laminar, and the atmospheric dynamic stability parameter is set to a high value (range 0.8-1.0); when the Ri value is less than 0.25, turbulence is considered enhanced, and the parameter is set to a low value (range 0.0-0.4). After calculation, this dynamic stability parameter is multidimensionally and spatiotemporally correlated with the basic meteorological potential temperature sequence, wind direction angle (0-360 degrees), and corresponding pollutant concentration data (unit: μg / m³). A hash mapping technique is used to bind all meteorological and environmental parameters under the same spatiotemporal coordinates, generating a structured meteorological and thermodynamic feature dataset. This provides a complete input source containing both thermodynamic and dynamic attributes for subsequent transport analysis.

[0023] Please see Figure 3 The specific steps for obtaining the three-dimensional concentration distribution tensor record are as follows: S211: Call the meteorological and thermodynamic feature dataset, call the standard meteorological reference record, extract the thermal stability reference and the shear turbulence reference, perform differential comparison operation on the potential temperature vertical gradient parameter and the thermal stability reference, perform interval determination operation on the atmospheric dynamic stability parameter and the shear turbulence reference, integrate the comparison results to quantify the degree of vertical flux obstruction, and generate a vertical obstruction status identifier. The system calls upon a meteorological and thermodynamic feature dataset and retrieves standard meteorological baseline records from a pre-built local database. These baseline records contain a thermal stability baseline (potential temperature gradient threshold range) and a shear turbulence baseline (wind shear index threshold range) derived from statistical analysis of historical data from the past five years for the region. Specifically, the thermal stability baseline is determined by performing a Gaussian normal distribution fitting operation on the historical potential temperature vertical gradient set and taking the mathematical expectation parameters of the fitted distribution curve. As the baseline of the neutral stratification, i.e., the benchmark median, the standard deviation statistical parameter is taken. As the allowable fluctuation amplitude, the specific method for determining the shear turbulence benchmark is as follows: Perform bimodal clustering statistical analysis on the historical vertical wind shear index set to identify the cluster centers of low and high shear states, and take the arithmetic mean boundary value of the two cluster centers as the dynamic shear threshold. During the comparison calculation, firstly, subtract the median of the thermal stability benchmark from the real-time calculated potential temperature vertical gradient parameter. If the difference exceeds twice the benchmark standard deviation, the thermal vertical barrier is considered significant. Simultaneously, perform an interval comparison between the real-time atmospheric dynamic stability parameters (such as Richardson's number) and the shear turbulence benchmark (such as a critical value of 0.25). Based on the potential temperature vertical gradient parameter and the atmospheric dynamic stability parameter... The combined values ​​are used to determine the vertical barrier status of the 5-level discretized system according to the following rules: If the potential temperature gradient shows a strong temperature inversion (such as gradient...), the vertical barrier status is determined by the following rules: ℃ / 100m) and the dynamic parameters show weak turbulence ( If the vertical blocking state is marked as "complete blockage" (Level 5), then the vertical blocking state is determined to be "complete blockage"; if the potential temperature gradient is in the moderate inversion range (gradient is in the range of...), then the vertical blocking state is marked as "complete blockage" (Level 5); ... ℃ / 100m to (between ℃ / 100m) and Between to If the potential temperature gradient is between [a certain value], it is classified as "strong barrier" (Level 4); if the potential temperature gradient is in the weakly stable range (gradient is between [a certain value]), it is classified as "strong barrier" (Level 4). ℃ / 100m to (between ℃ / 100m) and Between to If the potential temperature gradient is between [a certain value], it is determined to be "moderate barrier" (Level 3); if the potential temperature gradient is close to the neutral stratification (gradient is between [a certain value]), it is considered to be "moderate barrier" (Level 3). ℃ / 100m to (between ℃ / 100m) and If it is a "weak barrier" (Level 2), then it is judged as a superadiabatic lapse rate (gradient). ℃ / 100m) and strong turbulence ( If the value is not specified, it is determined to be "free diffusion" (Level 1). This step transforms the complex physical layering state into a 5-level discretized vertical flux obstruction level indicator through quantization comparison. The vertical obstruction state indicators of each layer at all spatial grid locations in the monitoring area are arranged and assembled according to three-dimensional spatial coordinates to construct a vertical obstruction state matrix, which is used to guide the weight allocation of subsequent three-dimensional interpolation.

