CN117826112B - Soil water content inversion method based on sar - Google Patents
Soil water content inversion method based on sar Download PDFInfo
- Publication number
- CN117826112B CN117826112B CN202410245311.8A CN202410245311A CN117826112B CN 117826112 B CN117826112 B CN 117826112B CN 202410245311 A CN202410245311 A CN 202410245311A CN 117826112 B CN117826112 B CN 117826112B
- Authority
- CN
- China
- Prior art keywords
- soil
- data
- model
- sar
- particle size
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 239000002689 soil Substances 0.000 title claims abstract description 193
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 67
- 238000000034 method Methods 0.000 title claims abstract description 39
- 239000002245 particle Substances 0.000 claims abstract description 126
- 230000010287 polarization Effects 0.000 claims abstract description 44
- 238000005457 optimization Methods 0.000 claims abstract description 36
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 19
- 238000012937 correction Methods 0.000 claims abstract description 17
- 238000004458 analytical method Methods 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 10
- 230000001629 suppression Effects 0.000 claims abstract description 8
- 238000009826 distribution Methods 0.000 claims description 61
- 230000003746 surface roughness Effects 0.000 claims description 32
- 230000000694 effects Effects 0.000 claims description 13
- 238000005259 measurement Methods 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000005315 distribution function Methods 0.000 claims description 12
- 238000012544 monitoring process Methods 0.000 claims description 11
- 230000006870 function Effects 0.000 claims description 9
- 230000008569 process Effects 0.000 claims description 9
- 230000008859 change Effects 0.000 claims description 8
- 238000005516 engineering process Methods 0.000 claims description 7
- 238000012876 topography Methods 0.000 claims description 7
- 239000005416 organic matter Substances 0.000 claims description 6
- 230000001360 synchronised effect Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 238000013480 data collection Methods 0.000 claims description 4
- 238000012512 characterization method Methods 0.000 claims description 3
- 230000009977 dual effect Effects 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 238000003384 imaging method Methods 0.000 claims description 3
- 230000007774 longterm Effects 0.000 claims description 3
- 238000010801 machine learning Methods 0.000 claims description 3
- 230000014759 maintenance of location Effects 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 claims description 3
- 238000002360 preparation method Methods 0.000 claims description 3
- 230000005855 radiation Effects 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 230000017260 vegetative to reproductive phase transition of meristem Effects 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 abstract description 2
- 238000012795 verification Methods 0.000 description 6
- 238000007726 management method Methods 0.000 description 4
- 230000007613 environmental effect Effects 0.000 description 3
- 238000012800 visualization Methods 0.000 description 3
- 239000004927 clay Substances 0.000 description 2
- 230000001427 coherent effect Effects 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000012271 agricultural production Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013079 data visualisation Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000001932 seasonal effect Effects 0.000 description 1
- 238000004441 surface measurement Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Abstract
The invention relates to the technical field of soil data analysis, in particular to an inversion method of soil water content based on sar, which comprises the following steps: collecting multi-time sequence and multi-polarization SAR data; preprocessing the collected SAR data, wherein the preprocessing comprises radiometric calibration, speckle noise suppression, terrain correction and registration of multi-time sequence data, and performing polarization decomposition on the multi-polarization data to extract characteristic information related to soil humidity; establishing a mathematical relationship model between SAR data and soil water content by applying soil-vegetation analysis; carrying out parameter optimization on the data relationship model by using an optimization algorithm, and improving inversion accuracy; introducing an inversion model, and converting the processed SAR data into soil water content information; the soil moisture content data is visualized to provide a spatial profile. According to the invention, the model parameters are optimized through a Particle Swarm Optimization (PSO) algorithm, so that the estimation accuracy is further improved.
Description
Technical Field
The invention relates to the technical field of soil data analysis, in particular to an inversion method of soil water content based on sar.
Background
The water content of the soil is a key factor affecting agricultural production, water resource management, weather forecast and environmental monitoring, and accurate monitoring of the water content of the soil is important for improving agricultural yield, realizing sustainable utilization of water resources, responding to climate change in time and protecting ecological environment.
Traditionally, monitoring of soil moisture content has relied on surface measurement techniques, which are often labor intensive, costly, and difficult to cover wide and remote areas.
Synthetic Aperture Radar (SAR) technology is a powerful tool for monitoring soil moisture content because it can penetrate cloud layers and is independent of lighting conditions, and SAR technology can provide large-scale, high-resolution surface data, and is suitable for continuous and dynamic environmental monitoring.
However, the conversion of SAR data to accurate soil moisture content information presents a number of challenges. Factors such as soil moisture, type, surface roughness, and vegetation coverage affect the SAR signal, complicating inversion of soil moisture from SAR data.
Existing SAR-based soil moisture inversion techniques often rely on specific physical models or empirical formulas, which may not be flexible or accurate enough to handle complex surface conditions and diverse soil types, and in addition, the prior art has limitations in terms of data processing and visualization, and lacks an efficient method to convert complex SAR data into intuitive and understandable soil moisture information.
Disclosure of Invention
Based on the purposes, the invention provides an inversion method of soil water content based on sar.
An inversion method of soil water content based on sar comprises the following steps:
S1: collecting multi-time sequence and multi-polarization SAR data;
s2: preprocessing the collected SAR data, wherein the preprocessing comprises radiometric calibration, speckle noise suppression, terrain correction and registration of multi-time sequence data, and performing polarization decomposition on the multi-polarization data to extract characteristic information related to soil humidity;
S3: establishing a mathematical relationship model between SAR data and soil water content by applying soil-vegetation analysis;
s4: carrying out parameter optimization on the data relationship model by using an optimization algorithm, and improving inversion accuracy;
s5: introducing an inversion model, and converting the processed SAR data into soil water content information;
S6: and by combining with a GIS technology, the soil water content data is visualized to provide a spatial distribution map.
