Root layer soil water content prediction method based on remote sensing driving hydrodynamic simulation

Abstract

The invention discloses a root layer soil water content prediction method based on remote sensing driving hydrodynamic simulation. The method comprises the following steps: step 1, acquiring basic data of a research area; 2, inverting the water content of surface soil; step 3, correcting a root table relation; 4, soil water content simulation and correction relation determination; 5, predicting the water content of the root layer soil. According to the method, the limitation that optical remote sensing is influenced by weather is overcome, and the time continuity of data acquisition is improved; a root table relation is established, surface soil water content information is converted into root layer soil water content information, and the problem that root layer soil cannot be directly detected by a remote sensing technology is solved; the HYDRUS-1D model is used for dynamically simulating the soil moisture, so that the change of the moisture content of root layer soil in a long-time sequence can be simulated; by establishing a correction relation, the accuracy of a model prediction result is improved, and meanwhile, the calculation potential and time continuity of a regional scale are realized.

Description

Technical Field

[0001] This invention relates to the field of soil moisture content prediction technology, specifically to a method for predicting root soil moisture content using remote sensing-driven hydrodynamic simulation. Background Technology

[0002] Root soil moisture content is a crucial variable in hydrological and climate prediction models, regulating the water and energy balance between soil, vegetation, and the atmosphere by influencing topsoil evaporation and vegetation transpiration. Current research on root soil moisture content monitoring and prediction mainly employs three technical approaches: first, obtaining topsoil moisture content using remote sensing methods and then extrapolating root soil moisture content through empirical relationships; second, simulating or predicting root soil moisture content using physical models such as HYDRUS and SWAT; and third, predicting root soil moisture content using machine learning models such as random forests and LSTM.

[0003] However, existing methods for predicting root layer soil moisture content have several shortcomings. On the one hand, due to the limited penetration capability of satellite electromagnetic signals, they cannot detect root layer soil information. Single remote sensing methods can only retrieve surface soil moisture content, and their temporal resolution is limited, failing to obtain temporally continuous and spatially complete surface soil moisture content retrieval. On the other hand, while the HYDRUS model, with its clear physical mechanism, can simulate long-term root layer soil moisture content, it cannot simulate or predict root layer soil moisture content at the regional scale. Machine learning methods, being "black box" models, have unclear physical mechanisms, resulting in significant uncertainty in their output. Therefore, there is an urgent need for a root layer soil moisture content prediction method with a clear physical mechanism, regional-scale computational potential, and temporal continuity. Summary of the Invention

[0004] The purpose of this invention is to provide a method for predicting root layer soil moisture content through remote sensing-driven hydrodynamic simulation. This method fully utilizes the advantage of microwave remote sensing, which is unaffected by rainy weather, to obtain the surface soil moisture content (0-10 cm) through microwave remote sensing inversion. The time interval for obtaining the surface soil moisture content is stable. The root layer soil moisture content is obtained through the relationship between the root layer soil moisture content and the surface soil moisture content (root-surface relationship). This method can simulate and predict surface and root layer soil moisture content according to actual needs, thereby solving the technical problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] This invention discloses a method for predicting root zone soil moisture content using remote sensing-driven hydrodynamic simulation, the method comprising the following steps:

[0007] Step 1: Obtain basic data for the study area: Obtain basic data for the study area, including satellite remote sensing data, product data, and actual observation data; the satellite remote sensing data includes Sentinel-1 and Sentinel-2 satellite imagery data; the product data includes MOD15A2H data and soil property data; the soil property data includes soil bulk density, sand content, silt content, and clay content; the actual observation data includes meteorological data, soil moisture content data, and leaf area index data; the meteorological data includes solar radiation R... a Precipitation, maximum air temperature T max Minimum air temperature T min The soil moisture content data refers to the soil moisture content at different measurement depths.

[0008] The reference crop evapotranspiration (ET) was calculated using the Hargreaves-Samani model. o The calculation formula is as follows:

[0009] (2)

[0010] In the formula: R a For solar radiation, MJ / (m 2 ·d); T mean The mean of the maximum and minimum air temperatures is ℃; C is the temperature coefficient; a is the temperature constant; m is the temperature index. It is the sum of the maximum air temperature and the minimum air temperature;

[0011] Step 2, Inverting surface soil moisture content: Using the improved water cloud model based on the modified chlorophyll absorption rate index (MCARI), the surface soil moisture content is inverted based on satellite image data; the soil depth of the surface soil is 0-10cm.