[0024] S212: Based on the vertical barrier status indicator, analyze the inhibition strength of atmospheric stratification on pollutant diffusion, perform numerical quantification calculation on the inhibition strength, establish a vertical diffusion attenuation factor, construct the weight allocation relationship of spatial interpolation based on the vertical diffusion attenuation factor, calculate the normalized numerical index of the correlation between monitoring points, and generate correlation weight coefficients. Based on the generated vertical barrier status indicators, an exponential decay model is used to numerically quantify the atmospheric stratification suppression intensity. A vertical diffusion attenuation factor is defined. When the barrier status is Level 5, A value of 0.05 (indicating that 95% of vertical exchange is suppressed); when the blocking state is Level 4, the value is... (Indicates that 80% of vertical exchange is suppressed); when the blocking state is Level 3, the value is... (Indicates that 55% of vertical exchange is suppressed); when the blocking state is Level 2, the value is... (This indicates that 30% of vertical exchange is suppressed); when the state is Level 1, The value is 0.95. When constructing the weighting relationship for spatial interpolation, Gaussian weighting is used in the horizontal direction, and this is introduced in the vertical direction. Factor-adjusted weights. Specifically, when calculating the three-dimensional Euclidean distance between the point to be estimated and the sampling point, the vertical distance component is divided by... The vertical distance is artificially "increased" to reduce the weight contribution of adjacent points in the vertical direction during interpolation. Finally, the reciprocal of the corrected three-dimensional distance is taken and normalized to generate the correlation weight coefficients of each sampling point to be estimated interpolation points, ensuring that the interpolation results conform to the physical barrier characteristics of the current atmospheric stratification, rather than simply geometric proximity.

[0025] S213: Call the correlation weight coefficient, combine it with the meteorological and thermodynamic feature dataset, perform three-dimensional spatial gridded weighted filling operation on the pollutant concentration sampling data according to the correlation weight coefficient, calculate the concentration estimate of the non-sampling area in space, perform structured encapsulation and tensor definition on the entire field concentration data according to the spatial dimension, and establish a three-dimensional concentration distribution tensor record. Using the generated correlation weighting coefficients, a three-dimensional spatial gridded weighted filling process is performed on the sparse pollutant concentration sampling data (sensors only exist at specific altitudes) in the meteorological and thermodynamic feature dataset. A three-dimensional voxel grid covering the monitoring area is established with a length of 5 km, a width of 5 km, and a height of 500 m, with a resolution of 50 m × 50 m × 10 m. For each unsampled voxel center point, K (e.g., K=8) valid sampling points within its three-dimensional neighborhood are searched, and the aforementioned correlation weighting coefficients are applied for weighted summation to estimate the concentration of that voxel. During the weighted filling process, a weighted filling confidence flag is generated simultaneously for each spatial grid voxel: if actual sensor sampling data exists at the voxel location, its weighted filling confidence flag is set to 0. This indicates complete confidence; if the concentration value of the voxel is obtained by interpolation from surrounding sampling points, then the weighted calculation will be based on the values ​​of the voxels involved. The sum of the association weight coefficients of each valid sampling point is normalized and assigned a value between [value missing]. to The weighted fill confidence markers between the data points are used to calculate the interpolation results. A higher sum indicates a higher confidence level, meaning a stronger interpolation result, and vice versa. For example, under a strong temperature inversion layer, the weight of high-concentration ground-level sampling points on upper-layer voxels will be extremely low, causing the estimated upper-layer concentration to remain at the background value. Simultaneously, the weighted fill confidence marker for this upper-layer voxel will also have a low value, indicating greater uncertainty in its concentration estimation result. This thus reproduces the physical phenomenon of "pollutants being suppressed at the ground" at the data level. After completing the full field filling, the spatial grid concentration values ​​of all voxels are compared with their corresponding weighted fill confidence markers. The dimensions are assembled into a four-dimensional tensor structure, and the vertical barrier state matrix is ​​integrated simultaneously. This makes the three-dimensional concentration distribution tensor record contain a set of three parts of data: spatial grid concentration values, weighted filling confidence markers, and the vertical barrier state matrix. This establishes a three-dimensional concentration distribution tensor record and realizes the holographic digital reconstruction of the environmental field.