Further, the S1 specifically includes:
Determining data acquisition parameters: selecting Sentinel-1 as a SAR satellite system for monitoring the water content of soil, and setting a frequency band of SAR imaging as a C band;
Setting a polarization mode to capture comprehensive information of soil surface roughness, wherein the polarization mode comprises a single polarization (HH or VV), a dual polarization (HH+HV or VV+VH) or a full polarization mode;
Planning time sequence data collection: planning data acquisition time points based on a climate mode and a vegetation growth period of a target area, and collecting data before and after a rainy season and in a seedling stage, a flowering stage, a setting stage and a maturing stage of vegetation growth;
arranging for data acquisition at desired points in time using satellite reservation functions to form a long-term, consistent time series;
coordinating ground synchronous measurements: ground synchronous measurements are made while the satellite is in transit, including in-field recordings of soil texture, vegetation type and coverage parameters.
Further, the preprocessing in S2 specifically includes:
Radiation calibration: performing radiometric calibration on the original SAR data to convert the original signal into a comparable reflectance value, and quantitatively analyzing the data by using calibration parameters provided by satellites to ensure that the data of different time and different sensors are comparable;
speckle noise suppression: due to the coherent nature of SAR data, images are often affected by speckle noise. Filtering the image by using a Lee filter or a frame filter to reduce noise and preserve details of the image;
terrain correction: for areas with large topography relief, a Digital Elevation Model (DEM) is used for topography correction so as to eliminate geometrical distortion caused by topography, and through correction, pixels in an image are ensured to accurately correspond to the positions of the ground, particularly in slopes and mountain areas;
Registration of multi-temporal data: the SAR data acquired at different times are accurately registered, so that the spatial consistency of multi-time sequence data is ensured, and a pixel-to-pixel registration method is adopted to reduce space-time dislocation;
Polarization decomposition: and carrying out polarization decomposition on the multi-polarization SAR data, extracting the characteristic of sensitivity to soil humidity by using an H/A/Alpha decomposition method, and analyzing different polarization channels and polarization decomposition parameters including phase information and a scattering mechanism so as to identify the change of the soil humidity.
Feature extraction and analysis: based on the result of polarization decomposition, characteristic parameters related to soil humidity are extracted, including scattering intensity, polarization ratio and coherence.
Further, the soil-vegetation analysis in the step S3 considers the influence of soil humidity and vegetation coverage, integrates soil types and organic matter contents, and adopts a physical basis method to simulate the relation between SAR signals and soil humidity, and simultaneously considers the influence of soil particle size distribution, surface roughness and vegetation parameters on the SAR signals;
Data-driven model optimization is also included: training and optimizing a data relationship model by using a machine learning algorithm, and combining the preprocessed SAR data and ground measured soil water content data, so as to train the model to identify a complex nonlinear relationship.
Further, the mathematical relationship model framework is as follows:
Relationship between soil moisture and SAR signal: the relationship between soil moisture and SAR backscatter coefficients is described using a classical Oh model, which is used to estimate the effect of soil moisture on backscatter coefficients, expressed as:
Wherein/> Is the backscattering coefficient,/>Is the incident angle,/>Is the soil moisture (water content),Is a soil roughness parameter;
consider soil type and organic content: expanding an Oh model, adding the influence of soil type and organic matter content, and expressing the expanded model as follows:
Wherein/> Representing soil type (clay, sandy soil)/>Represents the organic content;
Effect of vegetation coverage: using vegetation parameters (vegetation moisture content, vegetation height, coverage) to simulate the effect of vegetation on SAR signals, introducing a vegetation correction factor to account for scattering effects of the vegetation layer, expressed as:
wherein/> Representing vegetation parameters;
Consider soil particle size distribution and surface roughness: based on actual experimental data, an empirical relationship between soil particle size distribution and surface roughness and backscattering coefficient is established, and the lognormal distribution in the particle size distribution function is used to characterize the soil surface and is incorporated into the expanded model.
Further, the characteristic of the soil surface using the lognormal distribution in the particle size distribution function specifically includes:
Characterization of soil particle size distribution:
selection of a lognormal distribution: the soil particle size is in lognormal distribution, namely the natural log of the soil particle size is in normal distribution, and the formula of lognormal distribution is as follows:
Wherein/> Is particle size/>Probability density of/>And/>The mean and standard deviation of the logarithmic particle size, respectively;
Soil particle size distribution was incorporated into the model: the particle size distribution influences the dielectric property and the surface roughness of soil, further influences SAR signals, and adjusts the soil roughness parameters To take into account the influence of the particle size distribution as a function/>Expression of/>, whereinSoil particle size defined by lognormal distribution;
relation of particle size distribution to SAR signal:
relationship between particle size and dielectric constant: the soil particle size affects the retention of the water content, and thus the dielectric constant, and an empirical formula is used to describe the relationship between particle size, water content and dielectric constant;
Relationship between particle size and surface roughness: the change of the particle size of the soil influences the surface roughness, the larger the particle size is, the coarser the surface is, and the average value and the variance of the surface roughness are calculated according to a particle size distribution function;
incorporating the particle size distribution into the SAR backscatter model: according to the particle size distribution function and the relation between the particle size and the surface roughness, the roughness parameters in the Oh model are adjusted, and the final Oh model is expressed as:
Wherein/> Is a surface roughness parameter calculated based on the particle size of the lognormal distribution.