[0012] The expression for retrieving surface soil moisture content is:

[0013] (3)

[0014] In the formula: SSM is the surface soil moisture content, in cm. 3 / cm 3 ; denoted as Sentinel-1 total backscattering coefficient, in dB; θ is the incident angle of the Sentinel-1 image; VI is the MCARI, calculated from Sentinel-2 satellite image data.

[0015] The calculation method for MCARI is as follows:

[0016] (4)

[0017] In the formula: MCARI is the modified chlorophyll absorptivity index; VNIR is the near-infrared reflectance; Red is the red band reflectance; Green is the green band reflectance;

[0018] Step 3, root-to-surface relationship correction: The relationship between surface soil water storage and root soil water storage known from the literature is converted into the relationship between surface soil water content and root soil water content using formula (5), i.e., root-to-surface relationship, and then corrected.

[0019] The formula for converting soil water storage into soil moisture content is:

[0020] (5)

[0021] In the formula: θ n Let n be the soil moisture content of the nth layer, in cm. 3 / cm 3 ;∆Z n To observe soil layer thickness, in cm; D n S represents the depth of the nth layer from the Earth's surface, in cm; Dn For a depth of D n Soil water storage capacity of the soil layer, mm; n=1, 2, 3, 4, 5, 6;

[0022] The soil moisture content of each layer obtained by formula (5) is compared with the measured values ​​of soil moisture content of each layer, and the coefficient of determination R is calculated. 2 To determine whether the root-table relationship meets the requirements, if not, the root-table relationship is established by measuring the soil moisture content over a long period of time in the study area.

[0023] Step 4: Soil moisture content simulation and correction relationship determination: Use the HYDRUS-1D model to simulate soil moisture content and establish the correction relationship for soil moisture content in each layer;

[0024] Input the soil bulk density, soil sand content, soil silt content, soil clay content, precipitation, leaf area index, and reference crop evapotranspiration obtained in step 1. Set the initial conditions, i.e., the initial soil moisture content and boundary conditions. The upper boundary is an atmospheric boundary with no water accumulation, and the lower boundary is a free drainage boundary. Establish a HYDRUS-1D model to simulate the daily soil moisture content of each layer. Compare the simulated soil moisture content with the measured soil moisture content to obtain the correction relationship of soil moisture content of each layer.

[0025] Soil moisture movement in the HYDRUS-1D model follows the law of conservation of mass and Darcy's law. Combining these two laws yields the fundamental equations describing soil moisture movement under unsaturated conditions:

[0026] (6)

[0027] In the formula: θ is the soil moisture content, in cm 3 / cm 3 h represents matrix potential in cm; t represents time in d. -1 K represents unsaturated hydraulic conductivity, cm / d; z represents spatial coordinates, cm; S represents the source and sink terms of water absorption by crop roots, d. -1 The Kh and θ-h relationships in the van Genuchten-Mualem model are used to describe the hydraulic properties of unsaturated soil and the characteristic curves of soil moisture.

[0028] The expression for the water-conducting function is as follows:

[0029] (7)

[0030] (8)

[0031] (9)

[0032] (10)

[0033] Where: K s θ represents saturated hydraulic conductivity, in cm / d; r Residual soil moisture content, in cm 3 / cm 3 ;θ s The saturated water content is expressed in cm. 3 / cm 3 α, m, and n are the shape coefficients of the soil moisture characteristic curve; S e denoted as dimensionless available water content; l is the porosity parameter, taken as 0.5; θ(h) is the soil moisture content, in cm³. 3 / cm 3 K(h) is the unsaturated hydraulic conductivity, cm / d;

[0034] Step 5: Prediction of root zone soil moisture content: Establish a HYDRUS-1D model for the prediction period of the study area. First, the surface soil moisture content on the first day of the prediction period is obtained by inversion. Then, the root zone soil moisture content is obtained from the root-surface relationship. As the initial condition of the model, the daily soil moisture content of each layer during the prediction period is simulated. The prediction results are then corrected using the correction relationship to obtain the corrected prediction value.

