Organic carbon extraction method based on microwave soil moisture sensor

Abstract

The present invention relates to an organic carbon extraction method based on a microwave soil moisture sensor. According to the present invention, the organic carbon extraction method based on a microwave soil moisture sensor comprises the steps of: acquiring soil initial data including soil temperature, physical properties of the soil, and soil organic matter content; generating a permittivity observation proxy from soil moisture sensors of a plurality of wavelengths; generating a soil moisture estimation initial value using a conventional multiphase permittivity model through repeated estimation; and extracting organic carbon in the soil by performing optimization of a permittivity maximum value improvement model using the soil moisture estimation initial value and a soil organic matter value sampled from a soil organic matter map. According to the present invention, it is possible to estimate organic carbon in soil without additional cost by utilizing a soil moisture sensor of an existing smart farm system.

Description

Organic carbon extraction method based on microwave soil moisture sensor

[0001] The present invention relates to an organic carbon extraction method based on a microwave soil moisture sensor, and more particularly, to an organic carbon extraction method based on a microwave soil moisture sensor that estimates the amount of organic matter by analyzing the combined water ratio of moisture in soil using dielectric constant information provided by the microwave soil moisture sensor.

[0002] Due to global warming, abnormal weather and pest damage are becoming more frequent, and the number of cases of crop failure in traditional agriculture is gradually increasing.

[0003] In addition, as the labor shortage problem is becoming more serious due to the aging and population decline in rural areas, investment and research and development in smart farms, which are relatively free from such problems, are becoming active.

[0004] In these smart farm systems, microwave-based soil moisture information is being used as important data.

[0005] However, the system for measuring soil organic carbon information is still under-researched.

[0006] If soil organic carbon information could be easily obtained from smart farm systems, it is expected that innovative improvements could be made in food production, water management, and carbon management. Therefore, research is needed to conveniently obtain soil organic carbon information.

[0007] However, conventional soil carbon measurement methods mainly rely on laboratory-based methods such as high-temperature combustion or wet oxidation, and are therefore not suitable for smart farm systems that require real-time information provision.

[0008] In addition, conventionally used visible light and near-infrared based soil organic carbon sensors have limitations such as high cost and difficulty in maintenance.

[0009] According to the present invention, the purpose is to provide an organic carbon extraction method based on a microwave soil moisture sensor that estimates the amount of organic matter by analyzing the combined water content ratio of moisture in soil using dielectric constant information provided by the microwave soil moisture sensor.

[0010] According to one embodiment of the present invention for achieving such technical task, a method for extracting organic carbon based on a microwave soil moisture sensor includes the steps of: acquiring initial soil data including soil temperature, soil physical properties, and soil organic matter content; generating a permittivity observation proxy from soil moisture sensors of multiple wavelengths; generating an initial value of soil moisture estimation using a conventional multi-phase permittivity model through iterative estimation; and extracting organic carbon from soil by optimizing a maximum permittivity improvement model using the initial value of soil moisture estimation and soil organic matter values ​​sampled from a soil organic matter map.

[0011] In this way, according to the present invention, soil organic carbon can be estimated without additional cost by utilizing the soil moisture sensor of an existing smart farm system.

[0012] Accordingly, the accuracy of real-time soil management information can be improved and an important foundation for sustainable agricultural practices can be established.

[0013] FIG. 1 is a diagram illustrating a flow of an organic carbon extraction method based on a microwave soil moisture sensor according to one embodiment of the present invention.

[0014] Figure 2 is a diagram illustrating an example of soil organic carbon estimation.

[0015] FIG. 3 is a diagram illustrating a flow of steps for extracting soil organic carbon according to one embodiment of the present invention.

[0016] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings. In this process, the thickness of lines and the sizes of components depicted in the drawings may be exaggerated for clarity and convenience of explanation.