[0026] Please see Figure 4 The specific steps for obtaining the effective transmission payload sequence record are as follows: S311: Call the meteorological and thermodynamic feature dataset, parse the horizontal wind speed component and wind direction angle data, retrieve the geospatial coordinate information of industrial emission sources and environmental monitoring points, construct a spatial relative position vector from the emission source location to the monitoring point location, perform vector projection decomposition on the horizontal wind speed component and wind direction angle parameters in the direction of the spatial relative position vector, calculate the effective component intensity of the wind speed vector on the transmission path, and generate wind direction projection component parameters. When performing transmission path analysis, the real-time wind field information of each monitoring point is first analyzed using a meteorological and thermodynamic feature dataset. The coordinates of industrial emission sources (such as a chemical plant chimney) are then retrieved using a Geographic Information System (GIS). Coordinates of environmental monitoring points Construct a spatial relative position vector from the source to the monitoring point. For the horizontal wind speed vector at the monitoring point (Combined from wind speed and wind direction angle), calculate its spatial relative position vector. Projection in the direction. The specific operation is: calculate the wind speed vector. With position vector The dot product, i.e. ,in This represents the angle between the wind direction and the transmission path. A positive projection value indicates that the wind is blowing pollutants toward the monitoring point; a negative value indicates that the wind is blowing pollutants away. This step generates wind direction projection component parameters, accurately quantifying the effective contribution of the current wind field to the transmission at a specific point, rather than solely relying on wind speed.

[0027] S312: Based on the wind direction projection component parameters, analyze the geometric consistency characteristics between the wind speed projection direction and the pollutant transport path, perform normalization mapping operation according to the positive and negative polarity and magnitude of the projection component values, determine the dynamic driving and blocking state of the wind field on pollutant diffusion under the current meteorological conditions, construct a quantitative weight index for the directional transport efficiency of pollutants, and generate normalized transport weight coefficients. Based on the wind direction projection component parameters, its geometric consistency characteristics are further analyzed. To eliminate the influence of different wind speed levels, a normalization mapping is performed on the projection components. A variant of the Sigmoid function is used for mapping, mapping the wind direction projection components (unit: m / s) to the [-1, 1] interval. The specific algorithm is as follows: Input projection components Output When the wind speed projection is positive and large, A value close to 1 indicates extremely strong positive transmission drive; when the projection is negative and very large, A value close to -1 indicates a strong cleaning and removal effect; when the projection is close to 0, A value close to 0 indicates no direct transmission relationship. The normalized transmission weight coefficients generated in this step essentially construct a direction-sensitive physical filter, capable of distinguishing the essential differences between upwind and downwind monitoring points when affected by industrial sources.

[0028] S313: Obtain time-series data of industrial electricity load in the factory, and map it into a non-steady-state emission source strength sequence based on the transformation relationship between energy consumption and pollutant generation. Perform point-by-point multiplication of the non-steady-state emission source strength sequence with the normalized transmission weight coefficient in the time dimension, and take into account the correction effect of meteorological transmission efficiency on source strength emissions to generate an effective transmission load sequence record. To quantify the actual environmental impact of industrial sources, time-series data of industrial electricity load (unit: kW, sampling interval 15 minutes) for corresponding factories were obtained. Based on a pre-calibrated energy-material conversion model, the electricity load was mapped to a non-steady-state emission source intensity sequence (unit: g / s). For example, for an electrolytic aluminum workshop, it was set that every 1000 kWh of electricity consumed corresponds to 1.2 kg of particulate matter emissions. Subsequently, this source intensity sequence was multiplied point-by-point with the aforementioned normalized transmission weighting coefficients. At time points... Effective transmission load = emission source strength ×Transmission weighting coefficient If a factory is operating at high capacity but is downwind (with a negative weighting) at a certain moment, the effective transmission load will be corrected to a negative value or zero, reflecting that although there are emissions at that moment, they are not transmitted to the monitoring point; conversely, if it is upwind and operating at high capacity, the value will be significantly enhanced. The effective transmission load sequence record generated by this process realizes the physical modulation of source strength data by meteorological conditions.

[0029] Please see Figure 5 The specific steps for obtaining pollutant transport lag parameters are as follows: S411: Call the three-dimensional concentration distribution tensor record, extract pollutant concentration time series data based on the spatial index of monitoring points, obtain the preset environmental background noise baseline, perform baseline difference processing on the pollutant concentration time series data, and generate net value concentration fluctuation sequence. Time-series data of pollutant concentrations at target monitoring points were extracted from the three-dimensional concentration distribution tensor record. Since the raw environmental data contained significant background noise from non-industrial sources such as traffic and dust, baseline differencing was performed first. The average concentration during the period of 3:00-4:00 AM daily (typically when industrial activity and traffic are weakest) was selected as the environmental background noise baseline (e.g., 35 μg / m³). This baseline value was subtracted from the total daily monitoring data, and values ​​below the baseline were set to zero, thus generating a net concentration fluctuation sequence that primarily reflects the characteristics of sudden industrial emissions. This sequence removed constant term drift and highlighted the concentration pulse signals caused by industrial production fluctuations, laying the foundation for high-sensitivity correlation analysis in terms of signal-to-noise ratio.

