A black soil cold disaster monitoring method based on multi-source data fusion

Abstract

This invention discloses a method for monitoring cold disasters in black soil based on multi-source data fusion, belonging to the field of agricultural information technology. This invention calculates the depth difference between freeze-thaw fronts and generates a cumulative freeze-thaw damage amount based on its cumulative value within a preset time window, thus achieving a quantitative characterization of the dynamic characteristics of the freeze-thaw process. Simultaneously, by calculating the topsoil structure compaction coefficient and determining a freeze-thaw correction coefficient based on the cumulative freeze-thaw damage amount, the cumulative freeze-thaw damage effect is effectively quantified. This invention weightedly fuses the effective active accumulated temperature deficit value with the corrected compaction coefficient to generate a black soil cold disaster risk index, and outputs graded early warning information by matching early warning rules with crop phenological stages. This overcomes the limitations of single meteorological parameters or static soil indicators, fully considers the progressive damage to soil structure during the freeze-thaw cycle and its amplification effect on cold disaster sensitivity, significantly improving the accuracy and timeliness of cold disaster risk assessment.

Description

Technical Field

[0001] This invention relates to the field of agricultural information technology, specifically to a method for monitoring cold disasters in black soil based on multi-source data fusion. Background Technology

[0002] The black soil region, located in a seasonally frozen soil area, exhibits distinct continental monsoon climate characteristics, characterized by long, harsh winters and drastic temperature fluctuations in spring. This leads to frequent cold-related disasters in early winter and early spring. These disasters not only directly affect the normal growth and development of crops but also alter the soil's physical structure through freeze-thaw cycles, causing changes in topsoil structure and the destruction of soil aggregates. Consequently, they exacerbate soil erosion and water loss, seriously threatening the sustainable use of black soil and the safety of agricultural production. Therefore, establishing scientific and precise methods for monitoring cold-related disasters in black soil is of significant practical importance and application value for timely early warning of cold-related disaster risks, guiding agricultural production decisions, and protecting black soil resources.

[0003] Currently, monitoring methods for chilling injury in black soil mainly include the following categories: First, threshold-based early warning methods based on single meteorological elements, which determine whether chilling injury has occurred by setting fixed temperature thresholds; second, methods that utilize satellite thermal infrared remote sensing to retrieve surface temperature or soil moisture, combined with crop models to assess chilling injury risk; and third, methods that deploy soil temperature and humidity sensors in the field to monitor changes in shallow soil temperature in real time, issuing alarms when the temperature falls below the crop's tolerance threshold. These methods have, to a certain extent, achieved early identification of chilling injury, providing reference information for agricultural production.

[0004] The main shortcomings of existing technologies are: they cannot establish a quantitative correlation between the dynamic physical characteristics of the freeze-thaw process and the degree of soil structural damage, making it difficult to quantify the cumulative damage effect of the freeze-thaw process on black soil. Specifically, traditional freeze-thaw monitoring methods can only obtain instantaneous state information of soil temperature or freeze-thaw interface, failing to reflect the dynamic evolution of freeze-thaw fronts over time and their cumulative destructive effect on soil structure. Furthermore, existing cold disaster early warning technologies are mostly based on single meteorological parameters or static soil physical indicators, failing to fully consider the gradual damage to soil structure during the freeze-thaw cycle and its amplification effect on cold disaster sensitivity. This results in insufficient accuracy and timeliness of cold disaster risk assessment, making it difficult to meet the actual needs of precise protection of black soil and agricultural production. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for monitoring cold disasters in black soil based on multi-source data fusion.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows: a method for monitoring cold disasters in black soil based on multi-source data fusion, comprising:

[0007] Acquire air temperature data, layered soil temperature data, and radar / satellite data for the monitored area;

[0008] The air temperature data is accumulated and processed to calculate the effective accumulated temperature deficit; the layered soil temperature data is processed for frontal identification and the freeze-thaw front depth difference is calculated; the radar satellite data is processed for polarization decomposition and the topsoil structure compaction coefficient is calculated.

[0009] Based on the cumulative value of the freeze-thaw front depth difference within a preset time window, a cumulative freeze-thaw damage amount is generated. Based on the cumulative freeze-thaw damage amount, a freeze-thaw correction coefficient is determined according to a preset monotonically increasing mapping relationship. The topsoil structure compactness coefficient is multiplied by the freeze-thaw correction coefficient to obtain the corrected compactness coefficient.

[0010] The effective active accumulated temperature deficit value and the modified compaction coefficient are weighted and fused together to generate the black soil cold disaster risk index.

[0011] The system acquires the crop phenological stages of the monitored area, matches preset early warning rules with the black soil cold disaster risk index and the crop phenological stages, and outputs graded early warning information.

[0012] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0013] 1. This invention establishes a quantitative correlation between the dynamic physical characteristics of the freeze-thaw process and the degree of soil structural damage. By acquiring stratified soil temperature data and performing frontal identification processing, the depth difference between freeze-thaw fronts is calculated. Then, based on the cumulative value of the freeze-thaw front depth difference within a preset time window, the cumulative amount of freeze-thaw damage is generated, achieving a quantitative characterization of the dynamic physical characteristics of the freeze-thaw process. Simultaneously, through polarization decomposition processing of radar satellite data, the topsoil compaction coefficient is calculated, achieving a quantitative assessment of the degree of soil structural damage. More importantly, this invention establishes a quantitative correlation between the cumulative amount of freeze-thaw damage and the modified compaction coefficient through a preset monotonically increasing mapping relationship, thereby organically combining the dynamic physical characteristics of the freeze-thaw process with the degree of soil structural damage, effectively solving the technical problem that existing technologies cannot establish a quantitative correlation between the dynamic physical characteristics of the freeze-thaw process and the degree of soil structural damage.

[0014] 2. This invention quantifies the cumulative damage effect of freeze-thaw processes on black soil. It innovatively introduces the concept of cumulative freeze-thaw damage, calculating the cumulative value of the freeze-thaw front depth difference within a preset time window to achieve a quantitative assessment of the cumulative damage effect of the freeze-thaw process. This quantification of the cumulative damage effect not only considers the instantaneous state information of the freeze-thaw front but also focuses on the dynamic evolution of the freeze-thaw front over time, accurately reflecting the gradual damage to soil structure during the freeze-thaw cycle and its amplified effect on sensitivity to cold disasters. By multiplying the topsoil compaction coefficient by a freeze-thaw correction coefficient, a corrected compaction coefficient is obtained, further strengthening the impact of cumulative freeze-thaw damage on soil structure and effectively solving the technical problem that existing technologies struggle to quantify the cumulative damage effect of freeze-thaw processes on black soil.