[0017] Furthermore, the terms described below are defined based on their functions within the present invention, and may vary depending on the intent or custom of the user or operator. Therefore, the definitions of these terms should be based on the overall content of this specification.

[0018] Hereinafter, an organic carbon extraction method based on a microwave soil moisture sensor according to one embodiment of the present invention will be specifically described with reference to FIGS. 1 to 3.

[0019] FIG. 1 is a diagram illustrating a flow of an organic carbon extraction method based on a microwave soil moisture sensor according to one embodiment of the present invention, FIG. 2 is a diagram illustrating an example of soil organic carbon estimation, and FIG. 3 is a diagram illustrating a flow of steps for extracting soil organic carbon according to one embodiment of the present invention.

[0020] As illustrated in FIG. 1, an organic carbon extraction method based on a microwave soil moisture sensor according to one embodiment of the present invention first acquires initial soil data including soil temperature, soil physical properties, and soil organic matter content (S110).

[0021] More specifically, first, the soil temperature value is obtained using a temperature sensor installed at the monitoring point.

[0022] The thermal status of soil affects the evaporation and transpiration processes of soil moisture, and therefore soil temperature data can be used as input values ​​for soil moisture models.

[0023] Next, refer to the soil type map to determine the soil composition of your area.

[0024] The soil type map provides initial values ​​for the ratios of clay, silt, and sand, which are important factors in determining the physical properties of the soil and its ability to hold and move water.

[0025] These soil physical properties provide essential information for determining the ratio of bound and free water in the soil.

[0026] Next, the initial value of soil organic matter is extracted through the organic matter map.

[0027] Soil organic matter increases bound water, which significantly reduces the dielectric constant of bound water compared to free water. Therefore, soil organic matter indirectly influences the observed dielectric constant by inducing phase changes in soil moisture.

[0028] Finally, the dielectric constant of the soil is observed using multiple microwave soil moisture sensors.

[0029] Permittivity is a physical property that indicates how well soil absorbs and reflects electromagnetic waves, especially microwaves, and is closely related to soil moisture content.

[0030] After acquiring initial soil data, a dielectric constant observation agent is created from a soil moisture sensor of multiple wavelengths (S120).

[0031] Before explaining the step of estimating soil moisture using the multiphase dielectric constant maximum improvement model of the present invention, a method of estimating soil moisture in a dielectric constant surrogate using a conventional multiphase dielectric constant model is explained.

[0032] The permittivity refers to the polarizability of a material at a specific wavelength, and the dipole structure of water molecules is very sensitive to microwave electric fields due to its very high permittivity (about 80).

[0033] On the other hand, the permittivity of mineral soils is low, with values ​​ranging from 3 to 5, and has little response to microwave electric fields.

[0034] At this time, the effective permittivity of completely dry organic soil can be calculated using the following mathematical equation 1, which is a conventional multi-phase permittivity model.

[0035]

[0036] At this time, is the effective permittivity of dry organic soil, is the clay permittivity, is the volume ratio of clay, is the permittivity of sand, is the volume ratio of sand.

[0037] also, is the silt permittivity, is the volume fraction of silt, is the organic dielectric constant, is the volume ratio of organic matter.

[0038] At this time, the volume ratio of organic matter can be calculated using the following mathematical formula 2.

[0039]

[0040] Here, SOM is the weight ratio of organic matter, is the volume density of organic matter, is the bulk density of mineral soil.

[0041] Also, the organic matter weight ratio can be converted into organic carbon weight ratio OC through the following mathematical formula 3.

[0042]

[0043] At this time, is the conversion coefficient.

[0044] Also, a falsification point for estimating soil moisture (SM) is calculated using the following mathematical formula 4, and the saturation point p is calculated using the following mathematical formula 5.

[0045]

[0046]

[0047] Soil permittivity observations are essential to finding the closest optimal soil moisture and organic matter combination in simulations using permittivity models.