Further, the optimization algorithm in S4 adopts a particle swarm optimization PSO algorithm, which specifically includes:
The RMSE is used as an optimized objective function for evaluating the difference between the model predicted value and the actual observed value, and the calculation formula of the RMSE is as follows:
Wherein/> Is the total number of data points,/>Is a model predictive value,/>Is the actual observed value;
Application of particle swarm optimization PSO algorithm: in the PSO algorithm, each "particle" represents one potential solution to the model parameters, and the position update of the particle is based on the individual historical optimal position and the historical optimal position of the whole population;
Parameter coding and initial population setting: encoding parameters of the model to form positions of particles, and if the model parameters are coefficients of soil humidity inversion, the position of each particle can be a vector of the coefficients to generate an initial particle swarm, and randomly determining the initial position and the speed of each particle;
PSO iterative updating process: in each iteration, the speed and position of the particles are updated according to the historical optimal positions of the individuals and the groups, and the speed update formula of the particles is as follows:
Wherein/> Is particle/>At time/>Speed of/(I)Is an inertial weight,/>And/>Is a learning factor,/>And/>Is a random number,/>Is particle/>History of best positions,/>Is the historical optimal position of the population, and the position of the particles updates the formula: /(I)；
And using the RMSE as an evaluation index, finding out the particle position which minimizes the RMSE, namely, the optimized model parameter, applying the optimized model parameter to a mathematical relationship model of soil-vegetation analysis, and updating the model to improve inversion accuracy.
Further, the inversion model in the step S5 is based on an empirical model, and the empirical model considers SAR backward scattering coefficient, polarization characteristic and incidence angle;
Adjusting an empirical model by using parameters optimized by a Particle Swarm Optimization (PSO) algorithm, wherein the optimized parameters comprise coefficients related to SAR signals and characteristic parameters of soil and vegetation;
input and processing of SAR data: inputting the preprocessed multi-time sequence and multi-polarization SAR data into an empirical model, and ensuring that the data format and the scale are matched with the requirements of the model;
Implementation of inversion process and calculation of soil water content: the processed SAR data is input into an empirical model, and the water content of the soil is estimated by using a calculation formula of the empirical model.
Further, the calculation formula of the empirical model is as follows:
Wherein/> Is soil moisture (water content)/>Is SAR backscatter coefficient,/>Is a polarization parameter,/>Is the incident angle,/>And/>Is a model coefficient determined by optimization.
Further, the step S6 specifically includes:
data preparation: collecting inversion to obtain soil water content data, associating the soil water content data with geographical position information, and combining the soil water content data with geographical coordinates or area codes;
Importing data by using GIS software, wherein each soil water content data point is assigned a specific geographic position;
in GIS, soil moisture content is demonstrated by a heat map, contour map, or color-coded scatter plot.
The invention has the beneficial effects that:
According to the invention, by combining multi-time sequence and multi-polarization SAR data and an advanced soil-vegetation model, the estimation precision of the soil water content is effectively improved, and the abundant information of the SAR data including the captured soil and vegetation characteristics is utilized, so that the water content estimation is more accurate, model parameters are optimized through a Particle Swarm Optimization (PSO) algorithm, and the estimation accuracy is further improved, especially when the influence of various factors such as soil type, surface roughness, vegetation coverage and the like is considered.
The method is not only suitable for soil humidity monitoring in the agricultural field, but also can be applied to multiple fields such as weather forecast, environment monitoring and water resource management, and has wide adaptability and flexibility because the method can process various different environmental conditions and soil types, and by combining a GIS technology, not only is the numerical soil water content estimation provided, but also the spatial distribution of the soil water content can be displayed, and the visual mode is particularly important for understanding regional water resource conditions and performing land management planning.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only of the invention and that other drawings can be obtained from them without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of an inversion method according to an embodiment of the invention.
Detailed Description
The present invention will be further described in detail with reference to specific embodiments in order to make the objects, technical solutions and advantages of the present invention more apparent.
It is to be noted that unless otherwise defined, technical or scientific terms used herein should be taken in a general sense as understood by one of ordinary skill in the art to which the present invention belongs. The terms "first," "second," and the like, as used herein, do not denote any order, quantity, or importance, but rather are used to distinguish one element from another. The word "comprising" or "comprises", and the like, means that elements or items preceding the word are included in the element or item listed after the word and equivalents thereof, but does not exclude other elements or items. The terms "connected" or "connected," and the like, are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "upper", "lower", "left", "right", etc. are used merely to indicate relative positional relationships, which may also be changed when the absolute position of the object to be described is changed.
As shown in fig. 1, an inversion method of soil water content based on sar comprises the following steps:
S1: collecting multi-time sequence and multi-polarization SAR data;
s2: preprocessing the collected SAR data, wherein the preprocessing comprises radiometric calibration, speckle noise suppression, terrain correction and registration of multi-time sequence data, and performing polarization decomposition on the multi-polarization data to extract characteristic information related to soil humidity;
S3: establishing a mathematical relationship model between SAR data and soil water content by applying soil-vegetation analysis;
s4: carrying out parameter optimization on the data relationship model by using an optimization algorithm, and improving inversion accuracy;
s5: introducing an inversion model, and converting the processed SAR data into soil water content information;
S6: and by combining with a GIS technology, the soil water content data is visualized to provide a spatial distribution map.
S1 specifically comprises:
Determining data acquisition parameters: selecting Sentinel-1 as a SAR satellite system for monitoring the water content of soil, and setting a frequency band of SAR imaging as a C band;
Setting a polarization mode to capture comprehensive information of soil surface roughness, wherein the polarization mode comprises a single polarization (HH or VV), a dual polarization (HH+HV or VV+VH) or a full polarization mode;
Planning time sequence data collection: planning data acquisition time points based on a climate mode and a vegetation growth period of a target area, and collecting data before and after a rainy season and in a seedling stage, a flowering stage, a setting stage and a maturing stage of vegetation growth;
arranging for data acquisition at desired points in time using satellite reservation functions to form a long-term, consistent time series;
coordinating ground synchronous measurements: ground synchronous measurements are made while the satellite is in transit, including in-field recordings of soil texture, vegetation type and coverage parameters.