[0030] S412: Call the effective transmission load sequence record, calculate the energy integral of the environmental background noise baseline, obtain the estimated background noise energy, calculate the autocorrelation energy of the net concentration fluctuation sequence, obtain the total signal energy value, and set candidate time displacement variables within the standard cross-correlation search window parameter range, using the formula: ; Calculate the sequence dynamic coupling degree between the net value concentration fluctuation sequence and the effective transmission load sequence, and generate a time-shifted correlation strength distribution set; in, For the dynamic coupling degree of the sequence, Candidate time displacement variables were obtained by iterating through the standard cross-correlation search window parameters by step size. The sampling sequence length is obtained by counting the total number of data points in the net value concentration fluctuation sequence. The index for discrete sampling times is obtained by discretizing and numbering the time series. For the net asset value concentration fluctuation series values, the first value is extracted by... The concentration difference results at each time point were obtained. To effectively transmit the payload sequence values, by extracting the first... The load projection data at each time point is obtained. The estimated background noise energy value is obtained by performing a square integral on the baseline of the environmental background noise. The total energy value of the signal is obtained by summing the autocorrelation squares of the net value concentration fluctuation sequence. The effective transport load sequence and the net concentration fluctuation sequence were retrieved, and their dynamic coupling degree was calculated using a specific formula to identify the transport lag time. The standard cross-correlation search window was set to [0, 120] minutes, assuming the pollutant transport time did not exceed 2 hours. For the experimental data segment shown in Table 1 (with 10 minutes as the simplified discrete time index i, K=6), the time shift was calculated... Coupling degree at a lag of 10 minutes: Table 1. Basic data table for calculating the dynamic coupling degree of sequences: ; First, calculate the estimated value of background noise energy. If the background baseline is 5, then for Calculate the total energy value of the signal. ,Right now Sum of squares: .calculate Cross-correlation terms at time: and Align and multiply. Because... It does not exist, therefore from Start effective accumulation. Calculate the norm in the denominator for each: ; (Take the effective overlapping portion and pad with zeros) Then the normalized projection term (cosine similarity) = Substitute into the formula to calculate the coupling degree: The result shows that, after deducting noise interference ( After using the term as a penalty factor, the causal confidence with a lag of one time step is extremely high. If the signal-to-noise ratio is low ( (very small) The item will increase significantly, leading to A sharp drop, thus filtering out spurious correlations. This is achieved by iterating through all windows... Generate a time-shifted correlation strength distribution set.

[0031] S413: Retrieve the global peak point in the distribution set of time-shift correlation intensity, lock the time offset index corresponding to the global peak point, determine that the target index is a time delay feature, and convert the time delay feature into a physical time value in combination with the sampling frequency to construct pollutant transmission lag parameters for the target industrial emission source. Retrieve the global peak point in the distribution set of time-shift correlation strength. Assume it is within the search range. (corresponding to 40 minutes) If the value reaches its maximum of 0.96, then 40 minutes is identified as the time delay characteristic. Combined with the sampling frequency, the pollutant transport lag parameter of this monitoring point relative to the industrial source is confirmed to be 40 minutes. This parameter not only represents the physical transport time but also implicitly contains information about the current average wind speed and path tortuosity. If multiple peaks occur, the value is taken as... The point that is the largest and most significant (more than 20% above the second peak) is used. If there is no significant peak, a "very weak correlation" signal is output, and subsequent inversion is stopped to avoid forced prediction when there is no causal relationship.

[0032] Please see Figure 6 The specific steps for obtaining environmental state prediction data are as follows: S511: Call the pollutant transport lag parameter, calculate the geographical distance scalar of the monitoring point relative to the industrial emission source, perform rate calculation based on the ratio of the geographical distance scalar to the pollutant transport lag parameter, establish the benchmark diffusion rate parameter, obtain the current average wind speed parameter of the monitoring area, perform a division comparison operation on the current average wind speed parameter of the monitoring area and the benchmark diffusion rate parameter, and generate a real-time power transmission rate factor. The baseline diffusion rate parameter is calculated by inverting the obtained pollutant transport lag parameter (e.g., 40 minutes, or 2400 seconds) with the straight-line geographical distance from the monitoring point to the emission source (e.g., 12000 meters). At the same time, the current average wind speed parameters of the monitoring area are obtained from the meteorological monitoring network. (e.g., 7.5 m / s). Calculate the real-time power delivery rate factor. This factor indicates that the current actual wind speed is 1.5 times the historical average transmission rate, meaning that pollutants will reach downstream 50% faster than historically predicted. This factor will serve as a core driving variable, used to adjust the evolution step size of the state transition matrix and address the problem that static statistical models cannot adapt to dynamic wind speed changes.