[0015] 3. This invention improves the accuracy and timeliness of cold disaster risk assessment. It employs a multi-source data fusion method, organically combining air temperature data, stratified soil temperature data, and radar / satellite data. Through weighted fusion calculation using effective accumulated temperature deficit and a corrected compaction coefficient, a black soil cold disaster risk index is generated, enabling a comprehensive assessment of cold disaster risk. This multi-source data fusion method not only overcomes the limitations of single meteorological parameters or static soil physical indicators but also fully considers the progressive damage to soil structure during freeze-thaw cycles and its amplifying effect on cold disaster sensitivity, making the cold disaster risk assessment more comprehensive and accurate. Simultaneously, by acquiring the crop phenological stages of the monitoring area and matching the black soil cold disaster risk index with preset early warning rules, this invention outputs tiered early warning information, achieving dynamic assessment and tiered early warning of cold disaster risk for crops at different growth stages. This significantly improves the timeliness and practicality of cold disaster risk assessment, effectively solving the technical problems of insufficient accuracy and timeliness in existing cold disaster risk assessment technologies.

[0016] 4. This invention achieves precision and intelligence in monitoring cold damage in black soil. By constructing a dynamic early warning mechanism based on crop phenological stages, it overcomes the limitations of traditional fixed-threshold early warning, enabling accurate identification and tiered early warning of cold damage risks for crops at different growth stages. Simultaneously, this invention supports agriculturally suitable terminal interaction, achieving local caching and offline computation of multi-source data through an IoT gateway, ensuring the continuous operation of the monitoring system even during network interruptions. Furthermore, the zero-code voice interaction function lowers the barrier to entry for farmers, improving the system's ease of use and widespread adoption. These technological innovations not only improve the precision of black soil cold damage monitoring but also realize the intelligence and humanization of the monitoring process, providing strong technical support for precision protection of black soil and agricultural production. Attached Figure Description

[0017] The disclosure of this invention is illustrated with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. In the drawings, the same reference numerals are used to refer to the same parts. Wherein:

[0018] Figure 1 This is a flowchart of the method of the present invention;

[0019] Figure 2 This is a data processing flowchart of the present invention. Detailed Implementation

[0020] Those skilled in the art will understand that, based on the technical solutions of this invention, without altering the essential spirit of the invention, various interchangeable structural methods and implementations can be proposed. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solutions of this invention and should not be considered as the entirety of the invention or as limitations or restrictions on the technical solutions of this invention.

[0021] like Figure 1 and Figure 2 As shown, the specific implementation steps of the present invention include the following steps:

[0022] Acquire air temperature data, stratified soil temperature data, and radar satellite data for the monitored area.

[0023] Specifically, the air temperature data is collected by automatic weather stations deployed within the monitoring area. The sampling frequency can be set to once per hour, the measurement accuracy is ±0.1℃, and the spatial coverage density is no less than one station per 100 square kilometers to ensure the representativeness of the regional temperature field, which is used for subsequent calculation of the effective active accumulated temperature deficit.

[0024] The layered soil temperature data is acquired by soil temperature sensors buried at different depths. The sensor burial depth is strictly designed according to the typical profile structure of black soil, including at least shallow, middle and deep layers. The shallow layer corresponds to the black soil cultivated layer, with a depth range of 0 cm to 20 cm. The middle layer corresponds to the plow pan, with a depth range of 20 cm to 40 cm. The deep layer corresponds to the upper part of the subsoil, with a depth range of 40 cm to 60 cm. Each sensor outputs a temperature value independently. The sampling frequency and measurement accuracy are synchronized with the air temperature data, which is used to identify the location of freeze-thaw fronts and calculate the depth difference between freeze-thaw fronts.

[0025] The radar satellite data is spaceborne synthetic aperture radar imagery with a spatial resolution of 20 meters × 20 meters and a revisit period of 6 to 12 days. It is used for subsequent polarization decomposition processing to retrieve the surface soil structure compaction coefficient.

[0026] The selection of these data sources takes into account the specific needs of monitoring cold disasters in black soil regions. Air temperature reflects atmospheric thermal conditions, soil temperature reveals the dynamics of freeze-thaw processes, and radar data is sensitive to changes in surface structure; these three sources complement each other. To ensure temporal consistency of the multi-source data, timestamps of all data sources are aligned, Beijing time is uniformly adopted, and missing values ​​are filled using linear interpolation. For example, if air temperature data for a certain hour is missing, the average of the temperatures of the previous and following hours is taken as the estimated value for that moment.

[0027] This step serves as the data entry point for the entire black soil cold disaster monitoring method. Its core function is to provide the raw data foundation for calculating all subsequent features, including the effective accumulated temperature deficit, the freeze-thaw front depth difference, and the topsoil structure compaction coefficient. Specifically, air temperature data supports the quantification of heat deficit, stratified soil temperature data supports the identification of freeze-thaw misalignment intensity, and radar and satellite data supports the inversion of soil structure status. Without the data acquisition in this step, no subsequent processing can be performed; this step is a prerequisite and logical starting point for the implementation of the method.

[0028] By combining data from three different sources—air temperature, stratified ground temperature, and radar satellites—an integrated sky-ground monitoring capability is formed, compensating for the shortcomings of single data sources in terms of temporal resolution or spatial coverage. The multi-depth deployment of stratified soil temperature sensors can capture the vertical position of freeze-thaw fronts, providing crucial information that traditional single-point ground temperature monitoring cannot obtain for subsequent freeze-thaw misalignment analysis. The introduction of radar satellite data makes it possible to retrieve topsoil structure over a large area without contact, avoiding frequent manual field surveys. Unified timestamp alignment and missing value imputation processing ensure strict synchronization of multi-source data in the time dimension, eliminating calculation errors introduced by asynchronous data.

[0029] During implementation, air temperature data can also be derived from the interpolation results of regional meteorological station grids instead of plot-level meteorological stations, or surface temperature retrieved from thermal infrared remote sensing can be used as a substitute; layered soil temperature data can also be retrieved by ground-penetrating radar or resistivity tomography equipment to retrieve the temperature distribution at different depths, replacing physically buried sensors, or a soil thermal conduction model can be used in combination with surface temperature and meteorological driving data to simulate the ground temperature profile; radar and satellite data can also be used for local high-precision monitoring using UAV-borne synthetic aperture radar or hyperspectral imagery; data synchronization methods can also be adopted by asynchronous caching followed by unified resampling to the same time grid, or by using Kalman filtering for multi-source data fusion and alignment, rather than simple linear interpolation.

[0030] The air temperature data is accumulated and processed to calculate the effective accumulated temperature deficit; the layered soil temperature data is processed for frontal identification and the freeze-thaw front depth difference is calculated; the radar satellite data is processed for polarization decomposition and the topsoil structure compaction coefficient is calculated.

[0031] This step is the core feature extraction stage of the entire black soil cold disaster monitoring method. It uses targeted data processing algorithms to transform the raw observation data into cold disaster monitoring indicators with clear physical meaning.

[0032] The calculation of the effective accumulated temperature deficit is based on the theory of accumulated temperature in agricultural meteorology, and only the daily average temperature that is greater than 0°C is accumulated. This is because 0°C is the biological zero point for the growth of most crops. Below this temperature, crops basically stop growing. This indicator can quantify the degree of shortage of heat resources in the current growing season relative to the historical average level.

[0033] The calculation of the depth difference of the freeze-thaw front is based on the principle of soil freeze-thaw physics. The position of the freeze-thaw front is located by identifying the interface between positive and negative temperatures. The vertical movement of this position directly reflects the intensity of the soil freeze-thaw process and the potential damage to the soil structure.