The pretreatment in S2 specifically comprises:
Radiation calibration: performing radiometric calibration on the original SAR data to convert the original signal into a comparable reflectance value, and quantitatively analyzing the data by using calibration parameters provided by satellites to ensure that the data of different time and different sensors are comparable;
speckle noise suppression: due to the coherent nature of SAR data, images are often affected by speckle noise. Filtering the image by using a Lee filter or a frame filter to reduce noise and preserve details of the image;
terrain correction: for areas with large topography relief, a Digital Elevation Model (DEM) is used for topography correction so as to eliminate geometrical distortion caused by topography, and through correction, pixels in an image are ensured to accurately correspond to the positions of the ground, particularly in slopes and mountain areas;
Registration of multi-temporal data: the SAR data acquired at different times are accurately registered, so that the spatial consistency of multi-time sequence data is ensured, and a pixel-to-pixel registration method is adopted to reduce space-time dislocation;
Polarization decomposition: and carrying out polarization decomposition on the multi-polarization SAR data, extracting the characteristic of sensitivity to soil humidity by using an H/A/Alpha decomposition method, and analyzing different polarization channels and polarization decomposition parameters including phase information and a scattering mechanism so as to identify the change of the soil humidity.
Feature extraction and analysis: based on the result of polarization decomposition, characteristic parameters related to soil humidity are extracted, including scattering intensity, polarization ratio and coherence.
S3, analyzing soil-vegetation, considering the influence of soil humidity and vegetation coverage, integrating soil types and organic matter content, simulating the relation between SAR signals and the soil humidity by a mathematical relation model by adopting a physical foundation method, and simultaneously considering the influence of soil particle size distribution, surface roughness and vegetation parameters on the SAR signals;
Data-driven model optimization is also included: training and optimizing a data relationship model by using a machine learning algorithm, and combining the preprocessed SAR data and ground measured soil water content data, so as to train the model to identify a complex nonlinear relationship.
The mathematical relationship model framework is as follows:
Relationship between soil moisture and SAR signal: the relationship between soil moisture and SAR backscatter coefficients is described using a classical Oh model, which is used to estimate the effect of soil moisture on backscatter coefficients, expressed as:
Wherein/> Is the backscattering coefficient,/>Is the incident angle,/>Is the soil moisture (water content),Is a soil roughness parameter;
consider soil type and organic content: expanding an Oh model, adding the influence of soil type and organic matter content, and expressing the expanded model as follows:
Wherein/> Representing soil type (clay, sandy soil)/>Represents the organic content;
Effect of vegetation coverage: using vegetation parameters (vegetation moisture content, vegetation height, coverage) to simulate the effect of vegetation on SAR signals, a vegetation correction factor, such as WaterCloudModel, is introduced to account for scattering effects of the vegetation layer, expressed as:
wherein/> Representing vegetation parameters;
Consider soil particle size distribution and surface roughness: based on actual experimental data, an empirical relationship between soil particle size distribution and surface roughness and backscattering coefficient is established, and the lognormal distribution in the particle size distribution function is used to characterize the soil surface and is incorporated into the expanded model.
The realization steps are as follows:
Data collection and pretreatment:
SAR data, ground soil moisture data, vegetation information and meteorological data are collected.
The SAR data is preprocessed, including radiometric scaling and speckle noise suppression.
Parameter estimation and model correction:
The ground data is used to estimate model parameters (e.g., soil roughness coefficients).
The model is adjusted to improve accuracy by comparing the model output with the actual observations.
Model application and verification:
And (5) carrying out soil humidity inversion by using the model.
The model is verified using a separate dataset.
The use of a log-normal distribution in a particle size distribution function to characterize a soil surface specifically includes:
Characterization of soil particle size distribution:
selection of a lognormal distribution: the soil particle size is in lognormal distribution, namely the natural log of the soil particle size is in normal distribution, and the formula of lognormal distribution is as follows:
Wherein/> Is particle size/>Probability density of/>And/>The mean and standard deviation of the logarithmic particle size, respectively;
Soil particle size distribution was incorporated into the model: the particle size distribution influences the dielectric property and the surface roughness of soil, further influences SAR signals, and adjusts the soil roughness parameters To take into account the influence of the particle size distribution as a function/>Expression of/>, whereinSoil particle size defined by lognormal distribution;
relation of particle size distribution to SAR signal:
relationship between particle size and dielectric constant: the soil particle size affects the retention of the water content, and thus the dielectric constant, and an empirical formula is used to describe the relationship between particle size, water content and dielectric constant;
Relationship between particle size and surface roughness: the change of the particle size of the soil influences the surface roughness, the larger the particle size is, the coarser the surface is, and the average value and the variance of the surface roughness are calculated according to a particle size distribution function;
incorporating the particle size distribution into the SAR backscatter model: according to the particle size distribution function and the relation between the particle size and the surface roughness, the roughness parameters in the Oh model are adjusted, and the final Oh model is expressed as:
Wherein/> Is a surface roughness parameter calculated based on the particle size of the lognormal distribution.
In practical applications, specific soil particle size data needs to be collected for estimationAnd/>
The optimization algorithm in S4 adopts a particle swarm optimization PSO algorithm, and specifically comprises the following steps:
The RMSE is used as an optimized objective function for evaluating the difference between the model predicted value and the actual observed value, and the calculation formula of the RMSE is as follows:
Wherein/> Is the total number of data points,/>Is a model predictive value,/>Is the actual observed value;
Application of particle swarm optimization PSO algorithm: in the PSO algorithm, each "particle" represents one potential solution to the model parameters, and the position update of the particle is based on the individual historical optimal position and the historical optimal position of the whole population;
Parameter coding and initial population setting: encoding parameters of the model to form positions of particles, and if the model parameters are coefficients of soil humidity inversion, the position of each particle can be a vector of the coefficients to generate an initial particle swarm, and randomly determining the initial position and the speed of each particle;
PSO iterative updating process: in each iteration, the speed and position of the particles are updated according to the historical optimal positions of the individuals and the groups, and the speed update formula of the particles is as follows:
Wherein/> Is particle/>At time/>Speed of/(I)Is an inertial weight,/>And/>Is a learning factor,/>And/>Is a random number,/>Is particle/>History of best positions,/>Is the historical optimal position of the population, and the position of the particles updates the formula: /(I)；
And using the RMSE as an evaluation index, finding out the particle position which minimizes the RMSE, namely, the optimized model parameter, applying the optimized model parameter to a mathematical relationship model of soil-vegetation analysis, and updating the model to improve inversion accuracy. Optimization parameters include those parameters in the adjustment model that relate to soil type, vegetation coverage, soil particle size distribution, and surface roughness factors.