[0048] However, most soil moisture sensors focus only on measuring soil moisture and do not separately record the actual observed dielectric constant value.

[0049] Even if soil permittivity values ​​can be obtained from soil moisture sensors, the relationship between permittivity, soil moisture, and organic matter varies significantly at different wavelengths, requiring the development of individual permittivity models for each wavelength to estimate organic carbon from permittivity.

[0050] In such a situation, if there is no record of soil permittivity observation or if there is no permittivity model that physically considers organic matter in the relevant wavelength range, a permittivity surrogate can be created using a permittivity surrogate technique as in mathematical equation 6 below.

[0051]

[0052] At this time, refers to the soil moisture value observed from a soil moisture sensor in a wavelength range of a preset range (e.g., 50 MHz to 18 GHz), is a dielectric constant surrogate of a preset wavelength (e.g., 50 MHz).

[0053] At this time, the soil moisture value observed by the soil moisture sensor is used as a parameter of the regression model between the soil moisture and dielectric constant observation values ​​so that the error becomes 0 using a second-order polynomial-based regression model (e.g., Seyfried model). , and Decide.

[0054] The parameters of the regression model determined in this way can be used to convert soil moisture values ​​observed across multiple wavelengths into a permittivity proxy observable at a single wavelength (e.g., 50 MHz). This allows for application to both the existing 50 MHz-based multiphase permittivity model and the improved multiphase permittivity maximum model.

[0055] After creating a dielectric constant observation agent, an initial value for soil moisture estimation is created using a conventional multi-phase dielectric constant model through iterative estimation (S130).

[0056] When the soil moisture content is less than the withering point, the bound water acts dominantly due to the charge of the soil particles, attracting water molecules and attaching them to the surface of the soil particles, forming a thin moisture layer around the mineral particles and organic tissues.

[0057] As more water is added to the soil, this moisture layer becomes thicker, causing van der Waals forces to dominate over the charges on the particle surfaces.

[0058] When only these combinations exist, the soil moisture when the soil moisture is less than the withering point is calculated using the following mathematical formula 7. ) is estimated.

[0059]

[0060] In addition, when the soil moisture is greater than the withering point and less than the saturation point, the permittivity of the water in the soil is expressed as a complex of the permittivity of bound water and free water, and the soil moisture when the soil moisture is between the withering point and the saturation point is expressed by the following mathematical expression 8. ) is estimated.

[0061]

[0062] In addition, when the soil moisture is above the saturation point, the bound water no longer exists and all moisture becomes free water, and through the following mathematical equation 9, the soil moisture when the soil moisture is above the saturation point ( ) is estimated.

[0063]

[0064] Since soil moisture is the target of estimation, a model according to the soil moisture range cannot be applied, but soil moisture can be estimated using all three stages of models based on the observed permittivity, and then soil moisture can be estimated in a discontinuous model by selecting the model with the second largest soil moisture value.

[0065] At this time, the initial estimation value can be set through an iterative estimation method in soil moisture that shows the most similar simulation results to the dielectric constant proxy for estimating soil organic carbon.

[0066] In the iterative estimation process, the organic matter values ​​used are those of soil organic matter sampled from a global soil organic matter map with low resolution.

[0067] Soil moisture (SM) is measured in a preset range (e.g., 0 to 1 [cm 3 / cm 3 ] at preset intervals (e.g., 0.05 [cm) 3 / cm 3 ] is set to the interval.

[0068] The dielectric constant model applied in this case is a technique that does not consider organic matter in the maximum dielectric constant of bound and free water.

[0069] The iterative estimation heritability model is divided into three sections according to the soil moisture range, the first section is from soil moisture 0 to wilting point (WP), the second section is from wilting point to saturation point (P), and the third section is from saturation point to 1 [cm 3 / cm 3 ] is set to .