[0035] Furthermore, the leaf area index in step 1 is obtained by measuring the leaf length and maximum leaf width of each effective leaf of each plant with a ruler, and the calculation formula is as follows:

[0036] (1)

[0037] In the formula: LAI is the leaf area index; ρ 种 Planting density, plants / m 2 n represents the number of effective leaves in all representative plants; a i For each effective leaf length, m; b i The maximum leaf width for each effective leaf is measured in meters (m); c represents the number of crop plants being measured; leaf area index data are obtained by linear interpolation of the daily leaf area index during the crop's growth period.

[0038] Furthermore, regarding formula (2) in step 1, according to the recommendations of the Food and Agriculture Organization of the United Nations, when R a The unit is MJ / (m 2 When ·d), C, a and m take values ​​of 0.0023, 17.8 and 0.5 respectively.

[0039] Furthermore, the criterion for determining whether the root table relation meets the requirements in step 3 is R. 2 ≥0.7.

[0040] Furthermore, the long-term series described in step 3 includes at least 100 sample data.

[0041] The beneficial effects of this invention are:

[0042] 1) This invention utilizes remote sensing technology to retrieve surface soil moisture content, overcoming the limitations of optical remote sensing in response to weather conditions and improving the temporal continuity of data acquisition;

[0043] 2) This invention establishes a root-surface relationship, transforming surface soil moisture content information into root soil moisture content information, thus solving the problem that remote sensing technology cannot directly detect root soil.

[0044] 3) This invention uses the HYDRUS-1D model to simulate soil moisture dynamics. This model has a clear physical mechanism and can simulate long-term series of changes in root zone soil moisture content.

[0045] 4) This invention improves the accuracy of model prediction results by establishing a correction relationship, making them closer to the actual situation;

[0046] 5) The method of this invention combines the advantages of remote sensing and physical models, ensuring the clarity of the physical mechanism while realizing the computational potential and temporal continuity at the regional scale. Attached Figure Description

[0047] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram of the method flow described in Example 1;

[0049] Figure 2 This is a schematic diagram of the geographical locations of soil moisture monitoring points in Example 1;

[0050] Figure 3 This is a schematic diagram of soil water storage in Example 1;

[0051] Figure 4 This is a schematic diagram comparing the soil moisture content obtained from the root-surface relationship with the measured soil moisture content in Example 1.

[0052] Figure 5 This is a schematic diagram comparing the simulated and measured values ​​of soil moisture content at different layers of observation point 1 in Example 1.

[0053] Figure 6 This is a schematic diagram comparing the simulated and measured values ​​of soil moisture content at different layers in observation point 2 of Example 1.

[0054] Figure 7 This is a schematic diagram comparing the predicted values, corrected values, and measured values ​​of soil moisture content in different soil layers at observation point 1 in Example 1.

[0055] Figure 8 This is a schematic diagram comparing the predicted values, corrected values, and measured values ​​of soil moisture content in different soil layers at observation point 2 in Example 1. Detailed Implementation

[0056] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0057] Example 1

[0058] This embodiment discloses a method for predicting root zone soil moisture content using remote sensing-driven hydrodynamic simulation, such as... Figure 1 As shown, the method includes the following steps:

[0059] Step 1: Obtain basic data for the study area:

[0060] The study area will acquire basic data, including satellite remote sensing data, product data, and actual observation data. Satellite remote sensing data includes Sentinel-1 and Sentinel-2 satellite imagery, which can be obtained from the website: https: / / dataspace.copernicus.eu / . Product data includes MOD15A2H data and soil property data; soil property data includes soil bulk density, percentage of sand, percentage of silt, and percentage of clay. MOD15A2H data can be obtained from the website: https: / / ladsweb.nascom.nasa.gov / search; soil property data is sourced from the 1:1,000,000 Chinese Soil Database and can be obtained from the official website of the Nanjing Institute of Soil Science, Chinese Academy of Sciences. The actual observation data includes meteorological data, soil moisture content data, and leaf area index data. Meteorological data includes solar radiation, precipitation, maximum air temperature, and minimum air temperature, which can be obtained from the National Meteorological Information Center. Soil moisture content data includes soil moisture content measured at a depth of 0-60 cm. In this embodiment, the soil moisture content data was obtained through observations at the Daxing Test Base of the China Institute of Water Resources and Hydropower Research. See Table 1 for details.