[0034] The calculation of the topsoil structure compaction coefficient utilizes radar polarization scattering characteristics. Different soil structures exhibit significant differences in their electromagnetic wave scattering mechanisms. By polarization decomposition, surface scattering and secondary scattering components can be separated, thereby inverting the soil structure state.

[0035] These three indicators characterize the risk of cold disasters in black soil from three dimensions: thermal conditions, freeze-thaw dynamics, and soil physical structure, respectively, forming a multi-dimensional monitoring system and providing basic input parameters for subsequent risk index fusion calculation.

[0036] The calculation of the effective active accumulated temperature deficit is as follows:

[0037] Determine the start date of spring for the current year. The start date of spring is the first day when the average daily temperature is greater than or equal to a preset temperature threshold for a consecutive preset number of days in spring. Starting from the start date of spring for the current year, accumulate the accumulated temperature above zero degrees Celsius day by day to obtain the current actual accumulated temperature.

[0038] For each historical year within a preset number of years, determine the start date of spring for that historical year, and obtain the cumulative accumulated temperature of activities greater than zero degrees from the start date of spring of that historical year to the same month and day as the current date. Calculate the average of the cumulative accumulated temperature of activities greater than zero degrees obtained for all historical years as the baseline accumulated temperature for the same historical period.

[0039] Calculate the difference between the historical baseline accumulated temperature and the current actual accumulated temperature; if the difference is greater than 0, then the difference is taken as the effective active accumulated temperature deficit; if the difference is less than or equal to 0, then the effective active accumulated temperature deficit is determined to be 0.

[0040] In this calculation process, the preset number of days ranges from 3 to 7 days. This range fully considers the characteristics of spring temperature fluctuations in Northeast China. Shorter days can improve the sensitivity to temperature rises, while longer days can enhance the stability of the judgment and avoid misjudgments caused by short-term temperature fluctuations. The preset temperature threshold ranges from 3 to 8℃. This range covers the temperature requirements of major crops in the black soil region from germination to seedling emergence. Lower thresholds are suitable for cold-resistant crops or early-sown varieties, while higher thresholds are suitable for warm-loving crops or conventional sowing times. The preset number of years ranges from 20 to 50 years. This range can fully reflect the long-term characteristics of regional climate while taking into account the availability and timeliness of data. Longer years can smooth interannual fluctuations, while shorter years can better reflect recent climate change trends.

[0041] In practice, the determination of the start date of spring needs to be dynamically adjusted in combination with local meteorological data and crop growth patterns. The calculation of the historical benchmark accumulated temperature needs to take into account the natural variation of the start time of spring in different years. The continuity and accuracy of the accumulated temperature calculation are ensured by accumulating it day by day. The non-negative constraint is set after the difference calculation to ensure the physical meaning of the index and avoid logical confusion caused by negative values. This design enables the effective active accumulated temperature deficit to accurately reflect the degree of abnormality of the current growing season heat conditions.

[0042] The innovation of this method lies in introducing historical accumulated temperature as a dynamic reference system, avoiding the limitations of using a fixed accumulated temperature threshold. It can adapt to the natural variations in the start time of spring in different years. Furthermore, by setting a non-negative deficit constraint, it ensures the physical rationality and computational stability of the indicator, avoiding logical confusion caused by negative values ​​when the current accumulated temperature exceeds the historical average. This dynamic benchmark design enables the indicator to accurately reflect the inhibitory effect of abnormal low-temperature events on crop growth, providing a reliable thermal dimension input for subsequent risk assessment.

[0043] The calculation of the depth difference between the freeze-thaw fronts is as follows:

[0044] Acquire temperature values ​​collected by soil temperature sensors at at least three different depth layers, including shallow, middle and deep layers, which correspond to preset depth ranges of the black soil tillage layer, plow layer and upper subsoil layer, respectively.

[0045] Determine whether there is a first-type depth layer with a temperature greater than zero degrees Celsius and a second-type depth layer with a temperature less than zero degrees Celsius among all depth layers. If there is a layer with a temperature equal to 0 degrees Celsius, it can be classified into either side.

[0046] If both the first type of depth layer and the second type of depth layer exist simultaneously, the position with the greatest depth in the first type of depth layer is identified as the first front, and the position with the least depth in the second type of depth layer is identified as the second front. The vertical distance between the first front and the second front is calculated, and this vertical distance is taken as the freeze-thaw front depth difference.

[0047] If the first type of depth layer and the second type of depth layer do not exist simultaneously, the depth difference of the freeze-thaw front is determined to be 0.

[0048] In this calculation process, the preset depth range fully considers the typical structural characteristics and regional variations of the black soil profile. The shallow depth range is 0 to 20 cm, which covers the typical depth distribution of the cultivated layer and can effectively monitor the temperature changes of the main active layer of crop roots. The middle depth range is 20 to 40 cm, which corresponds to the typical depth of the plow pan and is a key transition layer in the soil freeze-thaw process. The deep depth range is 40 to 60 cm, which covers the main depth range of the upper subsoil and can reflect the freeze-thaw state of the deep soil.

[0049] During front identification, the determination of the first and second depth layers needs to consider the accuracy and stability of temperature measurement to avoid misjudgment due to measurement errors. The location of the first and second fronts adopts the extreme value principle, that is, the deepest or shallowest position is selected among the depth layers that meet the temperature conditions. This design can accurately capture the actual position of the freeze-thaw front. The vertical distance is calculated using the absolute difference, which reflects the vertical span of the freeze-thaw front. When the temperature of all depth layers is the same, it indicates that the soil is in a state of complete freezing or complete thawing. At this time, the depth difference of the freeze-thaw front is 0. This design conforms to the physical laws of the soil freeze-thaw process.

[0050] The innovation of this method lies in reconstructing the continuous freeze-thaw interface through discrete point temperature data, overcoming the limitation of traditional single-point freezing depth measurement that cannot reflect the dynamic changes of the freeze-thaw front. At the same time, by setting the depth difference to 0, it avoids generating meaningless calculation results in the state of complete freezing or complete thawing. This front identification algorithm can sensitively capture the vertical dynamics of the soil freeze-thaw process, providing a quantitative basis for assessing the degree of damage to soil structure caused by freeze-thaw. It is a key input parameter for subsequent calculation of cumulative freeze-thaw damage. Its design fully considers the physical characteristics and monitoring needs of the freeze-thaw process in black soil.

[0051] The calculation of the topsoil structure compaction coefficient is as follows:

[0052] Extract polarimetric synthetic aperture radar data from the radar satellite data;

[0053] Polarization decomposition parameters are extracted from the polarization synthetic aperture radar data, and the polarization decomposition parameters include at least the surface scattering ratio and the secondary scattering ratio;

[0054] The surface roughness index is calculated based on the ratio of the surface scattering ratio to the secondary scattering ratio.

[0055] Based on the surface roughness index, the topsoil structure compaction coefficient is determined according to a preset monotonically increasing mapping relationship.