[0070] Starting from 0, preset 0.05 [cm 3 / cm 3 ] and increase the dielectric constant ( ) for each of the three models as in the following mathematical expression 10 from one soil moisture input. ) and rank them according to their size, and take the median as the simulated heritability ( ) is selected.

[0071]

[0072] Finally, from 0 to 1 [cm 3 / cm 3 ] Among the soil moisture points of a preset number (e.g., 20) set within the range, the soil moisture value closest to the observed dielectric constant surrogate is set as the initial value of the optimization technique as in the following mathematical expression 11.

[0073]

[0074] These steps i is 0(0cm 3 / cm 3 ) to 100(1cm) 3 / cm 3 ) is repeated until the vertices of the soil moisture and soil organic matter pairs of the group converge to the optimal point where the dielectric constant simulation is closest to the dielectric constant observation.

[0075] This process is repeated until there are no more improvements, or until other user-defined termination conditions are met.

[0076] At this time, the conventional multi-phase dielectric constant model determines the maximum dielectric constant of bound water and free water by considering only the soil clay content as in the following mathematical equation 12.

[0077]

[0078] Afterwards, the initial soil moisture estimation value is generated as in the following mathematical expression 13.

[0079]

[0080] At this time, is the maximum permittivity of the bonding number, is the maximum permittivity of free water, OM is the organic matter content, It is an empirical constant.

[0081] However, the actual permittivity of soil is also greatly affected by its organic matter content.

[0082] Therefore, conventional multiphase permittivity models do not reflect organic matter in the setting of maximum permittivity, which may result in poor soil moisture estimation accuracy in soils with very high or low organic matter.

[0083] Therefore, in the present invention, dielectric constant information is obtained through a model for improving the maximum value of the bound water and free water dielectric constants considering the soil clay content and organic matter content, as in the following mathematical equation 14.

[0084]

[0085] At this time, is an empirical constant.

[0086] The empirical constant is determined based on the observational data on which it is based, and in the case of SMAPVEX12 observational data, and It is decided by .

[0087] Next, the values ​​calculated through mathematical expressions 6 and 10 are By applying to mathematical equations 7 to 9, and produce, The second value is determined as the final soil moisture.

[0088] The soil organic carbon is extracted by performing optimization of the maximum dielectric constant improvement model using the initial value of the soil moisture estimated in this way and the soil organic matter value sampled from the soil organic matter map (S140).

[0089] By estimating the optimal soil moisture and the ratio of bound water to this moisture while minimizing the difference between the observed and simulated values ​​of permittivity through model optimization, the soil organic matter value can be derived, thereby determining the optimal values ​​of the two unknowns, i.e., soil moisture and soil organic matter.

[0090] Soil moisture contains uncertainty due to organic matter, so setting the initial value of soil classification is an important factor for optimizing the model.

[0091] This is because when performing an optimization estimation, the three vertices representing organic matter information are uncertainly determined around the values ​​of the organic matter map on the x-axis, and if the soil moisture vertex on the y-axis is determined near the exact observed value, the error of the dielectric constant simulation may quickly converge below the desired level without sufficient exploration of the organic matter, which may lead to a problem of not being able to find the exact organic matter value.

[0092] When considering interactions between multiple variables in complex systems or models, it is important to set initial values ​​that include appropriate error bounds for all variables.

[0093] As shown in Fig. 2, it is not only practically impossible to use the soil moisture value measured by actual weight as an initial value, but also the high accuracy of the soil moisture value may hinder the search for organic matter, resulting in a decrease in the accuracy of the optimal estimate of soil organic carbon.

[0094] At this time, (a) of FIG. 2 is a drawing showing an example of estimating organic carbon using the optimization technique of the present invention, (b) of FIG. 2 is a drawing showing an example of estimating organic carbon using the soil moisture value measured by the actual weight formula as the initial value, and (c) of FIG. 2 is a drawing showing an example of estimating soil organic matter from the initial soil moisture value estimated from the dielectric constant.