[0033] S512: Call the three-dimensional concentration distribution tensor record, perform global discrete statistical analysis on the concentration level fluctuation sequence, obtain the state change frequency characteristics between adjacent discrete moments, normalize the data and construct the basic state transition probability matrix, perform quantization and rounding on the real-time power transmission rate factor, obtain the integer step level, using the formula: ; Calculate the state transition probability of the synthesis concentration and construct the synthesis transition matrix; in, The state transition probabilities for the synthetic concentration are obtained by power-law evolution and linear interpolation of the basic state transition probability matrix. The basic state transition probabilities are obtained by performing normalized statistics on the state change frequencies in the three-dimensional concentration distribution tensor record. The integer step level is obtained by rounding down the real-time power delivery rate factor. The identity matrix represents the probability of maintaining the state and serves as a benchmark. The current average wind speed parameter for the monitoring area is obtained in real time through meteorological monitoring equipment. The baseline diffusion rate parameter is obtained by inversion using the ratio of the geographical interval scalar to the pollutant transport hysteresis parameter; The system calls upon the three-dimensional concentration distribution tensor record and performs global discrete statistical analysis on the concentration level fluctuation sequence contained therein. First, it defines the discrete interval of concentration state, mapping continuous pollutant concentration values ​​to integer levels from 1 to 5 (level 1 represents excellent pollution, level 5 represents severe pollution). Then, it iterates through the historical time series, statistically analyzing adjacent discrete moments (…). to The frequency of state transitions is calculated by dividing the frequency of transitions by the total frequency of occurrence of that state, and then normalizing the result to construct the basic state transition probability matrix. For example, statistical data shows the probability of maintaining a "Level 1" weather pattern under current meteorological conditions. The probability of changing from "Level 1" to "Level 2" is 0.8. The initial value is 0.2, and the transition probabilities for other states are calculated similarly. Subsequently, the real-time power delivery rate factor is obtained. (1.5). Perform numerical decomposition on the factor and round down to obtain the integer step level. Using the formula Calculate the decimal weight hierarchy and substitute it into the previous... , and , and thus At this point, the formula is used. Calculate the state transition probability at the synthesis concentration. In the specific calculation process, first construct the identity matrix. Its diagonal elements are 1s, and the rest are 0s. Calculate the difference matrix. For element position The difference value is For element position The difference value is Next, calculate the linear interpolation term. Regarding location The calculation result is Regarding location The calculation result is This interpolation term matrix essentially constructs an intermediate state operator between "maintaining the status quo" and "evolving one step." Finally, the fundamental matrix... power (here) Perform matrix multiplication with the interpolation term matrix. The resulting matrix is ​​then multiplied. first row and first column element For example, its value is determined by The dot product of the first row vector and the first column vector of the interpolation matrix is ​​obtained, i.e. Through full matrix operations, a synthesis transfer matrix capable of accurately representing a diffusion step size of 1.5 times the standard size was finally constructed. The values ​​in this matrix directly reflect the dynamic transition probability distribution of each pollutant concentration level at the next moment under the current enhanced wind speed.

[0034] S513: Extract the current state vector from the three-dimensional concentration distribution tensor record, map the current state vector to the synthetic transition matrix, perform matrix multiplication to deduce the state distribution probability, integrate the deduction results of the spatiotemporal dimensions, and output environmental state prediction data. Extracting the current time from the three-dimensional concentration distribution tensor record ( The global state vector (Concentration levels for all grid points). Map this vector to the synthesis transition matrix. Perform matrix multiplication: This calculation outputs the probability distribution of each grid point's position at various concentration levels in the future. The level with the highest probability is selected as the prediction result, and spatial dimensional information is integrated to output three-dimensional environmental state prediction data covering the entire field. This prediction data not only includes the numerical distribution of pollutant concentrations but also includes prediction confidence levels calculated based on the probability distribution. Finally, it is presented to park managers as a dynamic heatmap via a visualization terminal to guide the formulation of emergency emission reduction measures.