[0056] In this calculation process, fully polarimetric synthetic aperture radar (SAR) data is extracted from the acquired radar satellite data. To ensure the accuracy of subsequent analysis, the SAR data must undergo rigorous preprocessing, including radiometric calibration to eliminate sensor system errors, geometric correction to accurately locate ground features, and filtering to effectively suppress the inherent speckle noise of the radar image, thereby obtaining a high-quality dataset that can be directly used for analysis.

[0057] Based on this, the preprocessed polarization data undergoes polarization decomposition. The core of this process lies in using a mature physical decomposition model to decompose the complex radar echo signal into several fundamental scattering mechanisms with clear physical meaning. Among them, this invention focuses on two scattering components directly related to the physical properties of the soil surface: surface scattering and secondary scattering, and calculates their respective proportions in the total scattering energy, namely the surface scattering proportion and the secondary scattering proportion.

[0058] Then, based on the surface scattering ratio and the secondary scattering ratio, a surface roughness index is constructed to reflect the micro-topographic undulation of the land surface. Specifically, the surface roughness index is calculated by the ratio of the surface scattering ratio to the secondary scattering ratio. This ratio effectively eliminates changes in overall signal strength caused by factors such as absolute calibration errors of the radar system or atmospheric attenuation, making the final index sensitive only to relative changes in surface scattering characteristics, thus significantly improving the stability and robustness of the index. When the soil surface is smoother and more compact, its surface scattering ratio is higher, while its secondary scattering ratio is relatively lower. Therefore, this ratio can effectively indicate the compaction state of the soil.

[0059] Finally, the calculated surface roughness index is substituted into a pre-defined, monotonically increasing mapping relationship to ultimately determine the topsoil structure compaction coefficient. This monotonically increasing mapping relationship is used to establish a quantitative bridge between the surface roughness index obtained from remote sensing inversion and the topsoil structure compaction, which has clear significance in soil physics.

[0060] The monotonically increasing mapping relationship is established based on mutual verification of a large amount of field observation data and remote sensing inversion results. Specifically, researchers conducted field sampling in typical plots in the Northeast black soil region while radar satellites were passing overhead, measuring the true physical parameters of the soil as the true values ​​of soil compaction. By analyzing the statistical regularity between these true values ​​and the simultaneously acquired surface roughness index, a clear negative correlation was found between the two: the rougher the surface, the looser the soil and the lower the bulk density. To ensure that the final output topsoil structure compaction coefficient can intuitively reflect the degree of soil compaction—that is, the larger the value, the more compact—this invention adopts a monotonically increasing mapping relationship. This means that when the surface roughness index increases due to the soil becoming smoother and more compacted, the output topsoil structure compaction coefficient will also increase accordingly. This accurately transforms radar remote sensing information into a quantitative indicator that is positively correlated with the soil physical state and can be directly used for risk assessment. Furthermore, the specific form and internal parameters of the monotonically increasing mapping relationship are localized and optimized based on the unique soil physicochemical properties of the Northeast Black Soil Region to ensure its accuracy and reliability in actual operational use.

[0061] The innovation of this method lies in utilizing radar polarimetric scattering characteristics to achieve large-scale, high-frequency monitoring of soil structure, avoiding the high cost and low efficiency of traditional field sampling methods. Simultaneously, the ratio method reduces the impact of absolute radiometric calibration errors. This non-contact monitoring method is particularly suitable for the large-scale monitoring needs of black soil, enabling rapid acquisition of regional-scale soil structure information and providing crucial structural dimension input for cold disaster risk assessment. Its design fully considers the characteristics of radar remote sensing technology and the actual needs of black soil monitoring.

[0062] This step, as the core feature extraction link of the entire black soil cold disaster monitoring method, transforms the raw observation data into cold disaster monitoring feature indicators with clear physical meaning. It provides quantitative input parameters in three dimensions—thermal conditions, freeze-thaw dynamics, and soil structure—for subsequent risk index fusion calculation. At the same time, it realizes a multi-angle characterization of black soil cold disaster risk through multi-source data fusion, ensuring the data processing integrity and indicator calculation accuracy of the monitoring system.

[0063] This step automates the conversion from raw data to characteristic indicators, significantly improving data processing efficiency and standardization. By introducing historical benchmarks and a dynamic spring start date determination mechanism, it enhances the sensitivity and adaptability of effective accumulated temperature deficit to abnormal low-temperature events. Utilizing multi-depth temperature data for front identification improves the accuracy and physical rationality of calculating freeze-thaw front depth differences. Large-scale soil structure monitoring based on radar polarization scattering characteristics overcomes the spatial limitations and cost constraints of traditional field sampling methods. The three characteristic indicators reflect cold disaster risks from different dimensions, forming a complementary monitoring system that provides comprehensive data support for subsequent risk assessment.

[0064] In the implementation process, in the calculation of effective accumulated temperature deficit, effective accumulated temperature can be used to replace active accumulated temperature to adapt to the biological characteristics of different crops, or a growth degree-day model can be used in combination with the biological zero degree of specific crops for calculation; in the calculation of freeze-thaw front depth difference, a time domain reflectometer can be used to directly measure the soil volumetric water content profile to identify the freeze-thaw interface, or the soil temperature field can be inverted through the heat conduction equation to locate the freeze-thaw front; in the calculation of topsoil structure compaction coefficient, the texture features of optical remote sensing images can be used to estimate surface roughness, or a microwave scattering model can be combined to invert soil parameters, or high-precision three-dimensional surface structure information can be obtained using UAV lidar.

[0065] Based on the cumulative value of the freeze-thaw front depth difference within a preset time window, a cumulative freeze-thaw damage amount is generated. Based on the cumulative freeze-thaw damage amount, a freeze-thaw correction coefficient is determined according to a preset monotonically increasing mapping relationship. The topsoil structure compactness coefficient is multiplied by the freeze-thaw correction coefficient to obtain the corrected compactness coefficient.

[0066] This step is a crucial damage quantification and correction step in the entire black soil cold disaster monitoring method. By introducing the concept of cumulative freeze-thaw damage, the destructive effect of the freeze-thaw process on soil structure is quantified and characterized, and this destructive effect is incorporated into the correction calculation of soil compaction.

[0067] Among them, the cumulative amount of freeze-thaw damage reflects the cumulative degree to which the depth difference between freeze-thaw fronts exceeds the preset depth threshold within the preset time window. The larger the cumulative value, the more severe the damage to the soil structure caused by the freeze-thaw process. The corrected compaction coefficient, as a bridge connecting freeze-thaw damage and soil structure, has a value range that starts from 1 and gradually increases with the increase of the cumulative amount of freeze-thaw damage, reflecting the enhancing effect of freeze-thaw damage on soil compaction. The final corrected compaction coefficient is obtained by multiplying the original topsoil structure compaction coefficient with the freeze-thaw correction coefficient. The design of this multiplication operation fully considers the superposition effect of freeze-thaw damage and the original soil structure, ensuring the physical rationality and calculation accuracy of the correction result.

[0068] The innovation of this step lies in establishing a quantitative relationship between the cumulative amount of freeze-thaw damage and the correction of soil compaction, which breaks through the limitation of traditional soil structure monitoring that ignores the dynamic influence of the freeze-thaw process, and provides more accurate soil structure parameters for the risk assessment of cold disasters in black soil.

