A method for estimating the bulk density of deep soil

Abstract

The present invention discloses a method for estimating the bulk density of deep soil, which relates to the technical field of soil science data estimation. In the present invention, the soil burial depth, the bulk density of shallow soil, the magnetic susceptibility, and the clay content are collected as input data. The soil compaction model is driven by the soil burial depth and the bulk density of shallow soil to calculate the change in the bulk density of deep soil caused by the overlying soil pressure stress. The soil transformation function is driven by the soil magnetic susceptibility and the clay content to calculate the change in the bulk density of deep soil determined by its own physical properties. The two parts of the bulk density are added together to obtain the estimated value of the bulk density of deep soil. Generally speaking, the present invention constructs a bulk density conceptual model with a certain mechanism. The model has high prediction accuracy, can directly estimate the bulk density of deep soil without parameter data calibration, has a wide range of applications, and helps to provide relatively reliable data support for research in aspects such as soil, hydrology, and ecology.

Description

Technical Field

[0001] The present invention relates to the technical field of soil science data estimation, and particularly to a method for estimating the bulk density of deep soil. Background Art

[0002] Soil bulk density, that is, the dry weight of undisturbed soil under unit volume, is a density characteristic representing the natural state of the soil mass. It is not only a mass-volume conversion coefficient for evaluating the reserves and fluxes of water, solutes, carbon and other substances in the soil, but also closely related to a series of soil, hydrological and ecological processes such as water infiltration, solute migration, and plant root growth, and is a key parameter for describing the material and energy processes of various critical zones of the earth.

[0003] However, different from other parameters (such as soil texture, water content, organic matter, etc.), the measurement of soil bulk density must meet the undisturbed nature of the soil structure, and common disturbed soils (filled and remolded soils) cannot reflect the true bulk density information of the soil. This is time-consuming, laborious and costly for the measurement of soil bulk density on a large scale and with high resolution in reality, especially for the bulk density of deep soil. The deeper the burial depth, the more difficult it is to collect undisturbed soil samples by the core method. Therefore, the measured data of deep soil bulk density have always been extremely limited.

[0004] For this reason, some studies have developed soil transfer functions (PTFs) for soil bulk density, using other related variables that are easy to measure to indirectly predict soil bulk density. There are many such methods, such as multiple linear regression, artificial neural network, etc. However, these methods are all empirical models, and relevant parameter data are required for calibration before the model runs. For the prediction of soil bulk density outside the data range driven by the model, the prediction result accuracy is not high, the prediction performance drops significantly, and it is difficult to be applied to the estimation of soil bulk density in different regions. Summary of the Invention

[0005] In view of the above application requirements and the existing problems of traditional methods for obtaining soil bulk density, a method for estimating the bulk density of deep soil with strong mechanism, good prediction performance and wide application range is needed; the present invention proposes a bulk density conceptual model with a certain mechanism to solve the problem of difficult acquisition of deep soil bulk density data at present. This model has high prediction accuracy, wide application range, and can directly estimate the bulk density of deep soil without parameter data calibration.

[0006] A method for estimating the bulk density of deep soil with a certain mechanism, good prediction performance and wide application range proposed by the present invention includes the following steps:

[0007] Obtain the burial depth h of the soil sample, as well as the bulk density ρ0 of the shallow soil, soil magnetic susceptibility MS, and clay content CL;

[0008] Build a soil compressive stress model, input the burial depth h of the soil sample into the soil compressive stress model, and output the compressive stress σ borne by the deep soil;

[0009] Build a soil compression model, input the compressive stress σ borne by the deep soil and the bulk density ρ0 of the shallow soil into the soil compression model, and output the change ρ in the compacted bulk density of the deep soil caused by the compressive stress; c ;

[0010] Build a multiple linear regression model, input the soil magnetic susceptibility MS and the clay content CL into the multiple linear regression model, and output the change ρ in the bulk density determined by the physical properties of the soil itself; p ;

[0011] According to the change mechanism of the deep soil bulk density, add the change ρ in the compacted bulk density of the deep soil caused by the compressive stress c to the change ρ in the bulk density determined by the physical properties of the soil itself p to obtain the predicted value of the deep soil bulk density

[0012] Furthermore, a method for estimating the deep soil bulk density provided by the present invention further includes:

[0013] Obtain the deep soil bulk density ρ according to the predicted value of the deep soil bulk density : s :

[0014]

[0015] wherein, is the predicted value of the deep soil bulk density; ρ c is the change value of the compacted bulk density of the deep soil caused by the compressive stress; ρ p is the change value of the bulk density determined by the physical properties of the soil itself; ε i is the error term, including sampling error and random error.