And verifying the optimized model by using an independent test data set, ensuring the accuracy and generalization capability of the model, and adjusting parameters (such as inertia weight and learning factor) of a PSO algorithm or performing more iterations according to a verification result to further optimize the model.
The method also comprises the step of calibrating and verifying the data relation model by combining ground actual measurement data, and specifically comprises the following steps:
collecting ground actual measurement data:
measured data is collected at a plurality of locations within the target area regarding soil moisture content, soil type, vegetation type, coverage, and the like.
Ensuring that these data cover different soil and vegetation conditions, as well as different climatic and seasonal variations.
Preprocessing ground actual measurement data:
the collected ground data is preprocessed, including data cleansing, outlier processing and format normalization, to facilitate integration with SAR data.
Model calibration:
the ground measured data is used to calibrate parameters of the "soil-vegetation model". This may include adjusting coefficients in the model to match the actual observations.
For example, if the model includes coefficients of soil moisture versus SAR signals, these coefficients may be adjusted by minimizing the difference between the model output and the ground observation data.
Model verification: the calibrated model is applied to a separate test dataset, which is not used for the calibration process. Error indicators, such as Root Mean Square Error (RMSE) or correlation coefficients, between the model predictions and the measured data are calculated to evaluate the accuracy and reliability of the model.
Result analysis and feedback adjustment: analyzing the results of the calibration and verification process identifies situations where the model performs poorly. Based on these analysis results, the model parameters are further adjusted, or the model structure is modified, to improve performance.
Continuous monitoring and iterative updating: and (3) carrying out model calibration and verification by using new ground actual measurement data at regular intervals, so as to ensure that the model is accurate and reliable along with time.
S5, the inversion model is based on an empirical model, and the empirical model considers SAR backward scattering coefficient, polarization characteristic and incidence angle;
Adjusting an empirical model by using parameters optimized by a Particle Swarm Optimization (PSO) algorithm, wherein the optimized parameters comprise coefficients related to SAR signals and characteristic parameters of soil and vegetation;
input and processing of SAR data: inputting the preprocessed multi-time sequence and multi-polarization SAR data into an empirical model, and ensuring that the data format and the scale are matched with the requirements of the model;
Implementation of inversion process and calculation of soil water content: the processed SAR data is input into an empirical model, and the water content of the soil is estimated by using a calculation formula of the empirical model.
The calculation formula of the empirical model is:
Wherein/> Is soil moisture (water content)/>Is SAR backscatter coefficient,/>Is a polarization parameter,/>Is the incident angle,/>And/>Model coefficients determined by optimization;
in empirical models, coefficients (e.g ) Representing the weights or influence of certain variables in the model, these coefficients determine the extent to which each variable in the model affects the predicted outcome (soil moisture content).
Optimization process-optimization refers to the use of particle swarm optimization PSOs to adjust these coefficients based on actual data so that the predicted outcome of the model is as close as possible to reality, in which the algorithm tries different combinations of coefficients, evaluates the effect of each combination, and selects coefficient values that optimize the model performance (minimize prediction errors).
For example, there is a simple empirical model to predict soil moisture content:
In this model,/> Is the predicted soil moisture contentIs a different feature (backscatter coefficient, polarization information, etc.) extracted from SAR data, coefficient/>It is determined by the data and optimization algorithm.
Data driven optimization such optimization methods are typically based on historical data sets containing known input features and corresponding soil moisture content measurements, and the optimization algorithm attempts to find a set of coefficients that minimize the difference between the model predicted moisture content and the actual measurements when applied to the historical data set.
The calculated soil moisture content may be a point estimate or a spatial distribution map, depending on the nature of the input data and the design of the model.
And verifying the water content of the soil obtained by inversion by using ground actual measurement data. And further adjusting parameters or algorithms of the empirical model according to the verification result to improve the accuracy of prediction.
S6 specifically comprises the following steps:
data preparation: collecting inversion to obtain soil water content data, associating the soil water content data with geographical position information, and combining the soil water content data with geographical coordinates or area codes;
Importing data by using GIS software, wherein each soil water content data point is assigned a specific geographic position;
In GIS, the soil moisture content is displayed through a heat map, a contour map or a color-coded scatter map;
Color schemes are designed to represent different levels of moisture content, for example, using a gradual change from light blue to dark blue to represent low to high moisture content, and other layers of geographic information, such as land use types, river networks, or terrain, are contemplated to provide more background information.
To enhance the usability of the visualization, interactive functionality may be incorporated into the GIS software, such as clicking on an area to display detailed moisture content data, allowing the user to display or hide specific geographic information or moisture content data as desired.
Results presentation and sharing: and a final soil water content distribution map is created, so that the soil water content distribution map is clear, accurate and rich in information. The generated map is output as a high resolution image or PDF file or shared on a GIS platform for use by decision makers and related stakeholders.
By combining GIS technology to perform data visualization, the invention not only provides quantitative estimation of soil water content, but also enhances understanding of spatial distribution characteristics, which is particularly important for regional water resource management and agricultural planning.
Through geographical visualization, the distribution condition of the soil moisture content in different areas can be visually displayed, and the identification of potential arid areas or over-wet areas is facilitated.