[0069] The calculation of the cumulative freeze-thaw damage is as follows:

[0070] Obtain the depth difference of the freeze-thaw front at each sampling moment within the preset time window;

[0071] For each sampling moment, if the depth difference between the freeze-thaw front at that moment is greater than a preset depth threshold, then the difference between the depth difference between the freeze-thaw front at that moment and the preset depth threshold is calculated as the instantaneous excess depth at that moment.

[0072] The cumulative damage amount is obtained by integrating the instantaneous excess depth at all sampling moments within the preset time window, and is used as the cumulative freeze-thaw damage amount.

[0073] In this calculation process, the preset time window ranges from 7 to 30 days. This range is chosen to fully consider the typical time scale and monitoring needs of the black soil freeze-thaw process. A shorter time window can improve the real-time response capability to freeze-thaw damage, while a longer time window can better reflect the cumulative effect of freeze-thaw damage. The preset depth threshold ranges from 5 to 15 centimeters. This range is determined based on the sensitivity of black soil structure to the freeze-thaw process. A lower threshold can improve the detection sensitivity of minor freeze-thaw damage, while a higher threshold can avoid misjudgment caused by normal freeze-thaw fluctuations.

[0074] In practical implementation, the sampling time should be consistent with the soil temperature data acquisition frequency to ensure data continuity and integrity. The calculation of instantaneous excess depth is only performed when the depth difference between the freeze-thaw fronts exceeds a preset depth threshold. This design can effectively filter out freeze-thaw fluctuations within the normal range and focus on the damage quantification of abnormal freeze-thaw events. The integral operation adopts numerical integration methods, such as the trapezoidal rule or Simpson's rule, to transform discrete instantaneous excess depth data into continuous cumulative damage. This design not only ensures the accuracy of the calculation but also facilitates practical programming implementation.

[0075] The calculation process of cumulative freeze-thaw damage reflects a deep understanding of the dynamic characteristics of the freeze-thaw process. By setting preset depth thresholds and instantaneous excess depths, the cumulative impact of the freeze-thaw process on soil structure can be accurately quantified, providing reliable input parameters for subsequent calculation of the corrected compaction coefficient.

[0076] The freeze-thaw correction factor is calculated as follows:

[0077] An exponential saturation mapping relationship is adopted, and the minimum value of the freeze-thaw correction coefficient is set to 1, and the maximum value is set to a preset upper limit value;

[0078] When the cumulative amount of freeze-thaw damage is 0, the freeze-thaw correction coefficient takes the minimum value of 1;

[0079] When the cumulative amount of freeze-thaw damage is greater than 0, the freeze-thaw correction coefficient gradually increases from 1 as the cumulative amount of freeze-thaw damage increases, and the rate of increase gradually slows down as the cumulative amount of freeze-thaw damage increases, eventually approaching the preset upper limit value.

[0080] The rate of increase is controlled by a preset saturation rate constant.

[0081] The freeze-thaw correction factor is calculated using an exponential saturation mapping method. This mapping ensures that the value of the freeze-thaw correction factor always falls between 1 and a preset upper limit. Specifically:

[0082] ;

[0083] in This is the freeze-thaw correction factor. The upper limit value is a preset value, which ranges from 1.5 to 3.0. The selection of this range fully considers the physical limits of the impact of freeze-thaw damage on soil compaction and actual observation data. The lower upper limit value is suitable for areas with less freeze-thaw damage, while the higher upper limit value is suitable for areas with severe freeze-thaw damage.

[0084] The preset saturation rate constant is defined as 0.01 to 0.1. This preset saturation rate constant controls the rate at which the freeze-thaw correction coefficient increases from 1 to near the preset upper limit. The larger the saturation rate constant, the faster the freeze-thaw correction coefficient increases, and the faster it can respond to changes in the cumulative amount of freeze-thaw damage. The smaller the saturation rate constant, the slower the freeze-thaw correction coefficient increases, and the smoother the response to changes in the cumulative amount of freeze-thaw damage.

[0085] The cumulative amount of freeze-thaw damage is denoted as 1. When the cumulative amount of freeze-thaw damage is zero, the freeze-thaw correction coefficient is equal to 1, indicating that the freeze-thaw process does not produce any additional compaction effect on the soil structure. As the cumulative amount of freeze-thaw damage increases, the freeze-thaw correction coefficient gradually increases from 1, but the rate of increase shows a decreasing trend. That is, for every increase of the cumulative amount of freeze-thaw damage, the increment of the freeze-thaw correction coefficient gradually becomes smaller. When the cumulative amount of freeze-thaw damage reaches a sufficiently large value, the freeze-thaw correction coefficient will gradually approach the preset upper limit value, but will never exceed the upper limit value. This gradual approach characteristic reflects the physical saturation phenomenon of the effect of freeze-thaw damage on soil compaction.

[0086] The calculation process of the freeze-thaw correction coefficient fully considers the potential impact of freeze-thaw damage on soil structure. By multiplying the topsoil structure compaction coefficient with the freeze-thaw correction coefficient, the dynamic correction of the topsoil structure compaction coefficient is achieved. This method can accurately reflect the comprehensive impact of the freeze-thaw process on soil structure and provide more accurate soil structure parameters for subsequent cold disaster risk assessment.

[0087] This step, a crucial damage quantification and correction step in the entire black soil cold disaster monitoring method, primarily aims to quantify and characterize the destructive effect of the freeze-thaw process on soil structure and integrate this destructive effect into the correction calculation of soil compaction. This provides soil structure parameters corrected for freeze-thaw damage for the subsequent fusion calculation of the cold disaster risk index. At the same time, by introducing the concept of cumulative freeze-thaw damage, a quantitative relationship between the dynamic characteristics of the freeze-thaw process and the state of soil structure is established, ensuring that the monitoring system accurately captures and reasonably corrects the freeze-thaw damage effect.

[0088] This step achieves a quantitative characterization of the impact of freeze-thaw damage on soil structure, overcoming the limitation of traditional soil structure monitoring that ignores the dynamic influence of the freeze-thaw process. By calculating and integrating the instantaneous excess depth, the cumulative amount of freeze-thaw damage obtained simultaneously considers the intensity and duration of freeze-thaw misalignment, which is more accurate than simply counting the number of freeze-thaw cycles. By setting preset depth thresholds and instantaneous excess depths, freeze-thaw fluctuations within the normal range are effectively filtered out, focusing on the quantification of damage from abnormal freeze-thaw events. An exponential saturation mapping relationship is used to realize the nonlinear response of the corrected compaction coefficient, which conforms to the physical laws of the impact of freeze-thaw damage on soil structure. By multiplying the topsoil compaction coefficient with the freeze-thaw correction coefficient, the superposition effect of freeze-thaw damage and the original soil structure state is accurately reflected. The flexible setting of parameters such as preset time window, preset depth threshold, preset upper limit value, and preset saturation rate constant allows this method to adapt to the monitoring needs of different regions and different freeze-thaw conditions.