[0016] Furthermore, the building of the soil compressive stress model specifically includes:

[0017] According to experimental analysis, the compressive stress σ borne by the deep soil shows a linear correlation with the soil burial depth h:

[0018] σ = ρgh ≈ 17.2h - 28.9, R 2 = 0.99;

[0019] wherein, ρ is the average density of the soil mass covering above the depth of the deep soil sample / g.cm -3 ; g is the acceleration due to gravity / N.kg -1 ; h is the burial depth of the soil sample / m; R 2is the correlation coefficient;

[0020] Therefore, according to the calculation formula of the compressive stress σ borne by the deep soil, σ = ρgh ≈ 17.2h - 28.9, a soil compressive stress model is constructed.

[0021] Furthermore, the construction of the soil compression model specifically includes:

[0022] According to the experimental analysis, the change in the compacted soil bulk density ρ caused by the compressive stress in the deep soil c has the following calculation formula:

[0023]

[0024] where ρ0 is the initial bulk density of the model; A, B, and C are the coefficients of the soil compaction model;

[0025] According to the experimental analysis, when the compressive stress above the soil mass > 0.03 MPa, obvious compression changes will occur in the soil mass, and the load on the soil mass below a burial depth of 2 m > 0.03 MPa. Therefore, the average bulk density of the shallow soil with a depth of 0 - 2 m is used as the initial bulk density ρ0 of the model;

[0026] Through the optimization and fitting of experimental data, the change in the compacted soil bulk density ρ caused by the compressive stress in the deep soil c has the following calculation formula

[0027] According to the calculation formula of the change in the compacted soil bulk density ρ caused by the compressive stress in the deep soil c a soil compression model is constructed.

[0028] Furthermore, the construction of the multiple linear regression model specifically includes:

[0029] According to the experimental analysis, the change in the bulk density ρ of the soil determined by its own physical properties p has the following calculation formula:

[0030] ρ p = 0.0005MS + 0.003CL - 0.11;

[0031] where MS is the soil magnetic susceptibility / m 3 .kg -1 ; CL is the content of soil clay particles with a particle size less than 0.002 mm / %;

[0032] According to the calculation formula of the change in the bulk density ρ of the soil determined by its own physical properties p a multiple linear regression model is constructed.

[0033] Furthermore, the change in the bulk density ρ of the soil determined by its own physical properties pAmong them, the physical properties of the soil itself include soil particle composition and soil structure.

[0034] Compared with the prior art, a method for estimating the bulk density of deep soil provided by the present invention has certain mechanism, good prediction performance and wide application range, and its beneficial effects are as follows:

[0035] The present invention constructs a new concept model of soil bulk density by using easily measurable soil parameter data to estimate the distribution of deep soil bulk density; according to the soil compression principle, the present invention increases the change in soil bulk density caused by the compressive stress of the upper soil layer on the deep soil; and quantitatively describes the influence of the physical properties of the soil itself on the soil bulk density according to the soil transfer function. The deep soil bulk density model has certain mechanism, high prediction accuracy, and can directly estimate the deep soil bulk density without parameter data calibration, so that it can be widely applied in different places, which helps to provide relatively reliable data support for research in soil, hydrology, ecology and other aspects, and promotes the development of related sciences and technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a flowchart of a method for estimating the bulk density of deep soil provided by the present invention;

[0037] Figure 2 is a scatter diagram of the burial depth h, the bulk density ρ0 of the shallow soil within 0-2 m deep from the ground surface, the magnetic susceptibility MS and the clay content CL of the soil profile samples observed in the field at WN, ZW, and XY provided in the embodiment of the present invention;

[0038] Figure 3 is a linear correlation diagram of the compressive stress σ borne by the deep soil and the soil burial depth h provided in the embodiment of the present invention;

[0039] Figure 4 is a scatter distribution diagram of the change ρ in the compacted bulk density of the deep soil at the three field sites of WN, ZW, and XY calculated by the compressive stress model provided in the embodiment of the present invention c ;

[0040] Figure 5 is a scatter distribution diagram of the change ρ in the bulk density of the deep soil determined by its own physical properties at the three field sites of WN, ZW, and XY provided in the embodiment of the present invention p ;

[0041] Figure 6 is a scatter diagram of the estimated value and the measured value of the deep soil bulk density at the three field sites of WN, ZW, and XY provided in the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0042] Currently, the methods for obtaining the bulk density ρ s of deep soil can be divided into two categories:

[0043] (1) Core cutter in-situ measurement method: Use a standard core cutter (with a diameter and length of 5 cm and a volume of 100 cm 3 ) to collect undisturbed original soil samples from natural soil, and then put them into an oven at 105 °C and dry for more than 24 hours until constant weight. Weigh its total dry weight (including the core cutter), subtract the weight of the core cutter (m T ) from the total dry weight and divide by the volume (V) of the core cutter to obtain the soil bulk density ρ s (g / cm 3 ):

[0044]

[0045] The disadvantage of the core cutter in-situ measurement method is that in actual application, it is often time-consuming, laborious and costly; moreover, the deeper the soil burial depth, the more difficult it is to collect original soil samples with a core cutter. Therefore, the measured deep soil bulk density data by such methods have always been very limited and it is difficult to be widely promoted and used on a large scale.