The introduction of interactive functionality enables end users to analyze and explore data more deeply, making more informative decisions.
Those of ordinary skill in the art will appreciate that: the discussion of any of the embodiments above is merely exemplary and is not intended to suggest that the scope of the invention is limited to these examples; the technical features of the above embodiments or in the different embodiments may also be combined within the idea of the invention, the steps may be implemented in any order and there are many other variations of the different aspects of the invention as described above, which are not provided in detail for the sake of brevity.
The present invention is intended to embrace all such alternatives, modifications and variances which fall within the broad scope of the appended claims. Therefore, any omission, modification, equivalent replacement, improvement, etc. of the present invention should be included in the scope of the present invention.
Claims (5)
1. The sar-based inversion method for the soil water content is characterized by comprising the following steps of:
S1: collecting multi-time sequence and multi-polarization SAR data;
s2: preprocessing the collected SAR data, wherein the preprocessing comprises radiometric calibration, speckle noise suppression, terrain correction and registration of multi-time sequence data, and performing polarization decomposition on the multi-polarization data to extract characteristic information related to soil humidity;
S3: establishing a mathematical relation model between SAR data and soil moisture content by applying soil-vegetation analysis, wherein the soil-vegetation analysis considers the influence of soil moisture and vegetation coverage, integrates soil types and organic matter content, and adopts a physical foundation method to simulate the relation between SAR signals and soil moisture, and simultaneously considers the influence of soil particle size distribution, surface roughness and vegetation parameters on the SAR signals;
Data-driven model optimization is also included: training and optimizing a data relationship model by using a machine learning algorithm, and combining the preprocessed SAR data and ground measured soil water content data, so as to train the model to identify a complex nonlinear relationship;
The mathematical relationship model framework is as follows:
Relationship between soil moisture and SAR signal: the relationship between soil moisture and SAR backscatter coefficients is described using a classical Oh model, which is used to estimate the effect of soil moisture on backscatter coefficients, expressed as:
σ 0 =f (θ, ω, s), where σ 0 is the backscattering coefficient, θ is the angle of incidence, ω is the soil humidity, s is the soil roughness parameter;
Consider soil type and organic content: expanding an Oh model, adding the influence of soil type and organic matter content, and expressing the expanded model as follows:
σ 0 =f (θ, ω, s, t, o), where t represents soil type and o represents organic content;
Effect of vegetation coverage: using vegetation parameters to simulate the effect of vegetation on SAR signals, introducing a vegetation correction factor to account for scattering effects of the vegetation layer, expressed as:
σ 0 =f (θ, ω, s, t, o, v) where v represents a vegetation parameter;
Consider soil particle size distribution and surface roughness: based on actual experimental data, establishing an empirical relationship between soil particle size distribution and surface roughness and a backscattering coefficient, characterizing the characteristics of the soil surface by using lognormal distribution in a particle size distribution function, and incorporating the lognormal distribution into an expanded model;
s4: parameter optimization is carried out on the data relation model by using an optimization algorithm, inversion precision is improved, the optimization algorithm adopts a particle swarm optimization PSO algorithm, and the method specifically comprises the following steps:
The RMSE is used as an optimized objective function for evaluating the difference between the model predicted value and the actual observed value, and the calculation formula of the RMSE is as follows:
where n is the total number of data points,/> Is a model predictive value, y i is an actual observed value;
Application of particle swarm optimization PSO algorithm: in the PSO algorithm, each "particle" represents one potential solution to the model parameters, and the position update of the particle is based on the individual historical optimal position and the historical optimal position of the whole population;
parameter coding and initial population setting: encoding parameters of the model to form positions of particles, generating an initial particle group, and randomly determining the initial position and speed of each particle;
PSO iterative updating process: in each iteration, the speed and position of the particles are updated according to the historical optimal positions of the individuals and the groups, and the speed update formula of the particles is as follows:
wherein/> Is the velocity of particle i at time t, w is the inertial weight, c 1 and c 2 are the learning factors, r 1 and r 2 are random numbers, p i is the historical optimal position of particle i, g is the historical optimal position of the population, and the position of the particle updates the formula: /(I)
Using the RMSE as an evaluation index, finding out the particle position which minimizes the RMSE, namely, the optimized model parameter, applying the optimized model parameter to a mathematical relationship model of soil-vegetation analysis, and updating the model to improve inversion accuracy;
S5: introducing an inversion model, and converting the processed SAR data into soil water content information, wherein the inversion model is based on an empirical model, and the empirical model considers SAR backscattering coefficient, polarization characteristic and incidence angle;
Adjusting an empirical model by using parameters optimized by a Particle Swarm Optimization (PSO) algorithm, wherein the optimized parameters comprise coefficients related to SAR signals and characteristic parameters of soil and vegetation;
input and processing of SAR data: inputting the preprocessed multi-time sequence and multi-polarization SAR data into an empirical model, and ensuring that the data format and the scale are matched with the requirements of the model;
Implementation of inversion process and calculation of soil water content: inputting the processed SAR data in an empirical model, and estimating the water content of the soil by using a calculation formula of the empirical model;
The calculation formula of the experience model is as follows:
ω=a·σ 0 +b·p+c·θ+d, where ω is soil humidity, σ 0 is SAR backscatter coefficient, P is polarization parameter, θ is incident angle, a, b, c and d are model coefficients determined by optimization;
S6: and by combining with a GIS technology, the soil water content data is visualized to provide a spatial distribution map.