[0061] Table 1. Sources of Basic Data

[0062]

[0063] This embodiment involved a summer maize field experiment conducted at the Daxing Experimental Base of the China Institute of Water Resources and Hydropower Research from 2022 to 2023. Daily soil moisture content (0-5, 5-10, 10-15, 15-20, 20-40, and 40-60 cm) was monitored in five plots from July 17 to October 4, 2022, and July 17 to October 4, 2023. Simultaneously, an automatic weather station was installed at the experimental base to monitor daily rainfall and temperature. The measured soil moisture data from 2022 were used for model calibration, while the measured soil moisture data from 2023 were used to verify the predicted daily root zone soil moisture content at the five monitoring points. The specific locations of the five monitoring points are as follows: Figure 2 As shown.

[0064] The satellite imagery data used in this invention comprises Sentinel-1 and Sentinel-2 satellite imagery data. Sentinel-1 is a key component of the European Space Agency's proposed global environmental and safety monitoring system. It consists of two satellites (Sentinel-1A and Sentinel-1B), successfully launched on April 3, 2014, and April 25, 2016, respectively. The temporal resolution of a single satellite is 12 days, decreasing to 6 days after the two satellites are deployed together. Sentinel-1 is equipped with a C-band synthetic aperture radar (SAR) sensor with two polarization modes (VV and VH) and operates at a frequency of 5.4 GHz. The Sentinel-1 imagery used in this invention is a Level 1 product in IW mode with SLC format, while the Sentinel-2 satellite imagery data is multispectral imager (MSI) spectral band data.

[0065] Leaf area index (LAI) is obtained by measuring the length and maximum width of each effective leaf of each plant with a ruler. The calculation formula is as follows:

[0066] (1)

[0067] In the formula: LAI is the leaf area index; ρ 种 Planting density, plants / m 2 n represents the number of effective leaves in all representative plants; a i For each effective leaf length, m; b i The maximum leaf width for each effective leaf is expressed in meters (m); c represents the number of crop plants measured. The leaf area index data in this embodiment were obtained by linear interpolation of the daily leaf area index during the summer maize growing season.

[0068] This embodiment obtained soil property data from the official website of the Nanjing Institute of Soil Science, Chinese Academy of Sciences, and obtained the soil bulk density and percentage content of sand, loam, and clay in two soil layers: 0-30 cm and 30-100 cm. In reality, the root layer involved in this embodiment is only 60 cm, therefore, the soil bulk density, percentage content of sand, silt, and clay in two soil layers (0-30 cm and 30-60 cm) were obtained. Based on the soil property data, the saturated hydraulic conductivity K of the two soil layers was predicted using the Rosetta Lite module. s Residual moisture content θ r saturated water content θ s The scale parameter α and shape parameter n are shown in Table 2.

[0069] Table 2 Soil parameters at monitoring points

[0070]

[0071] Meteorological data were measured by automatic weather stations installed at the Daxing test base, and the observed data included solar radiation R. a Maximum air temperature T max Minimum air temperature T min Rainfall data is output every 30 minutes.

[0072] The reference crop evapotranspiration (ET) was calculated using the Hargreaves-Samani (HS) model. o The calculation formula is as follows:

[0073] (2)

[0074] In the formula: R a For solar radiation, MJ / (m 2 ·d); T mean R is the average of the maximum and minimum air temperatures, in °C; C is the temperature coefficient; a is the temperature constant; m is the temperature index. The Food and Agriculture Organization of the United Nations (FAO) recommends that R... a The unit is MJ / (m 2 When ·d), C, a, and m take values ​​of 0.0023, 17.8, and 0.5, respectively; It is the sum of the maximum air temperature and the minimum air temperature.

[0075] Soil moisture content was measured using an FDR62 soil temperature and humidity sensor. The sensor was placed horizontally in the soil layer, buried 5 cm below the surface. Six layers of the sensor were installed at depths of 5, 10, 15, 20, 40, and 60 cm below the surface. Data was automatically recorded by a CK300 industrial field data logger, outputting data every 30 minutes. The soil moisture content described above is volumetric moisture content, which is the ratio of the volume of water in the soil to the total soil volume.

[0076] Step 2: Invert the surface soil moisture content:

[0077] The surface soil moisture content retrieved by remote sensing has a soil depth of 0-10 cm, which can obtain spatial soil moisture content data over a large area. The acquisition date is determined based on the Sentinel-1 satellite's transit date.

[0078] The surface soil moisture content retrieved by remote sensing is based on a water cloud model. The principle of the water cloud model is that the total backscattering coefficient equals the sum of the vegetation backscattering coefficient and the soil backscattering coefficient. The surface soil moisture content retrieval method used in this embodiment utilizes a modified water cloud model based on the Modified Chlorophyll Absorption Index (MCARI), retrieving surface soil moisture content from satellite imagery data. This model has been applied in North China, demonstrating high estimation accuracy. 2≥0.85, RMSE≤0.028 cm 3 / cm 3 MAE ≤ 0.023 cm 3 / cm 3 .

[0079] The expression for retrieving surface soil moisture content is as follows:

[0080] (3)

[0081] In the formula: SSM is the surface soil moisture content, in cm. 3 / cm 3 ; θ is the total backscattering coefficient of Sentinel-1, in dB; θ is the incident angle of the Sentinel-1 image; VI is the MCARI, calculated from Sentinel-2 satellite image data.

[0082] The calculation method for MCARI is as follows:

[0083] (4)

[0084] In the formula: MCARI is the modified chlorophyll absorptivity index; VNIR is the near-infrared reflectance; Red is the red band reflectance; Green is the green band reflectance.

[0085] Step 3: Root table relationship correction:

[0086] The relationship between surface soil water storage and root soil water storage known from the literature (source: Liu Suxia, Acta Phytoecologica Sinica, Analysis of the Relationship between Root and Surface Soil Moisture in China) was transformed into the relationship between surface soil water content and root soil water content, i.e., the root-surface relationship, through formula (5), and then corrected.

[0087] The relationship between known surface soil water storage and root zone soil water storage is used to convert soil water storage into soil moisture content using formula (5). For example, the root zone soil of summer maize is divided into 6 layers. Figure 3 As shown, the soil moisture content of the nth (n=1, 2, 3, 4, 5, 6) layer is θ. n (cm) 3 / cm 3 The observed soil layer thickness is ∆Z. n (cm), the depth of the nth layer from the surface is D. n (cm). Depth measured in water height equivalent (mm) is D. n The soil water storage capacity of the soil layer is S Dn The formula for converting soil water storage into soil moisture content is:

[0088] (5)

[0089] In the formula: θ n Let n be the soil moisture content of the nth layer, in cm. 3 / cm 3 ;∆Z n To observe soil layer thickness, in cm; D n S represents the depth of the nth layer from the Earth's surface, in cm; Dn For a depth of D n Soil water storage capacity of the soil layer, mm; n=1, 2, 3, 4, 5, 6.

[0090] The soil moisture content of each layer obtained by formula (5) is compared with the measured values ​​of soil moisture content of each layer, and the coefficient of determination R is calculated. 2 Determine whether the root table relation meets the requirements, i.e., R. 2 If the value is ≥0.7, the root-table relationship is established by measuring the soil moisture content over a long time series in the study area; the long time series includes at least 100 sample data.

[0091] The relationship between the transformed surface soil moisture content and the root layer soil moisture content at different depths is shown in Table 3.

[0092] Table 3. Relationship between surface soil moisture content and root zone soil moisture content

[0093]

[0094] The soil moisture content obtained by correcting the root-surface relationship was obtained by using the measured soil moisture content of each layer at monitoring point 1 from July 17 to October 4, 2022. Since the soil depth of the measured soil moisture content is inconsistent with the soil depth defined by the model, the soil moisture content needs to be processed. The soil depths of the measured soil moisture content are 0-5, 5-10, 10-15, 15-20, 20-40, and 40-60 cm, while the soil depths defined by the model are 0-10, 10-20, 20-30, 30-40, 40-50, and 50-60 cm. The 0-5 and 5-10 cm soil moisture content is used to calculate the 0-10 cm water storage using formula (5), and then the 0-10 cm soil moisture content is obtained as the surface soil moisture content in the root-surface relationship. Soil moisture content at depths of 10-20, 20-30, 30-40, 40-50, and 50-60 cm was obtained using equation (5) and compared with the measured values ​​of soil moisture content at each layer, as shown in the figure. Figure 4 As shown, the coefficient of determination R between the soil moisture content calculated from the root surface relationship and the measured soil moisture content increases with soil depth. 2 As the value gradually decreases, the fit gradually diminishes, but R... 2All values ​​were ≥0.82. This confirms the applicability of the root table relationship. If it is not applicable, a root table relationship needs to be established using measured soil moisture content from a long-term time series (at least 100 samples) in the local area. The initial soil moisture content for the prediction period is then used... Figure 4 The relational expressions in the text are corrected.

[0095] Step 4: Soil moisture content simulation and correction relationship determination:

[0096] Soil moisture content was simulated using the HYDRUS-1D model, and a correction relationship for soil moisture content in each layer was established.

[0097] For monitoring points 1-5, from July 17 to October 4, 2022, input the soil bulk density, soil sand content, soil silt content, soil clay content, precipitation, LAI, and ET obtained in step 1. o An initial soil moisture content and boundary conditions (upper boundary: atmospheric boundary with no water accumulation; lower boundary: free drainage boundary) were set to establish a HYDRUS-1D model to simulate daily soil moisture content at depths of 0-10, 10-20, 20-30, 30-40, 40-50, and 50-60 cm. It should be noted that the boundary conditions were set based on actual conditions. The model's selectable upper boundary conditions include: constant head, constant flux, atmospheric boundary with water accumulation, atmospheric boundary without water accumulation, variable head, variable flux, and irrigation; the lower boundary conditions include: constant head, constant flux, variable head, variable flux, free drainage, deep drainage, seepage surface, and horizontal drainage.

[0098] Taking monitoring points 1 and 2 as examples, this paper presents a comparison between simulated and measured soil moisture content values ​​from July 17 to October 4, 2022. Figure 5 and Figure 6 As shown. By combining measured and simulated data of soil moisture content at each layer from five observation points, the correction relationships for soil moisture content at each layer were obtained, with the aim of using them to correct soil moisture content during the predicted summer maize growing season of 2023.

[0099] The soil moisture content at each of the five monitoring points showed similar trends from July 17, 2022 to October 4, 2022. The simulated and measured values ​​exhibited similar trends and a high degree of overlap. By summing the simulated and measured data from the five observation points (400 sets of data for each layer), the correction relationships for each soil layer were obtained, as shown in Table 4. (Coefficient of determination R...) 2 The values ​​range from 0.417 to 0.808, and the root mean square error (RMSE) ranges from 0.053 to 0.091 cm. 3 / cm 3 The mean absolute error (MAE) ranges from 0.213 to 0.255 cm. 3 / cm 3 .

[0100] Table 4 Correction relationships for soil moisture content at different depths

[0101]

[0102] The HYDRUS model, developed in 1991 by the USDA Saline Soil Laboratory, is a computer model used to simulate soil moisture movement in variable-saturation porous media. The model's computational dimensions include one-dimensional, two-dimensional, and three-dimensional models, corresponding to HYDRUS-1D, HYDRUS-2D, and HYDRUS-3D, respectively. This embodiment only considers one-dimensional vertical soil moisture movement. The HYDRUS-1D model has a simple and flexible input / output interface. Upper boundary conditions include constant pressure head, atmospheric conditions, constant flux, and variable head; lower boundary conditions include deep drainage, constant head, and water flux. Initial conditions include pressure head and soil moisture content. For simulations of different crops, the HYDRUS-1D model provides corresponding root water uptake modules.

[0103] Specifically, soil moisture movement in the HYDRUS-1D model follows the law of conservation of mass and Darcy's law. Combining these two laws yields the fundamental equations describing soil moisture movement under unsaturated conditions:

[0104] (6)

[0105] In the formula: θ is the soil volumetric water content, in cm. 3 / cm 3 h represents matrix potential in cm; t represents time in d. -1 K represents unsaturated hydraulic conductivity, cm / d; z represents spatial coordinates, cm; S represents the source and sink terms of water absorption by crop roots, d. -1 This embodiment uses the Kh and θ-h relationships in the van Genuchten-Mualem (VG) model to describe the hydraulic properties of unsaturated soil and soil moisture characteristic curves. The expression for the hydraulic conductivity function is as follows:

[0106] (7)

[0107] (8)

[0108] (9)

[0109] (10)

[0110] Where: K s θ represents saturated hydraulic conductivity, in cm / d; r Residual soil moisture content, in cm 3 / cm 3 ;θ sThe saturated water content is expressed in cm. 3 / cm 3 α, m, and n are the shape coefficients of the soil moisture characteristic curve; S e denoted as dimensionless available water content; l is the porosity parameter, taken as 0.5; θ(h) is the soil moisture content, in cm³. 3 / cm 3 K(h) is the unsaturated hydraulic conductivity, cm / d.