[0046] (2) Traditional empirical models driven by easily measurable data: Generally, identify the influencing factors significantly related to bulk density at the measurement scale, and then establish soil transfer functions (PTFs) to indirectly predict soil bulk density using other easily measurable relevant variables. There are many such methods, such as multiple linear regression, artificial neural networks, etc. Among them, multiple linear regression is the most widely used, and the bulk density prediction equation can generally be written as:

[0047]

[0048] In the formula, is the predicted value of soil bulk density; ε i is the error term; a0 is the intercept; a i is the regression coefficient; ρ si is the i-th independent variable related to bulk density; n is the number of independent variables in the regression equation.

[0049] Although such methods can predict soil bulk density, the equations are all empirical models and need to be calibrated with relevant parameter data before the model runs; in addition, when predicting soil bulk density outside the data range driven by the model, the prediction results have low accuracy and the prediction performance drops significantly, making it difficult to be applicable to soil bulk density prediction in different regions.

[0050] Therefore, aiming at the many problems of the above traditional soil bulk density acquisition methods, a deep soil bulk density estimation method with strong mechanism, good prediction performance and wide application range is needed.

[0051] The following combines the attached Figures 1 to 6, a further description of the specific embodiments of the present invention is given. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and cannot be used to limit the protection scope of the present invention.

[0052] Example 1: As Figures 1-6 shown, the present invention proposes a method for estimating the bulk density of deep soil, which specifically includes the following steps:

[0053] Step 1. According to the change mechanism of the bulk density of deep soil, the bulk density ρ of deep soil s is mainly decomposed into two parts: the change in soil compaction bulk density caused by external compressive stress and the change in bulk density determined by the physical properties of the soil itself (such as soil particle composition and soil structure).

[0054] In this example, the formula for the bulk density ρ of deep soil s is:

[0055]

[0056] In the formula, is the predicted value of soil bulk density; ρ c is the part of the change in soil compaction bulk density caused by compressive stress; ρ p is the part of the change in bulk density determined by the physical properties of the soil itself; ε i is the error term, which is mainly composed of random errors such as sampling and sample measurement, and can be approximately ignored in this example.

[0057] Step 2. Collect the input data of the soil bulk density model. The input data includes the burial depth h of the soil sample, the bulk density ρ0 of the shallow soil within 0 - 2 m from the surface, the magnetic susceptibility MS, and the clay content CL.

[0058] In this example, 3 sets of measured data of soil deep profiles are selected as the input data for the evaluation of the method of the present invention. As shown in Table 1, these 3 soil deep profiles WN, ZW, and XY are located in different regions of China (Weinan City, Shaanxi Province; Zhongwei City, Ningxia Hui Autonomous Region; Pingliang City, Gansu Province), representing different climate types (warm temperate semi - humid, mid - temperate semi - arid, mid - temperate semi - humid), land use types (farmland, wasteland grassland, forest land), and soil types (sandy loess, loess, clay loess). As Figure 2 shown, the scatter plots of the burial depth h, the bulk density ρ0 of the shallow soil, the magnetic susceptibility MS, and the clay content CL of the three field - observed soil profiles WN, ZW, and XY. Among them, soil samples are collected at 20 - cm depth intervals for each soil profile. The basic statistical information of the observed model input data is shown in Table 2.

[0059] Table 1 Basic information of soil profiles

[0060]

[0061] Table 2 Basic statistical information of the model input data

[0062]

[0063] Step 3: Drive the compressive stress model through the burial depth h of the soil sample to obtain the compressive stress σ borne by the deep soil;

[0064] In this example, the calculation formula for the compressive stress σ borne by the deep soil is:

[0065] σ = ρgh ≈ 17.2h - 28.9;

[0066] In the formula, ρ is the average density of the soil mass covering above the depth of the deep soil sample (g cm -3 ); g is the acceleration due to gravity (Nkg -1 ); h is the burial depth of the soil sample (m). According to the experimental analysis ( Figure 3 ), it is found that the compressive stress σ borne by the deep soil is not sensitive to the change in the average density ρ of the overlying soil mass, but has a highly linear correlation with the soil burial depth h (σ = ρgh ≈ 17.2h - 28.9, R 2 = 0.99). Therefore, the compressive stress σ borne by the deep soil can be estimated using the relationship formula between the compressive stress σ and the burial depth h.