[0095] Therefore, the steps of extracting soil organic carbon of the present invention to solve these problems are specifically described with reference to FIG. 3.

[0096] First, the maximum dielectric constant improvement model, which is subdivided into three continuous functions according to the cases occurring below the falsification point, between the falsification point and the saturation point, and above the saturation point, is optimized in the state of the dielectric constant proxy and the minimum simulation error (S141).

[0097] At this time, the estimated initial value including the error applies a more accurate multiphase dielectric constant maximum value improvement model (Mathematical Formula 12) in the step of estimating organic matter, unlike the previous multiphase dielectric constant model (Mathematical Formula 11).

[0098] Additionally, optimization can be performed using the Nelder-Mead technique, which forms a group using test points and repeatedly transforms this group to find the optimal point.

[0099] In the Nelder-Mead technique, if the unknown variable is a two-dimensional problem consisting of soil moisture (SM) and soil organic matter (OM), three starting vertices (test points) are randomly selected.

[0100] This involves determining three points around the estimated initial value of soil moisture and the initial value of organic matter sampled from the organic matter map.

[0101] At this time, is calculated as the observed permittivity - simulated permittivity starting point value as in the following mathematical expression 15.

[0102]

[0103] At this time, is the value estimated using the iterative estimation technique, The initial error starting point of the brightness temperature is determined by the values ​​sampled from the organic map.

[0104] Also, reflection points to reduce errors is calculated as shown in the following mathematical formula 16.

[0105]

[0106] At this time, In the beginning This becomes the point where it started from the previous step, is the point with the highest function value, Is is the center point of all other points except , and α is the reflection coefficient.

[0107] Next, a new point is created by shifting the vertex of the soil moisture and organic matter pair that simulates the permittivity with the lowest accuracy compared to the observed permittivity.

[0108] Through this, the vertices with the highest error can be improved.

[0109] Next, if a more improved function value is produced at the reflection point, a new point is created by moving further in the same direction as in Equation 17 below.

[0110]

[0111] At this time, is the expansion coefficient.

[0112] On the other hand, if the reflected point is not improved, the point with the highest function value is moved in the direction estimated to be optimal, as in Equation 18 below.

[0113]

[0114] At this time, 0<β<1 is the contraction coefficient.

[0115] Additionally, if the reflection does not produce a better point, a contraction is performed to a new point closer to the center point.

[0116] Finally, if all points are not improved, the size of the group is reduced by moving all remaining points to the best point except the best point.

[0117] This is a last resort to resolve deadlock in the algorithm. If there is no improvement even after shrinking, all points except the optimal point Pl (the point with the lowest function value) are reduced in the direction of Pl, as in Equation 19 below.

[0118]

[0119] Here, δ is the reduction coefficient for all i≠1.

[0120] Through this optimization process, each dielectric constant model is optimized as shown in the following mathematical expression 20.

[0121]

[0122] At this time, means the nth model.

[0123] After performing optimization of each dielectric constant model, the soil moisture converged in each model is ranked as in the following mathematical equation 21, and the soil moisture of the middle rank is selected (S142).

[0124]

[0125] Next, soil organic matter is calculated according to the selected soil moisture content as in the following mathematical equation 22 (S143).

[0126]

[0127] Finally, soil organic carbon is calculated from the soil organic matter produced using the following mathematical formula 23 (S144).

[0128]

[0129] At this time, since both the soil organic matter map and the observations were calculated using the soil organic carbon map and the observations using a conversion factor (e.g., 1.72), if an estimated soil organic carbon value is needed, it can be estimated using the change coefficient.

[0130] Based on the soil organic carbon information generated through the above process, smart farm systems can be controlled in real time, adjusting irrigation, fertilization, and growth environment control elements to maintain optimal moisture and carbon conditions for crop growth. This allows for efficient management of water usage and enhanced soil carbon storage capacity, contributing to sustainable agriculture and increased food productivity.