2. The sar-based soil moisture inversion method of claim 1, wherein S1 specifically comprises:
Determining data acquisition parameters: selecting Sentinel-1 as a SAR satellite system for monitoring the water content of soil, and setting a frequency band of SAR imaging as a C band;
Setting a polarization mode to capture comprehensive information of soil surface roughness, wherein the polarization mode comprises a single polarization mode, a dual polarization mode or a full polarization mode;
Planning time sequence data collection: planning data acquisition time points based on a climate mode and a vegetation growth period of a target area, and collecting data before and after a rainy season and in a seedling stage, a flowering stage, a setting stage and a maturing stage of vegetation growth;
arranging for data acquisition at desired points in time using satellite reservation functions to form a long-term, consistent time series;
coordinating ground synchronous measurements: ground synchronous measurements are made while the satellite is in transit, including in-field recordings of soil texture, vegetation type and coverage parameters.
3. The sar-based soil moisture inversion method of claim 2, wherein the pretreatment in S2 comprises:
Radiation calibration: performing radiometric calibration on the original SAR data to convert the original signal into a comparable reflectance value, and quantitatively analyzing the data by using calibration parameters provided by satellites to ensure that the data of different time and different sensors are comparable;
Speckle noise suppression: filtering the image by using a Lee filter or a frame filter to reduce noise and preserve details of the image;
terrain correction: for the region with larger topographic relief, using a digital elevation model to carry out topographic correction so as to eliminate geometric distortion caused by the topography, and ensuring that pixels in the image accurately correspond to the position of the ground through correction;
Registration of multi-temporal data: the SAR data acquired at different times are accurately registered, so that the spatial consistency of multi-time sequence data is ensured, and a pixel-to-pixel registration method is adopted to reduce space-time dislocation;
Polarization decomposition: carrying out polarization decomposition on multi-polarization SAR data, extracting soil humidity sensitive characteristics by using an H/A/Alpha decomposition method, and analyzing different polarization channels and polarization decomposition parameters including phase information and a scattering mechanism so as to identify the change of soil humidity;
Feature extraction and analysis: based on the result of polarization decomposition, characteristic parameters related to soil humidity are extracted, including scattering intensity, polarization ratio and coherence.
4. A method of inverting soil moisture content based on sar according to claim 3, wherein said characterizing the soil surface using the lognormal distribution in the particle size distribution function comprises:
Characterization of soil particle size distribution:
selection of a lognormal distribution: the soil particle size is in lognormal distribution, namely the natural log of the soil particle size is in normal distribution, and the formula of lognormal distribution is as follows:
Wherein P (d) is the probability density of particle size d, and μ and σ are the mean and standard deviation of logarithmic particle size, respectively;
Soil particle size distribution was incorporated into the model: taking into account the influence of the particle size distribution by adjusting the soil roughness parameter s, expressed as a function s (d), where d is the soil particle size defined by the lognormal distribution;
relation of particle size distribution to SAR signal:
Relationship between particle size and dielectric constant: the soil particle size affects the retention of the water content, and thus the dielectric constant, and an empirical formula is used to describe the relationship between particle size, water content and dielectric constant;
Relationship between particle size and surface roughness: the change of the particle size of the soil influences the surface roughness, the larger the particle size is, the coarser the surface is, and the average value and the variance of the surface roughness are calculated according to a particle size distribution function;
incorporating the particle size distribution into the SAR backscatter model: according to the particle size distribution function and the relation between the particle size and the surface roughness, the roughness parameters in the Oh model are adjusted, and the final Oh model is expressed as:
σ 0 =f (θ, ω, s, s (d), t, o, v), where s (d) is a surface roughness parameter calculated based on the particle size of the lognormal distribution.
5. The sar-based soil moisture inversion method of claim 1, wherein S6 comprises:
data preparation: collecting inversion to obtain soil water content data, associating the soil water content data with geographical position information, and combining the soil water content data with geographical coordinates or area codes;
Importing data by using GIS software, wherein each soil water content data point is assigned a specific geographic position;
in GIS, soil moisture content is demonstrated by a heat map, contour map, or color-coded scatter plot.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410245311.8A CN117826112B (en) | 2024-03-05 | Soil water content inversion method based on sar |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410245311.8A CN117826112B (en) | 2024-03-05 | Soil water content inversion method based on sar |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117826112A CN117826112A (en) | 2024-04-05 |
| CN117826112B true CN117826112B (en) | 2024-05-31 |
Family
ID=
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106990121A (en) * | 2017-03-30 | 2017-07-28 | 中国科学院遥感与数字地球研究所 | A kind of full-polarization SAR data soil moisture content inversion method |
| CN108681652A (en) * | 2018-06-04 | 2018-10-19 | 中国科学院遥感与数字地球研究所 | A kind of vegetation-covered area Soil Moisture Inversion method based on three number of high score |
| CN109977574A (en) * | 2019-04-02 | 2019-07-05 | 中国科学院遥感与数字地球研究所 | A kind of Soil Moisture Inversion method based on improvement Freeman-Durden polarization decomposing model |
| CN111307643A (en) * | 2019-04-04 | 2020-06-19 | 西北大学 | Soil moisture prediction method based on machine learning algorithm |