[0089] In the implementation process, in calculating the cumulative amount of freeze-thaw damage, a weighted integral method can be used instead of a simple integral, assigning different weights according to the freeze-thaw sensitivity of different depth layers; or a peak detection method can be used, accumulating only the peak freeze-thaw front depth difference exceeding a preset depth threshold; in calculating the corrected compaction coefficient, a linear mapping relationship, a logarithmic mapping relationship, or a power function mapping relationship can be used instead of an exponential saturation mapping relationship; or machine learning methods, such as support vector regression and random forest regression, can be introduced to establish a nonlinear mapping relationship between the cumulative amount of freeze-thaw damage and the corrected compaction coefficient; or a more physically based freeze-thaw damage correction model can be established by combining measured data of soil mechanical parameters.

[0090] The effective accumulated temperature deficit and the modified compactness coefficient are weighted and fused together to generate the black soil cold disaster risk index.

[0091] This step is the final risk assessment stage of the entire black soil cold disaster monitoring method. By organically integrating the effective active accumulated temperature deficit value of the heat condition dimension with the modified compaction coefficient of the soil structure dimension, a comprehensive black soil cold disaster risk index is generated, providing a quantitative basis for agricultural production decisions.

[0092] Among them, the effective accumulated temperature deficit reflects the degree of shortage of heat resources in the current growing season relative to the historical average level, and is a heat dimension indicator of cold disaster risk; the modified compaction coefficient integrates the original soil structure state and the cumulative effect of freeze-thaw damage, and is a structural dimension indicator of cold disaster risk; the weighted fusion calculation design fully considers the differences in the contribution of the two dimensions to cold disaster risk, and ensures the scientificity and accuracy of the risk index through reasonable weight allocation. The innovation of this step lies in establishing a multi-dimensional indicator fusion calculation framework, breaking through the limitations of traditional single indicator assessment, and realizing a comprehensive quantitative characterization of cold disaster risk in black soil.

[0093] The calculation of the black soil cold disaster risk index is as follows:

[0094] The effective active accumulated temperature deficit value is normalized to obtain the heat deficit sub-index;

[0095] The modified compactness coefficient is normalized to obtain the structural damage sub-index;

[0096] Obtain the preset first weight and second weight, calculate the first product of the heat deficit sub-index and the first weight, and the second product of the structural damage sub-index and the second weight, and use the sum of the first product and the second product as the black soil cold disaster risk index.

[0097] In this calculation process, the purpose of normalization is to convert the original indices with different dimensions and magnitudes into a unified dimensionless index, facilitating subsequent weighted fusion calculations. The normalization range for the heat deficit sub-index is set to 0 to 1, where 0 represents no heat deficit and 1 represents a heat deficit reaching a historical extreme. Similarly, the normalization range for the structural damage sub-index is also set to 0 to 1, where 0 represents no structural damage and 1 represents structural damage reaching a historical extreme. The normalization process employs a linear mapping method, mapping the original index values ​​to the interval between 0 and 1. This design ensures both the comparability of the normalization results and preserves the relative magnitudes of the original indices. The first and second weights both range from 0 to 1, and their sum equals 1. This constraint ensures the mathematical rationality of the weighted fusion calculation.

[0098] In practice, the heat deficit sub-index is multiplied by the first weight to obtain the contribution of the heat dimension to the risk index, and the structural damage sub-index is multiplied by the second weight to obtain the contribution of the structural dimension to the risk index. The sum of the two is the final black soil cold disaster risk index, which ranges from 0 to 1, with a higher value indicating a higher risk of cold disaster. This weighted fusion calculation design fully considers the complementarity and importance differences of the multi-dimensional indicators, and achieves a comprehensive assessment of cold disaster risk through reasonable weight allocation.

[0099] The first weight and the second weight are obtained as follows:

[0100] Acquire historical monitoring data of the monitoring area recorded during the same phenological stage as the current date within a preset number of years in the past. The historical monitoring data includes historical heat deficit sub-index sequences and historical structural damage sub-index sequences.

[0101] Calculate the variance of the historical heat deficit sub-index sequence and the variance of the historical structural damage sub-index sequence, respectively.

[0102] The ratio of the variance of the historical heat deficit sub-index sequence to the sum of the variances of the two is used as the first weight; the ratio of the variance of the historical structural damage sub-index sequence to the sum of the variances of the two is used as the second weight.

[0103] In this calculation process, the preset range of years is 15 to 30 years. This range is chosen to fully consider the availability and representativeness of historical data. Shorter years better reflect recent climate change characteristics, while longer years can fully capture historical variability. The determination of phenological stages needs to be combined with the growth cycle of local major crops, typically including key growth stages such as sowing, seedling, jointing, and heading, to ensure that historical data is comparable to the current monitoring period. The system can use default regional weights during initialization. As monitoring data accumulates, the historical database is updated annually or quarterly, and the variance is recalculated and the weights are updated.

[0104] Obtaining the historical heat deficit sub-index sequence and the historical structural damage sub-index sequence requires normalizing monitoring data from the same phenological stage within a preset number of years to form standardized time series data. The variance is calculated using the sample variance formula, reflecting the dispersion of historical data. A larger variance indicates greater historical fluctuation in that dimension's indicators and a more significant contribution to the risk of cold disasters.

[0105] The first weight is calculated by dividing the variance of the historical heat deficit sub-index sequence by the sum of the two variances; the resulting ratio is the first weight. The second weight is calculated by dividing the variance of the historical structural damage sub-index sequence by the sum of the two variances; the resulting ratio is the second weight. This weight determination method based on historical data variance has dynamic adaptive characteristics, automatically adjusting the relative importance of the two-dimensional indicators according to the statistical characteristics of historical monitoring data, avoiding the limitations of subjectively setting weights.

[0106] In practical applications, if a certain dimension indicator has a large historical fluctuation, it indicates that the indicator has a strong ability to identify cold disaster risks, and the corresponding weight will automatically increase; conversely, if a certain dimension indicator has a small historical fluctuation, it indicates that the indicator has a weak ability to identify cold disaster risks, and the corresponding weight will automatically decrease. This weight determination method fully considers the statistical characteristics of historical data and the actual needs of cold disaster risk assessment, ensuring the scientific and reasonable allocation of weights.

[0107] This step, as the final risk assessment stage of the entire black soil cold disaster monitoring method, mainly plays the role of organically integrating the effective active accumulated temperature deficit value of the heat condition dimension with the corrected compaction coefficient of the soil structure dimension to generate a comprehensive black soil cold disaster risk index, providing a quantitative basis for agricultural production decisions. At the same time, through normalization processing and weighted fusion calculation, the comparability of indicators of different dimensions and the scientific nature of the fusion results are ensured. By using a weight determination method based on historical data variance, dynamic adaptation and objectivity of weight allocation are achieved, providing the final risk assessment output for the entire monitoring system.