[0131] In this way, according to the present invention, soil organic carbon can be estimated without additional cost by utilizing the soil moisture sensor of an existing smart farm system.

[0132] Accordingly, the accuracy of real-time soil management information can be improved and an important foundation for sustainable agricultural practices can be established.

[0133] While the present invention has been described with reference to the embodiments illustrated in the drawings, these are merely exemplary, and those skilled in the art will understand that various modifications and equivalent embodiments are possible. Therefore, the true technical protection scope of the present invention should be determined by the technical spirit of the following claims.

Claims

1. A step of acquiring initial soil data including soil temperature, soil physical properties, and soil organic matter content; A step of generating a dielectric constant observation agent from a soil moisture sensor of multiple wavelengths; A step of generating an initial value for soil moisture estimation using a conventional multiphase dielectric constant model through iterative estimation; and A method for extracting organic carbon from soil based on a microwave soil moisture sensor, comprising a step of extracting organic carbon from soil by performing optimization of a maximum dielectric constant improvement model using the initial soil moisture estimation value and soil organic matter values ​​sampled from a soil organic matter map.

2. In paragraph 1, The step of acquiring the above soil initial data is: A method for extracting organic carbon based on a microwave soil moisture sensor, which obtains a soil temperature value using a temperature sensor installed at a monitoring point, obtains an initial value for the ratio of clay, silt, and sand by referring to a soil type map, obtains an initial value of soil organic matter through an organic matter map, and obtains a soil permittivity using a microwave soil moisture sensor of multiple wavelengths.

3. In paragraph 1, The step of generating the above dielectric constant observation agent is: A method for extracting organic carbon based on a microwave soil moisture sensor that generates a dielectric constant observation surrogate using the following mathematical formula: At this time, refers to the soil moisture value observed from the soil moisture sensor in the preset wavelength range, is a dielectric constant agent of a preset wavelength range, are parameters of the regression model between soil moisture and dielectric constant observations.

4. In paragraph 1, The above conventional multiphase dielectric constant model is, A method for extracting organic carbon based on a microwave soil moisture sensor that estimates soil moisture from a dielectric surrogate as shown in the following mathematical formula: At this time, is the maximum permittivity of the bonding number, is the maximum permittivity of free water, It is an empirical constant.

5. In paragraph 4, The above dielectric constant maximum improvement model is, Organic carbon extraction method based on microwave soil moisture sensor, which is a model for improving the maximum value of bound and free water permittivity considering soil clay content and organic matter content as shown in the following mathematical formula in the above multiphase permittivity model: At this time, is a preset empirical constant.

6. In paragraph 1, To perform the above iterative estimation, the multiphase dielectric constant model is divided into three sections according to the soil moisture range, the first section is from soil moisture 0 to wilting point (WP), the second section is from wilting point to saturation point (P), and the third section is from saturation point to 1 [cm 3 / cm 3 ] is set to , and the dielectric constant for each model is calculated according to the mathematical formula below, and the median value is ranked according to the size and set as the simulated dielectric constant. Organic carbon extraction method based on microwave soil moisture sensor: At this time, is the permittivity for each model, is the simulated dielectric constant.

7. In paragraph 1, The step of extracting organic carbon from soil by performing optimization of the maximum dielectric constant model using the above optimal estimated initial value is as follows. A step of optimizing the dielectric constant maximum improvement model, which is subdivided into three continuous functions according to cases occurring below the saturation point, between the saturation point and the saturation point, and above the saturation point, respectively, under the dielectric constant proxy and the minimum simulation error; A step of ranking the converged soil moisture in each of the above maximum dielectric constant models and selecting the soil moisture of the middle rank; A step of calculating soil organic matter according to the selected soil moisture content; and A method for extracting organic carbon based on a microwave soil moisture sensor, comprising a step of calculating soil organic carbon from the above-mentioned produced soil organic matter.

Images