[0035] A natural environment monitoring system based on big data analytics is used to execute the aforementioned natural environment monitoring method based on big data analytics. The system includes: The data acquisition and calculation module collects air pressure, temperature, horizontal wind speed components, wind direction angle, pollutant concentration sampling data and industrial power load data, calculates potential temperature vertical gradient parameters and atmospheric dynamic stability parameters, and generates meteorological and thermodynamic characteristic datasets. The concentration-weighted construction module calls the meteorological and thermodynamic feature dataset, compares the potential temperature vertical gradient parameter, atmospheric dynamic stability parameter and standard meteorological and physical benchmark record to generate correlation weight coefficients, and fills the pollutant concentration sampling data with weights to construct a three-dimensional concentration distribution tensor record. The transmission load calculation module calls the meteorological and thermal feature dataset, constructs a spatial relative position vector, calculates the projection component of the horizontal wind speed component on the spatial relative position vector to generate the transmission weight coefficient, and performs point-by-point multiplication operation on the industrial electricity load data and the transmission weight coefficient to generate an effective transmission load sequence record. The lag parameter extraction module calculates the correlation parameters between the effective transport load sequence record and the pollutant concentration time series data at the corresponding monitoring point locations in the three-dimensional concentration distribution tensor record, searches for the peak positions of the correlation parameters, and constructs pollutant transport lag parameters. The prediction data generation module combines pollutant transport lag parameters and geometric displacement vectors to obtain real-time power transport rate factors. It then uses these real-time power transport rate factors to correct the time step of the basic state transition probability matrix and generate a synthetic transition matrix. This matrix maps the three-dimensional concentration distribution tensor record and outputs environmental state prediction data.

[0036] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A natural environment monitoring method based on big data analysis, characterized in that, Includes the following steps: S1: Collect air pressure, temperature, horizontal wind speed component, wind direction angle, pollutant concentration sampling data and industrial power load data, calculate potential temperature vertical gradient parameters and atmospheric dynamic stability parameters, and generate meteorological and thermodynamic feature datasets. S2: Call the meteorological and thermodynamic feature dataset, compare the potential temperature vertical gradient parameter, atmospheric dynamic stability parameter and standard meteorological and physical benchmark record to generate correlation weight coefficients, fill the pollutant concentration sampling data with weights, and construct a three-dimensional concentration distribution tensor record; S3: Call the meteorological and thermal feature dataset, construct a spatial relative position vector, calculate the projection component of the horizontal wind speed component on the spatial relative position vector to generate a transmission weight coefficient, and perform point-by-point multiplication operation on the industrial electricity load data and the transmission weight coefficient to generate an effective transmission load sequence record. S4: Calculate the correlation parameter between the pollutant concentration time series data at the corresponding monitoring point locations of the effective transmission load sequence record and the three-dimensional concentration distribution tensor record, search for the peak position of the correlation parameter, and construct the pollutant transmission lag parameter. S5: Combine the pollutant transport lag parameter and geometric displacement vector to obtain the real-time power transport rate factor, use the real-time power transport rate factor to correct the time step of the basic state transition probability matrix and generate a synthetic transition matrix, map the three-dimensional concentration distribution tensor record, and output environmental state prediction data.

2. The natural environment monitoring method based on big data analysis according to claim 1, characterized in that, The meteorological and thermodynamic feature dataset includes a potential temperature vertical distribution table, a dynamic stability feature vector, and a multidimensional meteorological element index. The three-dimensional concentration distribution tensor record specifically consists of a set of spatial grid concentration values, weighted filling confidence markers, and vertical barrier state matrices. The effective transmission load sequence record includes normalized wind field projection flux, industrial load time-series mapping values, and source point transmission correlation factors. The pollutant transmission lag parameters specifically refer to transmission time delay values, interrelated peak offsets, and time window search confidence. The environmental state prediction data includes a concentration distribution tensor for the next preset time step, a regional diffusion evolution trend map, and a synthetic state transition probability table.