[0067] Step 4: Drive the soil compression model through the compressive stress σ and the bulk density ρ0 of the shallow soil to calculate the change in the soil compacted bulk density ρ caused by the compressive stress on the deep soil c ;

[0068] In this example, the change in the soil compacted bulk density ρ c caused by the compressive stress is calculated as follows:

[0069]

[0070] In the formula, ρ0 is the initial bulk density of the model. Extensive research shows that obvious compression changes in the soil mass occur only when the compressive stress above the soil mass > 0.03 MPa, and the load on the soil mass below a burial depth of 2 m is basically > 0.03 MPa. Therefore, in the present invention, the average bulk density of the shallow soil within a depth of 0 - 2 m is used as the initial bulk density ρ0 of the model; A, B, and C are the coefficients of the soil compression model. Through optimization fitting of a large amount of experimental data, the following are obtained:

[0071]

[0072] Such as Figure 4As shown, the scatter plot of the variation of the compacted bulk density ρ of the deep soil at three field sites WN, ZW, and XY calculated by the compaction stress model driven by the burial depth of the three soil profiles of WN, ZW, and XY and the average bulk density ρ0 of the shallow soil c Scatter plot of variation

[0073] Step 5: Use the data of soil sample magnetic susceptibility and clay content to drive a multiple linear regression model, establish a soil conversion function, and simulate the change in bulk density ρ determined by the physical properties of the soil itself p Was carried out

[0074] In this example, the change in bulk density ρ determined by the physical properties of the soil itself p The calculation formula is:

[0075] ρ p = 0.0005MS + 0.003CL - 0.11;

[0076] In the formula, MS is the soil magnetic susceptibility; CL is the soil clay content

[0077] Figure 5 Is the scatter plot of the change in the bulk density ρ of the deep soil determined by the physical properties of the soil itself calculated from the data of the soil magnetic susceptibility and clay content of the three soil profiles of WN, ZW, and XY p Scatter plot of variation

[0078] Adding the above calculated ρ c To ρ p Correspondingly, the bulk density ρ of the deep soil can be obtained s . The scatter plot of the estimated value and the measured value of the bulk density of the deep soil calculated by the method of the present invention for the three soil profiles of WN, ZW, and XY is as shown in Figure 6 Shown

[0079] Table 3 and Table 4 respectively show the evaluation of the estimation performance of the bulk density of the deep soil of the soil profiles of WN, ZW, and XY by the method of the present invention and a traditional empirical model estimation method (such as ρ s = 1.22 + 0.0019h + 0.0074CL - 0.021OC). The evaluation indexes include the adjusted determination coefficient adj-R 2 , mean deviation ME, standard deviation of error SDE, and root mean square error RMSE. Among them, adj-R 2 Is a measure of the strength of the linear relationship between the measured value and the estimated value. The adj-R 2 For the bulk density estimation of the WN, ZW, and XY profiles by the method of the present invention is 0.70, 0.75, and 0.77 (Table 3), all higher than the adj-R 2(Table 4) reflects the high interpretability of the method of the present invention for the bulk density of deep soil, with the interpretability reaching over 70%; MPE is the systematic bias of the prediction model. The ME of the method of the present invention for estimating the bulk density of the WN, ZW, and XY profiles is less than that of the traditional empirical model estimation method, reflecting the high prediction accuracy of the method of the present invention; SDE represents the random variation of the predicted value after correcting the global bias. In the examples, the SDE of the method of the present invention for estimating the bulk density of the WN, ZW, and XY profiles is much less than that of the traditional empirical model estimation method, reflecting that the prediction accuracy of the method of the present invention is also better than that of the empirical model estimation method; It can also be seen from the examples that the standard deviation RMSE of the total error of the estimation method of the present invention is much lower than that of the traditional empirical model method, indicating that the prediction performance of the estimation method of the present invention for the bulk density of deep soil has been greatly improved compared with the traditional empirical model method.