| CN112906310A (en) * | 2021-04-19 | 2021-06-04 | 中国农业科学院农业资源与农业区划研究所 | Method for optimizing BP neural network microwave remote sensing soil moisture inversion by considering firefly algorithm |
| CN114065643A (en) * | 2021-11-24 | 2022-02-18 | 电子科技大学长三角研究院(湖州) | Plant soil water content estimation method and system based on SAR and polarization decomposition |
| CN114444290A (en) * | 2022-01-20 | 2022-05-06 | 天津智云水务科技有限公司 | Method and system for automatically generating pressure and flow monitoring threshold of water supply system |
| CN114782834A (en) * | 2022-04-19 | 2022-07-22 | 电子科技大学长三角研究院(湖州) | Farmland soil moisture retrieval method and system based on ratio method |
| CN115479960A (en) * | 2022-09-13 | 2022-12-16 | 安徽师范大学 | Crop growth process soil humidity monitoring method combining SAR and optical remote sensing data |
| CN115931917A (en) * | 2022-11-17 | 2023-04-07 | 电子科技大学长三角研究院(湖州) | Soil moisture inversion method and system based on fully-polarized SAR data |
| CN116047500A (en) * | 2022-12-02 | 2023-05-02 | 自然资源部国土卫星遥感应用中心 | Vegetation coverage area soil moisture inversion method considering polarization scattering information |
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106990121A (en) * | 2017-03-30 | 2017-07-28 | 中国科学院遥感与数字地球研究所 | A kind of full-polarization SAR data soil moisture content inversion method |
| CN108681652A (en) * | 2018-06-04 | 2018-10-19 | 中国科学院遥感与数字地球研究所 | A kind of vegetation-covered area Soil Moisture Inversion method based on three number of high score |
| CN109977574A (en) * | 2019-04-02 | 2019-07-05 | 中国科学院遥感与数字地球研究所 | A kind of Soil Moisture Inversion method based on improvement Freeman-Durden polarization decomposing model |
| CN111307643A (en) * | 2019-04-04 | 2020-06-19 | 西北大学 | Soil moisture prediction method based on machine learning algorithm |
| CN112906310A (en) * | 2021-04-19 | 2021-06-04 | 中国农业科学院农业资源与农业区划研究所 | Method for optimizing BP neural network microwave remote sensing soil moisture inversion by considering firefly algorithm |
| CN114065643A (en) * | 2021-11-24 | 2022-02-18 | 电子科技大学长三角研究院(湖州) | Plant soil water content estimation method and system based on SAR and polarization decomposition |
| CN114444290A (en) * | 2022-01-20 | 2022-05-06 | 天津智云水务科技有限公司 | Method and system for automatically generating pressure and flow monitoring threshold of water supply system |
| CN114782834A (en) * | 2022-04-19 | 2022-07-22 | 电子科技大学长三角研究院(湖州) | Farmland soil moisture retrieval method and system based on ratio method |
| CN115479960A (en) * | 2022-09-13 | 2022-12-16 | 安徽师范大学 | Crop growth process soil humidity monitoring method combining SAR and optical remote sensing data |
| CN115931917A (en) * | 2022-11-17 | 2023-04-07 | 电子科技大学长三角研究院(湖州) | Soil moisture inversion method and system based on fully-polarized SAR data |
| CN116047500A (en) * | 2022-12-02 | 2023-05-02 | 自然资源部国土卫星遥感应用中心 | Vegetation coverage area soil moisture inversion method considering polarization scattering information |
Non-Patent Citations (3)
| Title |
|---|
| 《Design of Input Matching Network for 150W PT Power Converter》;Rui Zhang et.al;《2008 Australasian Universities Power Engineering Conference》;20090410;全文 * |
| 裸露土壤参数空间分布对被动微波遥感反演土壤水分的影响;张涛;张立新;蒋玲梅;赵少杰;赵天杰;李云青;;中国科学:地球科学;20120920(09);全文 * |
| 近景摄影测量提高裸露地表粗糙度测量精度;陶浩然;陈权;李震;曾江源;陈晓宁;;农业工程学报;20170808(15);全文 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| El Hajj et al. | Soil moisture retrieval over irrigated grassland using X-band SAR data | |
| CN107909607B (en) | A kind of year regional vegetation coverage computational methods | |
| Choi et al. | A microwave-optical/infrared disaggregation for improving spatial representation of soil moisture using AMSR-E and MODIS products | |
| Jin et al. | Downscaling AMSR-2 soil moisture data with geographically weighted area-to-area regression kriging | |
| CN106501186B (en) | A kind of soil moisture content product NO emissions reduction method | |
| CN112699959B (en) | Multi-source multi-scale precipitation data fusion method and device based on energy functional model | |
| Bocchiola et al. | Spatial estimation of Snow Water Equivalent at different dates within the Adamello Park of Italy | |
| CN110927120B (en) | Early warning method for vegetation coverage | |
| Hajdu et al. | Modelling of near-surface soil moisture using machine learning and multi-temporal sentinel 1 images in New Zealand | |
| CN114819737B (en) | Method, system and storage medium for estimating carbon reserves of highway road vegetation | |
| Liu et al. | Development of a 250-m downscaled land surface temperature data set and its application to improving remotely sensed evapotranspiration over large landscapes in northern china | |
| Jin et al. | Deriving temporally continuous soil moisture estimations at fine resolution by downscaling remotely sensed product | |
| Zhao et al. | Triangle Space-Based Surface Soil Moisture Estimation by the Synergistic Use of $ In\Situ $ Measurements and Optical/Thermal Infrared Remote Sensing: An Alternative to Conventional Validations | |
| Lou et al. | An effective method for canopy chlorophyll content estimation of marsh vegetation based on multiscale remote sensing data | |
| CN105528523A (en) | Soil thickness inversion method based on remote sensing data | |
| Das et al. | Ensemble surface soil moisture estimates at farm-scale combining satellite-based optical-thermal-microwave remote sensing observations | |
| CN114062439A (en) | Method for jointly estimating salinity of soil profile by using time series remote sensing images | |
| Hong et al. | 15 Global Precipitation Estimation and Applications | |
| Jiang et al. | Desertification in the south Junggar Basin, 2000–2009: Part I. Spatial analysis and indicator retrieval | |
| Liu et al. | Bi-LSTM model for time series leaf area index estimation using multiple satellite products | |
| CN117075138A (en) | Remote sensing measurement and calculation method, system and medium for canopy height of 30-meter forest in area | |
| De Lannoy et al. | Adaptive soil moisture profile filtering for horizontal information propagation in the independent column-based CLM2. 0 | |
| CN117826112B (en) | Soil water content inversion method based on sar | |
| Zhan et al. | A data assimilation method for simultaneously estimating the multiscale leaf area index from time-series multi-resolution satellite observations | |
| CN114611699A (en) | Soil moisture downscaling method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |