Method for assessing the traceability of soil physical and chemical properties based on digital soil mapping products

Abstract

The application discloses a kind of based on digital soil mapping product's soil physicochemical property traceability ability evaluation method, the preprocessing of soil sample data and digital soil mapping product;Obtain the laboratory determination error of soil physicochemical property and the prediction uncertainty of digital soil mapping product;By superimposing two groups of error range, construct soil possible source place screening standard;Similarity matching algorithm is used to carry out soil traceability simulation experiment;From two dimensions of traceability accuracy and traceability convergence, using point coverage, administrative division coverage and spatial likelihood three standards, the traceability ability of soil physicochemical property is quantified;The priority of soil physicochemical property traceability index is sorted out.This method can solve the problem that there is no complete soil database in the area involved, soil evidence is trace sample, etc., based on digital soil mapping product, the soil physicochemical property index for traceability work is preferentially used, which can provide technical support for efficient and accurate traceability of soil evidence in forensic investigation.

Description

Technical Field

[0001] This invention belongs to the field of forensic soil tracing, specifically involving a method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products. Background Technology

[0002] Soil source tracing involves using expert interpretation of soil physicochemical properties, spectral characteristics, or other soil evidence to infer the soil's formation environment and possible origin. In recent years, soil source tracing has played a significant role in fields such as forensic soil science and forensic geology, providing scientific evidence for the identification of criminal evidence and the investigation of criminal cases.

[0003] In forensic soil science, basic soil physicochemical properties are the earliest, most widespread, and most common indicators for tracing soil evidence. However, soil evidence is usually collected by police, crime scene investigators, and forensic personnel from the surfaces of items used as evidence, such as shoes, clothing, and tools used in the crime. Therefore, soil evidence samples are typically small or even trace. Although numerous soil physicochemical properties can be used for tracing, the limited sample size makes it impossible to measure all indicators, and cost constraints further complicate matters. Therefore, prioritizing the tracing capabilities of individual physicochemical properties is necessary to provide a basis for selecting and measuring which indicators to prioritize in practical tracing applications.

[0004] Soil physicochemical properties have become important traceability indicators not only because of their direct and objective nature, but also because of the support of large-scale soil databases accumulated in the past, such as the World Soil Database covering the global scale and the LUCAS surface soil database covering the regional scale. However, this traditional method has certain limitations. The location involved in the case may not be within the coverage of existing soil databases, and even if it is, the traceability results will be affected to some extent by the accuracy and quantity of the data. Furthermore, the cost of constructing a soil database is commensurate with the size of the survey area and the required sampling density. Therefore, a traceability model that relies entirely on discrete samples may have limited application scope.

[0005] Currently, numerous digital soil mapping products covering different scales and varying levels of accuracy have been developed both domestically and internationally. Their advantages, such as spatial continuity and regional integrity, can effectively compensate for the shortcomings of existing databases in soil source tracing. However, there are few cases where digital soil mapping products have been directly applied to soil source tracing.

[0006] In summary, the current technology has the following shortcomings: (1) Although there are various soil physicochemical property traceability indicators, the ability of different indicators to distinguish traceability varies greatly. There is a lack of systematic priority screening methods. The selection of indicators relies heavily on experience, which can easily lead to information redundancy or missing key features; (2) Existing traceability studies mainly focus on similarity matching between soil samples and existing databases. There are very few studies that apply digital soil mapping products to soil traceability.

[0007] Therefore, there is an urgent need to invent a method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products, in order to solve the problems of strong subjectivity in indicator selection, insufficient utilization of spatial information, and limited traceability efficiency and accuracy in existing technologies. Summary of the Invention

[0008] Purpose of the invention: To address the shortcomings of existing technologies, this invention provides a method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products. This method enables the classification, sorting, and priority screening of soil physicochemical property indicators based on digital soil mapping products, providing technical support for the efficient and accurate traceability of soil evidence in forensic investigations.

[0009] Technical Solution: The present invention provides a method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products, which specifically includes the following steps:

[0010] S1: Obtain measured data of the physicochemical properties of soil samples in the study area and raster data of digital soil mapping products;

[0011] S2: Calculate the measurement range of each physicochemical property of the soil sample based on the laboratory measurement error; calculate the predicted range of each grid cell based on the predicted root mean square error of the digital soil mapping product.

[0012] S3: Establish the overlap matching rules between the measured value range of each physicochemical property of soil samples and the predicted value range of the corresponding raster unit, and select the raster that meets the conditions as the preliminary traceability range.

[0013] S4: Calculate the traceability accuracy of each physicochemical property indicator from two dimensions: location coverage and county-level administrative region coverage;

[0014] S5: Construct a spatial likelihood model to generate high-probability soil source areas and priority search areas, and quantitatively evaluate the convergence of source tracing;

[0015] S6: Based on the comprehensive accuracy and convergence of traceability, physicochemical property indicators are ranked according to their traceability capabilities, and priority traceability indicators are selected.

[0016] Further, step S1 specifically includes:

[0017] Soil bulk density was measured using the ring cutter method, soil pH was measured using the water extraction method, total phosphorus was measured using the molybdenum-antimony colorimetric method, total potassium was measured using the flame photometry method, soil particle size was measured using the laser particle size analyzer method, soil organic matter was measured using the potassium dichromate-sulfuric acid digestion method, total nitrogen was measured using the sulfuric acid digestion-distillation method, and soil cation exchange capacity was measured using the ammonium acetate-EDTA exchange method.

[0018] The digital soil mapping products for the study area were cropped using ArcMap, and raster data of the surface digital soil mapping products with a depth of 0 to 5 cm were extracted.

[0019] Furthermore, step S2 is implemented as follows:

[0020] Based on laboratory measurement errors, the measurement range of each physicochemical property of the soil sample is calculated using the following formulas:

[0021]

[0022] in, The range of values ​​for a certain attribute of an unknown soil sample; This represents the measured value of a certain attribute of an unknown soil sample. Laboratory measurement error for a certain attribute;

[0023] Based on the root mean square error of prediction from digital soil mapping products, the prediction range of each raster cell is calculated using the following formula:

[0024]

[0025]

[0026] in, The root mean square error of a certain attribute; For the first Measured values ​​of soil sample properties; For the first Predicted values ​​of soil sample properties; The predicted value range for each raster cell, i.e., the maximum predicted value of soil attributes. Minimum predicted value of soil properties ; Raster values ​​for digital soil mapping products.

[0027] Furthermore, the overlap matching rule in step S3 is as follows:

[0028]

[0029] in, and These are the first digital soil main map products. Predicted minimum and maximum values ​​for soil attributes in the raster; and The first The upper and lower limits of the attribute value range for each soil sample.

[0030] Furthermore, step S4 is implemented as follows:

[0031] The accuracy of source tracing refers to whether the source tracing range of each physicochemical property of each soil sample covers the actual location of the soil sample point. The higher the proportion of samples whose source tracing range of a certain property includes the actual location of the sample point, the better its source tracing accuracy.

[0032] County-level traceability accuracy refers to whether the traceability range of each physicochemical attribute of each soil sample covers the county where the soil sample point is located and the percentage of samples whose coverage area reaches or exceeds a preset threshold T. The higher the percentage of samples whose traceability range of a certain attribute covers the county where the sample point is located, and the higher the percentage of samples whose traceability range covers the county area where the sample point is located and exceeds the preset threshold T, the better its traceability accuracy.

[0033] Furthermore, the implementation process of step S5 is as follows:

[0034] N raster cells are randomly selected from the digital soil mapping product and subjected to Hermit normal transformation;

[0035] The optimal variogram model is fitted based on the transformed data, and the center point of each grid cell is predicted by Kriging to obtain the mean and standard deviation of the Kriging prediction.

[0036] The Kriging prediction values ​​for the study area are divided into M intervals, and the probability of each grid cell falling into its corresponding interval is calculated using the following formula:

[0037]

[0038] in, For the grid to fall into the first The probability of each interval; and The first The upper and lower limits of each interval; The Kriging forecast mean; For Kriging, predict the standard deviation; The cumulative distribution function of the standard normal distribution;

[0039] The measured values ​​of soil samples are matched to a preset normal distribution interval, and the probability values ​​of all grid cells in the corresponding interval are extracted to generate a spatial likelihood distribution map.

[0040] The eight-neighbor method is used to remove outliers in the spatial likelihood values, and only high-value clustered rasters that meet the following conditions are retained: the likelihood value of the raster itself is greater than or equal to the threshold A, and at least three rasters in its eight-neighbor area have likelihood values ​​greater than or equal to the threshold A; all other rasters are identified as outliers and removed; the value of A is 0.6.

[0041] The region with the highest probability α proportion is designated as the high-probability soil source area. The higher the proportion of samples whose actual locations are covered by the high-probability soil source area, the higher the accuracy of the source tracing for that soil attribute. Furthermore, the smaller the average ratio of the high-probability soil source area to the total area of ​​the study area, the better the convergence of the source tracing for that attribute. The specific formula is as follows:

[0042]

[0043] in, This represents the average ratio of the area of ​​the high-probability soil source region to the total area of ​​the study area. This represents the total amount of soil samples taken. For the first Area of ​​high-probability soil source area corresponding to each soil sampling point; The total area of ​​the study region; α is 30%;

[0044] Regions with spatial likelihood values ​​higher than the sample point's own likelihood value are designated as priority search regions. The smaller the average ratio of the priority search region area to the total study area, the better the source convergence of that attribute. The specific formula is as follows:

[0045]

[0046] in, This represents the average ratio of the area of ​​the priority search zone to the total area of ​​the study area. For the first The area of ​​the priority search zone corresponding to each soil sampling point;

[0047] The convergence of source tracing for different soil physicochemical properties is reflected by the average ratio of the source tracing range of each soil sample point to the total area of ​​the study area. The specific formula is as follows:

[0048]

[0049] in, This represents the average ratio of the source tracing area to the total area of ​​the study region. For the first The area corresponding to each soil sample point for tracing the source.

[0050] Beneficial effects: Compared with the prior art, the beneficial effects of the present invention are: The present invention can address the problems of strong subjectivity in the selection of traceability indicators, insufficient utilization of spatial information, and limited traceability efficiency and accuracy. Based on digital soil mapping products, a similarity matching algorithm is used to rank the traceability capabilities of soil physicochemical property indicators, providing strong technical support for efficient and accurate traceability of soil evidence in forensic investigation. Attached Figure Description

[0051] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0052] The present invention will now be described in further detail with reference to the accompanying drawings.

[0053] like Figure 1 As shown, this invention proposes a method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products. The specific implementation process is as follows:

[0054] Step 1: Obtain measured data of the physicochemical properties of soil samples in the study area and raster data of digital soil mapping products.

[0055] Soil bulk density (BD) was measured using the ring cutter method; soil pH was measured using the water extraction method; total phosphorus (TP) was measured using the molybdenum-antimony colorimetric method; total potassium (TK) was measured using the flame photometry method; soil particle size (Sand, Silt, Clay) was measured using the laser particle size analyzer method; soil organic matter (SOM) was measured using the potassium dichromate-sulfuric acid digestion method; total nitrogen (TN) was measured using the sulfuric acid digestion-distillation method; and cation exchange capacity (CEC) was measured using the ammonium acetate-EDTA exchange method. Digital soil mapping products for the study area were cropped using ArcMap, and predicted topsoil data at depths of 0–5 cm were extracted.

[0056] The 108 soil samples from Anhui Province used in this embodiment are derived from China's terrestrial geography survey and soil carbon pool survey. The digital soil mapping product used is the China High-Resolution National Soil Information Grid Basic Attribute Dataset (1 km). One typical soil physicochemical property index (pH) and one typical soil sampling point (84) are selected from the dataset. The process of this invention is explained regarding data preprocessing, soil potential source area screening rules, point coverage, administrative division coverage, and spatial likelihood.

[0057] Step 2: Calculate the measurement range of each physicochemical property of the soil sample based on the laboratory measurement error; calculate the predicted range of each grid cell based on the root mean square error of the digital soil mapping product.

[0058] Standardized methods were used to measure the measured values ​​of soil physicochemical properties for each soil sample. The possible ranges of these properties were determined based on laboratory measurement errors, as shown in Table 1. The specific formulas are as follows:

[0059]

[0060] in, Let be the range of values ​​for a certain attribute of an unknown soil sample, i.e. ( , ); This represents the measured value of a certain attribute of an unknown soil sample. The laboratory measurement error for a certain attribute.

[0061] In this embodiment, the measured values ​​and ranges of the 10 soil physicochemical properties of soil sample point 84 are shown in Table 1.

[0062] Table 1. Measured values ​​and ranges of soil physicochemical properties at sample point 84

[0063]

[0064] Spatial clipping was performed using ArcMap to obtain digital soil mapping products for Anhui Province. The root mean square error (RMSE) of the predictions for each attribute was calculated using these digital soil mapping products, as shown in Table 2. The RMSE was added to and subtracted from each raster cell in the digital soil mapping products for each attribute to obtain the prediction range. The specific formula is as follows:

[0065]

[0066]

[0067] in, The root mean square error of a certain attribute; For the first Measured values ​​of soil sample properties; For the first Predicted values ​​of soil sample properties; The predicted value range for each raster cell, i.e., the maximum predicted value of soil attributes. Minimum predicted value of soil properties ; Raster values ​​for digital soil mapping products.

[0068] Table 2. Root mean square error of prediction for various physicochemical properties of digital soil mapping products

[0069]

[0070] After combining the root mean square error correction with each predicted grid cell for soil property pH, the upper and lower limit distribution results of the predicted values ​​are obtained.

[0071] Step 3: Establish the overlap matching rules between the measured value range of each physicochemical property of the soil sample and the predicted value range of the corresponding raster unit, and select the raster that meets the conditions as the preliminary traceability range.

[0072] By superimposing two types of numerical ranges—laboratory measurement error and digital soil mapping prediction error—the retrieval and screening conditions for the soil information grid are defined, generating the source tracing range corresponding to each soil physicochemical property. The screening criterion formula is as follows:

[0073]

[0074] in, and These are the first digital soil main map products. Predicted minimum and maximum values ​​for soil attributes in the raster; and The first The upper and lower limits of the value range of the corresponding attributes for each soil sample.

[0075] In this embodiment, the measured pH range of soil sample 84 is (8.2, 8.4). If the predicted pH range of a certain grid cell is (7.59, 9.01), then according to the screening criterion formula, this grid cell is a potential source area of ​​the soil sample.

[0076] Step 4: Calculate the traceability accuracy of each physicochemical property indicator from two dimensions: location coverage and county-level administrative region coverage.

[0077] Accuracy of source tracing: This involves statistically analyzing whether the source tracing range of each soil physicochemical property covers the actual location of the soil sample point, and the percentage of samples that successfully cover the actual location. In this embodiment, the source tracing range of each soil physicochemical property successfully covered the actual location of sample 84; the percentage of soil property pH that successfully covered the actual location of the sample point was 69.44%.

[0078] County-level traceability accuracy: This involves statistically analyzing whether the traceability scope of each physicochemical property covers the county where the soil sample point is located, and the percentage of samples with a coverage area of ​​over 80%. In this embodiment, the traceability scope of each physicochemical property covers the county where the soil sample point is located. Except for pH, the traceability scope of all other properties covers more than 80% of the area of ​​the county where the sample point is located. For the soil property pH, the percentage of samples covering more than 80% of the area of ​​the county where the soil sample point is located is 32.41%.

[0079] Step 5: Construct a spatial likelihood model to generate high-probability soil source areas and priority search areas, and quantitatively evaluate the convergence of source tracing.

[0080] 5000 raster cells were randomly selected from the digital soil mapping of various physical and chemical properties, and their predicted values ​​were subjected to Hermitian normal transformation.

[0081] The optimal variogram model is fitted based on the transformed data, and Kriging prediction is performed on the center point of each grid cell within the source tracing range to obtain the Kriging prediction mean and standard deviation. In this embodiment, the soil property pH is fitted to obtain a nested shared variogram model, which consists of the nugget effect and the spherical model.

[0082] The Kriging predictions (standardized values) for the study area were divided into 20 intervals, and the probability of each grid cell falling into its corresponding interval was calculated using the following formula:

[0083]

[0084] in, For the grid to fall into the first The probability of each interval; and The first The upper and lower limits of each interval; The Kriging forecast mean; For Kriging, predict the standard deviation; This is the cumulative distribution function of the standard normal distribution.

[0085] In this embodiment, the Kriging predicted values ​​for soil pH in the 20 ranges are (-∞, -1.6429), (-1.6429, -1.2827), (-1.2827, -1.0314), (-1.0314, -0.8427), (-0.8427, -0.6701), (-0.6701, -0.5195), (-0.5195, -0.3878), (-0.3878, -0.2443), (-0.2443, -0.1292), (-0.1292, 0.0048), (0.0048, 0.1272), (0.1272, 0.2515), (0.2515, 0.3842), (0.3842, 0.5224), (0.5224, 0.6726), (0.6726, 0.8384), (0.8384, 1.0334), (1.0334, 1.2867), (1.2867, 1.6497), (1.6497, +∞).

[0086] The measured values ​​of soil samples are matched to a preset normal distribution interval, and the probability values ​​of all grid cells within the corresponding interval are extracted to generate a spatial likelihood distribution map. In this embodiment, the measured value of pH of soil property at sample point 84 is 8.3, and the transformed value after normalization is 3.7191, corresponding to interval 20. The probability of all grid cells in interval 20 is extracted to generate a spatial likelihood distribution map of pH of soil property at sample point 84.

[0087] High-value clustered rasters that meet the following conditions are retained: the raster itself has a likelihood value greater than or equal to 0.6, and at least 3 of its eight neighboring rasters have a likelihood value greater than or equal to 0.6; all other rasters are identified as extreme values ​​and removed.

[0088] The top 30% of areas by probability were identified as high-probability soil source areas, and the pH value of sample 84 was also identified as a high-probability soil source area. The percentage of samples where high-probability soil source areas covered the actual locations of soil samples was calculated, along with the average ratio of the area of ​​high-probability soil source areas to the total area of ​​the study area. The specific formulas are as follows:

[0089]

[0090] in, This represents the average ratio of the area of ​​the high-probability soil source region to the total area of ​​the study area. This represents the total amount of soil samples taken. For the first Area of ​​high-probability soil source area corresponding to each soil sampling point; This represents the total area of ​​the study region.

[0091] In this embodiment, the proportion of soil samples whose actual location was covered by the high-probability soil source area (pH) was 16.67%, and the average ratio of the area of ​​the high-probability soil source area to the total area of ​​the study area was 0.119%.

[0092] Regions with spatial likelihood values ​​higher than the sample point's own likelihood value are designated as priority search regions. The average ratio of the priority search region area to the total study area is calculated using the following formula:

[0093]

[0094] in, This represents the average ratio of the area of ​​the priority search zone to the total area of ​​the study area. This represents the total amount of soil samples taken. For the first The area of ​​the priority search zone corresponding to each soil sampling point; This represents the total area of ​​the study region.

[0095] In this embodiment, the average ratio of the priority search area for soil property pH to the total area of ​​the study area was 0.108%.

[0096] The average ratio of the traceability range of each physicochemical property to the total area of ​​the study area is calculated using the following formula:

[0097]

[0098] in, This represents the average ratio of the source tracing area to the total area of ​​the study region. This represents the total amount of soil samples taken. For the first The area of ​​the source tracing range corresponding to each soil sampling point; This represents the total area of ​​the study region.

[0099] In this embodiment, the average ratio of the soil pH tracing range to the total area of ​​the study area is 0.409%.

[0100] Based on the comprehensive accuracy and convergence of source tracing, the source tracing capabilities of 10 physicochemical properties were ranked, and priority source tracing indicators were selected. In this embodiment, total phosphorus and cation exchange capacity were selected as the priority core indicators for soil source tracing using digital soil mapping products in the study area.

[0101] The above content merely illustrates the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made based on this technical solution fall within the scope of protection of the technical concept proposed in this invention and are included in the claims of this invention.

Claims

1. A method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products, characterized in that, Includes the following steps: S1: Obtain measured data of the physicochemical properties of soil samples in the study area and raster data of digital soil mapping products; S2: Calculate the measurement range of each physicochemical property of the soil sample based on the laboratory measurement error; calculate the predicted range of each grid cell based on the predicted root mean square error of the digital soil mapping product. S3: Establish the overlap matching rules between the measured value range of each physicochemical property of soil samples and the predicted value range of the corresponding raster unit, and select the raster that meets the conditions as the preliminary traceability range. S4: Calculate the traceability accuracy of each physicochemical property indicator from two dimensions: location coverage and county-level administrative region coverage; S5: Construct a spatial likelihood model to generate high-probability soil source areas and priority search areas, and quantitatively evaluate the convergence of source tracing; S6: Based on the comprehensive accuracy and convergence of traceability, rank the physicochemical property indicators according to their traceability capabilities and select priority traceability indicators; The implementation process of step S5 is as follows: N raster cells are randomly selected from the digital soil mapping product and subjected to Hermit normal transformation; The optimal variogram model is fitted based on the transformed data, and the center point of each grid cell is predicted by Kriging to obtain the mean and standard deviation of the Kriging prediction. The Kriging prediction values ​​for the study area are divided into M intervals, and the probability of each grid cell falling into its corresponding interval is calculated using the following formula: ； in, For the grid to fall into the first The probability of each interval; and The first The upper and lower limits of each interval; The Kriging forecast mean; For Kriging, predict the standard deviation; The cumulative distribution function of the standard normal distribution; The measured values ​​of soil samples are matched to a preset normal distribution interval, and the probability values ​​of all grid cells in the corresponding interval are extracted to generate a spatial likelihood distribution map. The eight-neighbor method is used to remove outlier extreme values ​​in the spatial likelihood values, and only high-value clustered rasters that meet the following conditions are retained: the likelihood value of the raster itself is greater than or equal to the threshold A, and at least three rasters in its eight-neighborhood have likelihood values ​​greater than or equal to the threshold A; all other rasters are identified as extreme values ​​and removed. The region with the highest probability α proportion is designated as the high-probability soil source area. The higher the proportion of samples whose actual locations are covered by the high-probability soil source area, the higher the accuracy of the source tracing for that soil attribute. Furthermore, the smaller the average ratio of the high-probability soil source area to the total area of ​​the study area, the better the convergence of the source tracing for that attribute. The specific formula is as follows: ； in, This represents the average ratio of the area of ​​the high-probability soil source region to the total area of ​​the study area. This represents the total amount of soil samples taken. For the first Area of ​​high-probability soil source area corresponding to each soil sampling point; The total area of ​​the study area; Regions with spatial likelihood values ​​higher than the sample point's own likelihood value are designated as priority search regions. The smaller the average ratio of the priority search region area to the total study area, the better the source convergence of that attribute. The specific formula is as follows: ； in, This represents the average ratio of the area of ​​the priority search zone to the total area of ​​the study area. For the first The area of ​​the priority search zone corresponding to each soil sampling point; The convergence of source tracing for different soil physicochemical properties is reflected by the average ratio of the source tracing range of each soil sample point to the total area of ​​the study area. The specific formula is as follows: ； in, This represents the average ratio of the source tracing area to the total area of ​​the study region. For the first The area corresponding to each soil sample point for tracing the source.

2. The method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products according to claim 1, characterized in that, Step S1 specifically includes: Soil bulk density was measured using the ring cutter method, soil pH was measured using the water extraction method, total phosphorus was measured using the molybdenum-antimony colorimetric method, total potassium was measured using the flame photometry method, soil particle size was measured using the laser particle size analyzer method, soil organic matter was measured using the potassium dichromate-sulfuric acid digestion method, total nitrogen was measured using the sulfuric acid digestion-distillation method, and soil cation exchange capacity was measured using the ammonium acetate-EDTA exchange method. The digital soil mapping products for the study area were cropped using ArcMap, and raster data of the surface digital soil mapping products with a depth of 0 to 5 cm were extracted.

3. The method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products according to claim 1, characterized in that, The implementation process of step S2 is as follows: Based on laboratory measurement errors, the measurement range of each physicochemical property of the soil sample is calculated using the following formulas: ； in, The range of values ​​for a certain attribute of an unknown soil sample; This represents the measured value of a certain attribute of an unknown soil sample. Laboratory measurement error for a certain attribute; Based on the root mean square error of prediction from digital soil mapping products, the prediction range of each raster cell is calculated using the following formula: ； ； in, The root mean square error of a certain attribute; For the first Measured values ​​of soil sample properties; For the first Predicted values ​​of soil sample properties; The predicted value range for each raster cell, i.e., the maximum predicted value of soil attributes. Minimum predicted value of soil properties ; Raster values ​​for digital soil mapping products.

4. The method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products according to claim 1, characterized in that, The overlap matching rule in step S3 is as follows: ； in, and These are the first digital soil main map products. Predicted minimum and maximum values ​​for soil attributes in the raster; and The first The upper and lower limits of the attribute value range for each soil sample.

5. The method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products according to claim 1, characterized in that, The implementation process of step S4 is as follows: The accuracy of source tracing refers to whether the source tracing range of each physicochemical property of each soil sample covers the actual location of the soil sample point. The higher the proportion of samples whose source tracing range of a certain property includes the actual location of the sample point, the better its source tracing accuracy. County-level traceability accuracy refers to whether the traceability range of each physicochemical attribute of each soil sample covers the county where the soil sample point is located and the percentage of samples whose coverage area reaches or exceeds a preset threshold T. The higher the percentage of samples whose traceability range of a certain attribute covers the county where the sample point is located, and the higher the percentage of samples whose traceability range covers the county area where the sample point is located and exceeds the preset threshold T, the better its traceability accuracy.

6. The method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products according to claim 1, characterized in that, The threshold A is set to 0.

6.

7. The method for assessing the traceability capability of soil physicochemical properties based on digital soil mapping products according to claim 1, characterized in that, The value of α is 30%.

Images