[0108] This step achieves a comprehensive quantitative representation of multi-dimensional cold disaster risk indicators, breaking through the limitations of traditional single-indicator assessments. Normalization transforms original indicators of different dimensions and magnitudes into a unified dimensionless index, improving the comparability and accuracy of indicator fusion. A weighting determination method based on historical data variance is adopted, achieving dynamic adaptability and objectivity in weight allocation and avoiding the limitations of subjective weight setting. The weighted fusion calculation design fully considers the complementarity and importance differences of multi-dimensional indicators, ensuring the scientific validity and practicality of the risk index. The black soil cold disaster risk index ranges from 0 to 1, facilitating risk level classification and decision-making applications. The entire calculation process exhibits good repeatability and verifiability, providing a standardized assessment method for black soil cold disaster monitoring.

[0109] In the implementation process, logarithmic normalization or exponential normalization can be used instead of linear normalization in the normalization process to adapt to the distribution characteristics of different indicators; in the determination of weights, subjective or objective weighting methods such as expert scoring, analytic hierarchy process, or entropy weighting can be used instead of variance-based weight determination methods; in the weighted fusion calculation, multiplicative fusion, geometric mean, or harmonic mean can be used instead of weighted summation methods; machine learning methods can also be introduced to establish a nonlinear mapping relationship between multidimensional indicators and cold disaster risk, or fuzzy comprehensive evaluation methods can be used to handle the uncertainty and fuzziness of indicators.

[0110] The system acquires the crop phenological stages of the monitored area, matches preset early warning rules with the black soil cold disaster risk index and the crop phenological stages, and outputs graded early warning information.

[0111] This step is the final decision-making output stage of the entire black soil cold disaster monitoring method. By combining the quantitatively calculated black soil cold disaster risk index with the current phenological stage of the crop, it achieves graded early warning of cold disaster risk. The crop phenological stage can be obtained through various methods such as remote sensing monitoring, ground observation, or farmer reporting, ensuring accurate identification of crop growth status in the monitoring area. The preset early warning rules are based on a decision rule library established using a large number of historical cold disaster cases and crop physiological characteristics, covering the correspondence between different crop types, different growth stages, and the cold disaster risk index. The output of graded early warning information adopts a standardized early warning level classification, typically including four levels: blue, yellow, orange, and red, corresponding to low risk, medium risk, high risk, and extremely high risk, respectively. The innovation of this step lies in establishing a dynamic early warning mechanism based on crop phenological stages, breaking through the limitations of traditional fixed threshold early warnings, and achieving accurate identification and graded early warning of cold disaster risk for crops at different growth stages.

[0112] The generation of the tiered early warning information is specifically as follows:

[0113] A table of cold disaster tolerance thresholds for major crops in black soil at different growth stages is pre-stored. The table of cold disaster tolerance thresholds includes at least the upper limit of the tolerance index of black soil cold disaster risk index for corn, soybean, and rice at the three-leaf stage, jointing stage, and pod-filling stage, respectively.

[0114] Based on the crop type and growth stage of the current monitoring area, the corresponding tolerance upper limit is queried from the cold disaster tolerance threshold table;

[0115] The black soil cold disaster risk index is compared with the tolerance limit. If it exceeds the tolerance limit, a preset level is matched according to the extent of the exceedance, and the corresponding level of level warning information is output.

[0116] In this generation process, the construction of the cold disaster tolerance threshold table is the core foundation of the entire early warning system. This threshold table is determined comprehensively based on historical cold disaster cases of major crops in the black soil region, physiological characteristic test data, and expert experience. Specifically, the upper limit of the cold disaster risk index tolerance for maize is set at 0.55 to 0.65 at the three-leaf stage, 0.60 to 0.70 at the jointing stage, and 0.65 to 0.75 at the pod-setting stage; for soybeans, the upper limit is set at 0.50 to 0.60 at the three-leaf stage, 0.55 to 0.65 at the jointing stage, and 0.60 to 0.70 at the pod-setting stage; and for rice, the upper limit is set at 0.45 to 0.55 at the three-leaf stage, 0.50 to 0.60 at the jointing stage, and 0.55 to 0.65 at the pod-setting stage. The selection of these threshold ranges fully considers the differences in sensitivity to cold disasters among different crop types and different growth stages; the lower the value, the more sensitive the growth stage is to cold disasters.

[0117] In practical implementation, crop type identification can be achieved through remote sensing image classification, ground surveys, or farmer reports to ensure accurate determination of crop types in the monitoring area. Determining the growth period requires combining local accumulated temperature data, crop growth models, and field observations, typically using phenological calendars or accumulated temperature methods. When querying the corresponding tolerance upper limit from the cold disaster tolerance threshold table, both crop type and growth period dimensions need to be matched simultaneously to ensure the accuracy of the query results. The comparison between the black soil cold disaster risk index and the tolerance upper limit uses a simple numerical comparison method. When the risk index is less than or equal to the tolerance upper limit, it indicates that the current cold disaster risk is within an acceptable range and no warning is triggered. When the risk index exceeds the tolerance upper limit, the extent of the exceedance needs to be further calculated. First, the difference between the black soil cold disaster risk index and the tolerance upper limit is calculated, i.e., subtracting the tolerance upper limit from the black soil cold disaster risk index to obtain the absolute value of the exceedance. Then, this absolute value of the exceedance is divided by the tolerance upper limit to obtain the relative proportion of the exceedance. Finally, this relative proportion is multiplied by 100% and converted to a percentage form, which is the extent of the exceedance. This calculation method can quantify the degree to which the current cold disaster risk exceeds the crop's tolerance capacity, providing an objective basis for tiered early warning.

[0118] The specific rules for matching the exceedance magnitude with the preset classification are as follows: when the exceedance magnitude is less than 10%, a blue alert is triggered, indicating a low risk of cold disaster, and it is recommended to pay attention to weather changes; when the exceedance magnitude is between 10% and 30%, a yellow alert is triggered, indicating a moderate risk of cold disaster, and it is recommended to take preventive measures; when the exceedance magnitude is between 30% and 50%, an orange alert is triggered, indicating a high risk of cold disaster, and it is recommended to take emergency protective measures; when the exceedance magnitude is greater than 50%, a red alert is triggered, indicating an extremely high risk of cold disaster, and it is recommended to take immediate emergency measures and report to relevant departments. This tiered early warning mechanism fully considers the severity of cold disaster risk and the urgency of response measures, ensuring the practicality and operability of the early warning information.

[0119] This step, serving as the final decision-making output of the entire black soil cold disaster monitoring method, primarily functions to combine the quantitatively calculated black soil cold disaster risk index with the current phenological stage of the crop, thereby achieving graded early warning of cold disaster risk. This provides timely and accurate risk warning information for agricultural production decisions. Furthermore, by constructing a cold disaster tolerance threshold table based on crop type and growth stage, it enables dynamic adjustment and precise matching of early warning thresholds, ensuring the relevance and practicality of early warning information and providing the final decision support output for the entire monitoring system.

[0120] This step realizes a dynamic early warning mechanism based on crop phenological stages, breaking through the limitations of traditional fixed threshold early warning. By constructing a cold disaster tolerance threshold table, it fully considers the differences in sensitivity to cold disasters among different crop types and different growth stages, improving the accuracy of early warning. The adoption of a hierarchical early warning mechanism, which divides different early warning levels according to the extent of exceedance, ensures the hierarchy and operability of early warning information. The output of early warning information adopts a standardized format, which is easy for farmers to understand and implement. The tolerance threshold table can be locally calibrated based on historical disaster damage data and regional variety characteristics, improving the adaptability of the method. The entire early warning process has good repeatability and verifiability, providing a standardized early warning method for monitoring cold disasters in black soil.