3. The natural environment monitoring method based on big data analysis according to claim 1, characterized in that, The specific steps for obtaining the meteorological and thermodynamic feature dataset are as follows: S111: Collect air pressure parameters, temperature parameters, horizontal wind speed components, wind direction angle parameters, and pollutant concentration sampling data at multiple altitude levels; calculate the thermodynamic correlation between air pressure parameters and temperature parameters; perform adiabatic compression state conversion processing; determine the potential temperature parameters corresponding to the air mass conservation properties at each monitoring point; spatially sequence and arrange data at different altitude levels; and construct a basic meteorological potential temperature sequence. S112: Call the basic meteorological potential temperature sequence, analyze the vertical distribution structure and retrieve the potential temperature parameters corresponding to adjacent height layers, calculate the physical difference of potential temperature parameters between adjacent layers, perform gradient calculation in combination with the height difference between layers, determine the atmospheric stratification thermal stratification state and establish vertical stability characteristics, and construct a vertical potential temperature gradient vector. S113: Call the vertical potential temperature gradient vector, combine it with the horizontal wind speed component and vertical stability characteristics at the same location, calculate the atmospheric dynamic stability parameters, integrate the basic meteorological potential temperature sequence, wind direction angle and pollutant concentration data, perform multi-dimensional spatiotemporal correlation mapping, and generate a meteorological thermodynamic feature dataset.

4. The natural environment monitoring method based on big data analysis according to claim 3, characterized in that, The specific steps for obtaining the three-dimensional concentration distribution tensor record are as follows: S211: Call the meteorological and thermodynamic feature dataset, call the standard meteorological reference record, extract the thermal stability reference and the shear turbulence reference, perform differential comparison operation on the potential temperature vertical gradient parameter and the thermal stability reference, perform interval determination operation on the atmospheric dynamic stability parameter and the shear turbulence reference, integrate the comparison results to quantify the degree of vertical flux obstruction, and generate a vertical obstruction status identifier. S212: Based on the vertical barrier status indicator, analyze the inhibition strength of atmospheric stratification on pollutant diffusion, perform numerical quantification calculation on the inhibition strength, establish a vertical diffusion attenuation factor, construct a spatial interpolation weight allocation relationship based on the vertical diffusion attenuation factor, calculate the normalized numerical index of the correlation degree between monitoring points, and generate correlation weight coefficients. S213: Call the associated weight coefficient, combine it with the meteorological and thermodynamic feature dataset, perform a three-dimensional spatial gridded weighted filling operation on the pollutant concentration sampling data according to the associated weight coefficient, calculate the estimated concentration value of the non-sampling area in space, perform structured encapsulation and tensor definition on the entire field concentration data according to the spatial dimension, and establish a three-dimensional concentration distribution tensor record.

5. The natural environment monitoring method based on big data analysis according to claim 4, characterized in that, The specific process for obtaining the standard meteorological baseline record is as follows: The system analyzes vertical stratified records of air pressure, temperature, and wind speed within the historical statistical period of the target monitoring area. It calculates the historical potential temperature vertical gradient set and performs Gaussian normal distribution fitting to obtain the expected mathematical parameters and standard deviation statistical parameters of the fitted distribution curve. These are set as the neutral stratification baseline and allowable fluctuation amplitude, respectively, to establish the thermal stability benchmark. The system calculates the historical vertical wind shear index set and performs bimodal clustering statistical analysis to identify cluster centers in low-shear and high-shear states. It calculates the arithmetic mean boundary value of the two cluster centers as the dynamic shear threshold to establish the shear turbulence benchmark. The system integrates the thermal stability benchmark and the shear turbulence benchmark to establish the standard meteorological benchmark record analysis. It identifies cluster centers in low-shear and high-shear states, calculates the arithmetic mean boundary value of the two cluster centers, and determines the arithmetic mean boundary value as the dynamic shear threshold for fluid state transitions to establish the shear turbulence benchmark. Finally, it integrates the thermal stability benchmark and the shear turbulence benchmark to establish the standard meteorological benchmark record.

6. The natural environment monitoring method based on big data analysis according to claim 4, characterized in that, The specific steps for obtaining the effective transmission payload sequence record are as follows: S311: Call the meteorological and thermodynamic feature dataset, parse the horizontal wind speed component and wind direction angle data, retrieve the geospatial coordinate information of industrial emission sources and environmental monitoring points, construct a spatial relative position vector from the emission source location to the monitoring point location, perform vector projection decomposition on the horizontal wind speed component and wind direction angle parameters in the direction of the spatial relative position vector, calculate the effective component intensity of the wind speed vector on the transmission path, and generate wind direction projection component parameters. S312: Based on the wind direction projection component parameters, analyze the geometric consistency characteristics between the wind speed projection direction and the pollutant transport path, perform normalization mapping operation according to the positive and negative polarity and magnitude of the projection component values, determine the dynamic driving and blocking state of the wind field on pollutant diffusion under the current meteorological conditions, construct a quantitative weight index for the directional transport efficiency of pollutants, and generate normalized transport weight coefficients. S313: Obtain time-series data of industrial electricity load in the factory, and map it into a non-steady-state emission source strength sequence based on the transformation relationship between energy consumption and pollutant generation. Perform point-by-point multiplication of the non-steady-state emission source strength sequence with the normalized transmission weight coefficient in the time dimension, and generate an effective transmission load sequence record by taking into account the correction effect of meteorological transmission efficiency on source strength emissions.