[0111] Step 5: Prediction of root zone soil moisture content:

[0112] A HYDRUS-1D model for the prediction period of the study area was established. First, the surface soil moisture content on the first day of the prediction period was obtained by inversion. Then, the root soil moisture content was obtained from the root-surface relationship. As the initial condition of the model, the soil moisture content of each layer on each day of the prediction period was simulated. The prediction results were then corrected using the correction relationship to obtain the corrected prediction values.

[0113] A HYDRUS-1D model was established to predict soil moisture content at depths of 0-10, 10-20, 20-30, 30-40, 40-50, and 50-60 cm in 2023. Input soil parameters, precipitation, LAI (Lower Irrigation Area), and ET (Earnings Transmission). o A HYDRUS-1D model was established for the prediction period at observation points in Daxing District to predict daily soil moisture content for 80 days from July 17 to October 4, 2023. The surface soil moisture content at observation points 1-5 on July 17, 2023, was obtained through inversion in step 2. Then, the soil moisture content at depths of 10-20, 20-30, 30-40, 40-50, and 50-60 cm was obtained from the root-surface relationship and used as the initial conditions for the model. Table 5 shows the initial soil moisture content at the five observation points.

[0114] Table 5 Initial soil moisture content at observation points

[0115]

[0116] Based on the above initial conditions, input soil parameters, precipitation, LAI, and ET. o Based on the data, HYDRUS-1D models were established for five monitoring points to obtain the predicted soil moisture content at depths of 0-10, 10-20, 20-30, 30-40, 40-50, and 50-60 cm for the prediction period from July 17 to October 4, 2023. Based on the correction relationships in Table 3, the corrected predicted values ​​were obtained. Taking monitoring points 1 and 2 as examples... Figure 7 and Figure 8The study presents a comparison of predicted soil moisture content at different soil layers for two monitoring points during the prediction period, the results of predicted value correction, and the measured values. After correcting the predicted values ​​using the correction relationships for soil moisture content at each layer, the corrected predicted values ​​are closer to the measured values, and the overlap between the predicted and measured values ​​is significantly improved.

[0117] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention.