[0080] Table 3 Performance of the method of the present invention for estimating the bulk density of deep soil

[0081]

[0082] Table 4 Performance of the traditional empirical model driven by data for estimating the bulk density of deep soil

[0083]

[0084] Generally speaking, the present invention constructs a new conceptual model of soil bulk density and estimates the bulk density of deep soil by using easily measurable soil parameter data. According to the soil compression principle, the change in soil bulk density caused by the compressive stress of the upper soil layer on the deep soil is increased; based on the soil transfer function, the magnetic susceptibility MS of the soil parameter is added at the same time to quantitatively describe the influence of the physical properties of the soil itself on the soil bulk density. This soil bulk density model has a certain mechanism, high prediction accuracy, and can directly estimate the bulk density of deep soil without parameter data calibration, making it widely applicable in different places and having strong applicability.

[0085] The above-described embodiments are only preferred specific embodiments of the present invention, and the protection scope of the present invention is not limited thereto. Any simple changes or equivalent replacements of the technical solutions that can be obviously obtained by those skilled in the art within the technical scope disclosed by the present invention all belong to the protection scope of the present invention.

Claims

1. A method for estimating the bulk density of deep soil, characterized in that, It includes the following steps: Obtain the burial depth h of the soil sample, as well as the bulk density ρ0 of the shallow soil, the soil magnetic susceptibility MS, and the clay content CL; Construct a soil compressive stress model, input the burial depth h of the soil sample into the soil compressive stress model, and output the compressive stress σ borne by the deep soil; Build a soil compression model, input the compressive stress σ borne by the deep soil and the unit weight ρ0 of the shallow soil into the soil compression model, and output the change ρ in the compacted unit weight of the deep soil caused by the compressive stress c ; Build a multiple linear regression model, input the soil magnetic susceptibility MS and clay content CL into the multiple linear regression model, and output the bulk density change ρ determined by the physical properties of the soil itself p ; According to the change mechanism of the deep soil bulk density, add the change in the compacted soil bulk density ρ c caused by the compressive stress in the deep soil to the change in the bulk density ρ p determined by the physical properties of the soil itself to obtain the predicted value of the deep soil bulk density 2. The method for estimating the bulk density of deep soil according to claim 1, characterized in that, It also includes: According to the predicted value of the deep soil bulk density Obtain the deep soil bulk density ρ s : Among them, is the predicted value of the deep soil bulk density; ρ c is the change value of the soil compaction bulk density caused by the compressive stress of the deep soil; ρ p is the change value of the bulk density determined by the physical properties of the soil itself; ε i is the error term, including sampling error and random error.

3. The method for estimating the bulk density of deep soil according to claim 1, characterized in that, The construction of the soil compressive stress model specifically includes: According to experimental analysis, the compressive stress σ borne by the deep soil shows a linear correlation with the soil burial depth h: σ = ρgh ≈ 17.2h - 28.9, R 2 = 0.99; where ρ is the average density of the overlying soil mass above the depth of the deep soil sample / g·cm -3 ; g is the acceleration due to gravity / N·kg -1 ; h is the burial depth of the soil sample / m; R 2 is the correlation coefficient; Therefore, according to the calculation formula of the compressive stress σ borne by the deep soil σ = ρgh ≈ 17.2h - 28.9, a soil compressive stress model is constructed.

4. The method for estimating the bulk density of deep soil according to claim 1, wherein, The construction of the soil compression model specifically includes: According to the test analysis, the change in the compacted soil bulk density ρ c caused by the compressive stress in the deep soil is calculated by the following formula: Among them, ρ0 is the initial bulk density of the model; A, B, and C are soil compaction model coefficients; According to experimental analysis, obvious compression changes in the soil will occur only when the compressive stress above the soil is > 0.03 MPa, and the load on the soil below a burial depth of 2 m is > 0.03 MPa. Therefore, the average bulk density of the shallow soil at a depth of 0 - 2 m is used as the initial bulk density ρ0 of the model; After optimizing and fitting the experimental data, the change in the soil compaction bulk density ρ c caused by the compressive stress in the deep soil is obtained, and the calculation formula is Based on the calculation formula of the change in the compacted soil bulk density ρ caused by the compressive stress in the deep soil c a soil compression model is constructed.

5. The method for estimating the bulk density of deep soil according to claim 1, characterized in that The construction of the multiple linear regression model specifically includes: According to the test analysis, the bulk density change ρ determined by the physical properties of the soil p is calculated by the following formula: ρ p = 0.0005MS + 0.003CL - 0.11; where MS is the soil magnetic susceptibility / m 3 .kg -1 ; CL is the soil clay content with particle size less than 0.002 mm / %; Construct a multiple linear regression model according to the calculation formula of the bulk density change ρ determined by the physical properties of the soil itself. p ​ 6. A method for estimating the bulk density of deep soil according to claim 1, characterized in that: The bulk density change ρ of the soil determined by its own physical properties p Among them, the soil's own physical properties include soil particle composition and soil structure.

Images