7. The natural environment monitoring method based on big data analysis according to claim 6, characterized in that, The specific steps for obtaining the pollutant transport hysteresis parameter are as follows: S411: Call the three-dimensional concentration distribution tensor record, extract pollutant concentration time series data according to the monitoring point spatial index, obtain the preset environmental background noise baseline, perform baseline difference processing on the pollutant concentration time series data, and generate net value concentration fluctuation sequence. S412: Call the effective transmission load sequence record, calculate the energy integral of the environmental background noise baseline, obtain the estimated value of the background noise energy, calculate the autocorrelation energy of the net value concentration fluctuation sequence, obtain the total signal energy value, set the candidate time displacement variable within the standard cross-correlation search window parameter range, calculate the sequence dynamic coupling degree between the net value concentration fluctuation sequence and the effective transmission load sequence record, and generate a time-shifted correlation strength distribution set; S413: Retrieve the global peak point in the time-shift correlation intensity distribution set, lock the time offset index corresponding to the global peak point, determine that the target index is a time delay feature, and convert the time delay feature into a physical time value in combination with the sampling frequency to construct pollutant transmission lag parameters for the target industrial emission source.

8. The natural environment monitoring method based on big data analysis according to claim 7, characterized in that, The specific steps for obtaining the environmental state prediction data are as follows: S511: Call the pollutant transport lag parameter, calculate the geographical distance scalar of the monitoring point relative to the industrial emission source, perform rate calculation based on the ratio of the geographical distance scalar to the pollutant transport lag parameter, establish a benchmark diffusion rate parameter, obtain the current average wind speed parameter of the monitoring area, perform a division comparison operation on the current average wind speed parameter of the monitoring area and the benchmark diffusion rate parameter, and generate a real-time power transmission rate factor. S512: Call the three-dimensional concentration distribution tensor record, perform global discrete statistical analysis on the concentration level fluctuation sequence, obtain the state change frequency characteristics between adjacent discrete moments, normalize and construct the basic state transition probability matrix, perform quantization and rounding on the real-time power transmission rate factor, obtain the integer step level, calculate the synthetic concentration state transition probability, and construct the synthetic transition matrix. S513: Extract the current state vector from the three-dimensional concentration distribution tensor record, map the current state vector to the synthetic transition matrix, perform matrix multiplication to deduce the state distribution probability, integrate the deduction results of the spatiotemporal dimensions, and output environmental state prediction data.

9. A natural environment monitoring system based on big data analysis, characterized in that, The system is used to implement the natural environment monitoring method based on big data analysis as described in any one of claims 1-8, and the system comprises: The data acquisition and calculation module collects air pressure, temperature, horizontal wind speed components, wind direction angle, pollutant concentration sampling data and industrial power load data, calculates potential temperature vertical gradient parameters and atmospheric dynamic stability parameters, and generates meteorological and thermodynamic characteristic datasets. The concentration-weighted construction module calls the meteorological and thermodynamic feature dataset, compares the potential temperature vertical gradient parameter, atmospheric dynamic stability parameter and standard meteorological and physical benchmark record to generate correlation weight coefficients, and fills the pollutant concentration sampling data with weights to construct a three-dimensional concentration distribution tensor record. The transmission load calculation module calls the meteorological and thermal feature dataset, constructs a spatial relative position vector, calculates the projection component of the horizontal wind speed component on the spatial relative position vector to generate a transmission weight coefficient, and performs point-by-point multiplication operation on the industrial electricity load data and the transmission weight coefficient to generate an effective transmission load sequence record. The lag parameter extraction module calculates the correlation parameters between the pollutant concentration time series data at the corresponding monitoring point locations of the effective transmission load sequence record and the three-dimensional concentration distribution tensor record, searches for the peak position of the correlation parameters, and constructs pollutant transmission lag parameters. The prediction data generation module combines the pollutant transport lag parameter and geometric displacement vector to obtain the real-time power transport rate factor, uses the real-time power transport rate factor to correct the time step of the basic state transition probability matrix and generate a synthetic transition matrix, maps the three-dimensional concentration distribution tensor record, and outputs environmental state prediction data.

Images