Claims

1. A method for predicting root layer soil moisture content using remote sensing-driven hydrodynamic simulation, characterized in that, The method comprises the following steps: Step 1, obtaining basic data of the research area: obtaining basic data of the research area, including satellite remote sensing data, product data, actual observation data; the satellite remote sensing data includes Sentinel-1 satellite image data and Sentinel-2 satellite image data; the product data includes MOD15A2H data and soil attribute data; the soil attribute data includes soil bulk density, sand percentage, silt percentage, clay percentage; the actual observation data includes meteorological data, soil water content data, leaf area index data; meteorological data includes solar radiation R a , precipitation, maximum air temperature T max , minimum air temperature T min ; the soil water content data is the soil water content at different measurement depths; The Hargreaves-Samani model is used to calculate the reference crop evapotranspiration ET o , and the calculation formula is as follows: (2) wherein: R a is solar radiation, MJ / (m 2 ·d); T mean is the mean of the maximum and minimum air temperature, °C; C is the temperature coefficient; a is the temperature constant; m is the temperature exponent; is the sum of the maximum and minimum air temperature; Step 2, inversion of surface soil moisture content: using the water cloud model improved based on the modified chlorophyll absorption ratio index (MCARI), the surface soil moisture content is inverted based on satellite image data; the soil depth of the surface soil is 0-10 cm; The expression for inversion of the surface soil moisture content is: (3) where SSM is the surface soil moisture, cm 3 / cm 3 ; is the Sentinel-1 total backscatter coefficient, db; 0 is the incidence angle of the Sentinel-1 image; VI is MCARI, calculated from Sentinel-2 satellite image data; The calculation method of MCARI is as follows: (4) In the formula: MCARI is the modified chlorophyll absorption ratio index; VNIR is the reflectivity of the near-infrared band; Red is the reflectivity of the red band; Green is the reflectivity of the green band; Step 3, root surface relationship correction: the relationship between the surface soil water storage and the root layer soil water storage known through literature is converted into the relationship between the surface soil moisture content and the root layer soil moisture content, i.e. the root surface relationship, through formula (5), and is corrected; The conversion formula for converting the soil water storage into the soil moisture content is: (5) where: θ n is the soil water content of the nth layer, cm 3 / cm 3 ; ΔZ n is the observed soil layer thickness, cm; D n is the depth from the nth layer to the surface, cm; S Dn is the soil water storage of the soil layer at depth D n , mm; n = 1, 2, 3, 4, 5, 6; The soil moisture content of each layer obtained by formula (5) is compared with the measured value of the soil moisture content of each layer, and the determination coefficient R is calculated 2 , whether the root table relationship meets the requirements is judged, and if not, the root table relationship is established by studying the long time series of measured soil moisture content in the region. Step 4, soil moisture content simulation and correction relationship determination: the soil moisture content is simulated by using the HYDRUS-1D model, and the correction relationship of the soil moisture content of each layer is established; The soil bulk density, soil sand percentage, soil silt percentage, soil clay percentage, precipitation, leaf area index, and reference crop evapotranspiration obtained in step 1 are input, the initial conditions, i.e. the initial soil moisture content and the boundary conditions, are set, the upper boundary is the atmospheric boundary without water accumulation, and the lower boundary is the free drainage boundary, the HYDRUS-1D model is established, the daily soil moisture content of each layer is simulated, and the simulated soil moisture content is compared with the measured soil moisture content, so that the correction relationship of the soil moisture content of each layer is obtained; The soil water movement in the HYDRUS-1D model follows the law of conservation of mass and Darcy's law, and the basic equation for describing the soil water movement in the unsaturated state is obtained by combining the two: (6) where θ is the soil water content, cm 3 / cm 3 ; h is the matric potential, cm; t is time, d -1 ; K is the unsaturated hydraulic conductivity, cm / d; z is the spatial coordinate, cm; S is the crop root water uptake sink term, d -1 ; the K-h and θ-h relationships in the van Genuchten-Mualem model are used to describe the unsaturated soil hydraulic properties and soil water retention curve; The hydraulic conductivity function expression is as follows: (7) (8) (9) (10) where: K s is the saturated hydraulic conductivity, cm / d; θ r is the residual water content of the soil, cm 3 / cm 3 ; θ s is the saturated water content, cm 3 / cm 3 ; α, m, n are the shape parameters of the soil water characteristic curve; S e is the dimensionless effective water content; l is the porosity correlation parameter, taken as 0.5; θ(h) is the soil water content, cm 3 / cm 3 ; K(h) is the unsaturated hydraulic conductivity, cm / d; Step 5, root layer soil moisture content prediction: the HYDRUS-1D model of the prediction period of the research area is established, the surface soil moisture content of the first day of the prediction period is obtained by inversion, and then the root layer soil moisture content is obtained from the root surface relationship and used as the initial condition of the model, the daily soil moisture content of each layer in the prediction period is simulated, and the correction relationship is used to correct the prediction results to obtain the corrected prediction values.

2. The method of claim 1, wherein the method is characterized by: The leaf area index in step 1 is obtained by measuring the leaf length and maximum leaf width of each effective leaf of each plant by a ruler, and the calculation formula is as follows: (1) In the formula: LAI is the leaf area index; p 种 Crop planting density, plants / m 2 ; n is the number of effective leaves for all representative plants; a i m is the leaf length of each effective leaf, m; b i m is the maximum leaf width of each effective leaf, m; c is the number of crop plants determined; leaf area index data is obtained by linear interpolation of the leaf area index per day during the crop growth period.

3. The method of claim 1, wherein, For equation (2) in step 1, according to the recommendation of the Food and Agriculture Organization of the United Nations, when R a is MJ / (m 2 ·d), C, a and m take values of 0.0023, 17.8 and 0.5, respectively.

4. The method of claim 1, wherein, The criterion for judging whether the root table relationship in Step 3 meets the requirements is R 2 ≥ 0.

7.

5. The method of claim 1, wherein, The long time sequence in step 3 comprises at least 100 sample data.

Images