[0121] During implementation, in constructing the cold disaster tolerance threshold table, machine learning methods can be used to automatically learn thresholds based on historical cold disaster case data, replacing the method of manually setting thresholds; in early warning classification, fuzzy logic methods can be used to handle the uncertainty and ambiguity of cold disaster risk, or probabilistic early warning methods can be used to output the probability of cold disaster occurrence; in early warning information output, multimodal output methods can be used to replace plain text output, or intelligent recommendation systems can be used to automatically generate response measures suggestions based on the early warning level; expert systems can also be introduced, combined with agronomic expert knowledge bases, to provide more personalized early warning suggestions and response strategies.

[0122] A method for monitoring cold disasters in black soil based on multi-source data fusion, which also includes an agriculturally suitable terminal, specifically:

[0123] At the IoT gateway in the monitoring area, the acquired air temperature data, layered soil temperature data, and radar satellite data are cached to local storage according to timestamps. The local storage retains the original data within the most recent preset number of days to support offline caching in the absence of network access.

[0124] The IoT gateway acquires the automatic location information of the monitored plots and stores the location information in association with cached data.

[0125] When a network interruption is detected, the calculation of the effective active accumulated temperature deficit, the freeze-thaw front depth difference, the topsoil structure compaction coefficient, the corrected compaction coefficient, and the black soil cold disaster risk index is continued using locally cached data, and the calculation results are cached.

[0126] When the network is restored, the cached calculation results will be synchronized to the cloud server;

[0127] The agricultural-friendly terminal supports zero-code voice interaction, receives user voice input, parses the plot identification and query command in the voice input, retrieves the corresponding black soil cold disaster risk index or graded early warning information from the local cache or cloud server according to the query command, and outputs it through voice broadcast.

[0128] In this agricultural-friendly terminal interaction, the IoT gateway, as the core data processing node of the entire monitoring system, undertakes multiple functions such as data caching, local computing, and network status monitoring. The preset number of days ranges from 3 to 15 days, preferably 7 days. This time range fully considers the actual needs of black soil cold disaster monitoring and the limitations of local storage capacity. Shorter days reduce storage pressure, while longer days provide longer offline computing capabilities. The local storage capacity is configured from 32 to 128 gigabytes, with the specific capacity adjusted according to the data acquisition frequency and data volume of the monitoring area to ensure that all raw data within the preset number of days can be stored. Air temperature data is acquired hourly, and stratified soil temperature data is acquired every hour. Based on the satellite transit cycle, the typical revisit period is 6 to 12 days. These acquisition frequencies ensure data timeliness while avoiding storage pressure caused by excessive acquisition.

[0129] The IoT gateway acquires automatic location information of monitored sites using the Global Positioning System (GPS) or the BeiDou Navigation Satellite System, achieving sub-meter accuracy to ensure the accuracy of the monitored site's location information. The location information and cached data are stored in a timestamp-aligned manner, meaning each piece of raw data is marked with a precise timestamp and location information, forming a spatiotemporally linked data record for easy subsequent data retrieval and analysis. When a network interruption is detected, the IoT gateway automatically switches to offline mode, continuing to calculate effective active accumulated temperature deficit, freeze-thaw front depth difference, topsoil structure compaction coefficient, corrected compaction coefficient, and black soil cold disaster risk index using locally cached data. These calculation processes are completely consistent with the calculation logic in network mode, ensuring the accuracy and consistency of the calculation results. The calculation results are cached using a first-in, first-out (FIFO) storage strategy; when cache space is insufficient, the oldest data is automatically deleted to ensure that the latest calculation results are always available.

[0130] When the network is restored, the IoT gateway automatically detects the network connection status and synchronizes the cached calculation results to the cloud server in chronological order. The synchronization process uses an incremental synchronization mechanism, meaning that only the data that has changed is transmitted, avoiding duplicate transmissions that would waste network bandwidth. After synchronization is complete, the cloud server verifies and integrates the uploaded data to ensure data integrity and consistency.

[0131] One of the innovations of this step is the zero-code voice interaction function of the agricultural-friendly terminal. This function adopts deep learning-based speech recognition technology, which can accurately recognize the user's voice input and parse the land parcel identifier and query command in the voice input through natural language processing technology. The recognition of land parcel identifier supports multiple expressions, including land parcel number, land parcel name, geographical location description, etc. The recognition of query command supports multiple query types, including current risk index query, historical data query, early warning information query, etc.

[0132] The voice broadcast output employs high-quality text-to-speech technology, supporting multiple dialects and adjustable speech rates to ensure farmers can clearly understand the broadcast content. In practice, when a user inputs information via voice, the system automatically recognizes the land parcel identifier. If the query command is "cold soil disaster risk query," the system retrieves the corresponding black soil cold soil disaster risk index and tiered early warning information from the local cache or cloud server. It then broadcasts the current black soil cold soil disaster risk index for the land parcel, its warning level, and recommended measures.

[0133] This step serves as the terminal interaction and data management link in the entire black soil cold disaster monitoring method. Its main function is to realize local caching and offline computing of multi-source data through the Internet of Things gateway, ensuring the continuous operation of the monitoring system in the event of network interruption. At the same time, the voice zero-code interaction function reduces the usage threshold for farmers, improves the ease of use and popularization of the system, provides reliable data management and a user-friendly interface for the entire monitoring system, and ensures the integrity of monitoring data and the timely transmission of early warning information.

[0134] This step enables local caching and offline computation of monitoring data, ensuring the continuous operation of the monitoring system even during network outages and improving system reliability and stability. By automatically associating location information with cached data, precise spatiotemporal management of monitoring data is achieved, facilitating subsequent data retrieval and analysis. An incremental synchronization mechanism synchronizes cached computation results to the cloud server, effectively reducing network bandwidth consumption and improving data transmission efficiency. The zero-code voice interaction function significantly lowers the barrier to entry for farmers, allowing them to query data and obtain early warning information without requiring specialized computer skills, thus improving the system's ease of use and accessibility. Support for multiple dialects and adjustable speech rates ensures farmers can clearly understand the broadcast content, improving the accuracy of information delivery.

[0135] During implementation, in terms of data caching, distributed storage technology can be used to replace local storage, or a cloud-edge collaborative storage architecture can be adopted to achieve hierarchical data storage. In terms of network status detection, various technical means such as heartbeat detection mechanisms and network quality monitoring can be used. In terms of voice interaction, various interaction methods such as graphical interfaces, touch screen interaction, and SMS interaction can be used to replace voice interaction, or multimodal interaction technology can be used in combination with voice, image, gesture and other interaction methods. In terms of data synchronization, various synchronization strategies such as real-time synchronization, timed synchronization, and event-triggered synchronization can be used. Edge computing technology can also be introduced to deploy lightweight artificial intelligence models at the IoT gateway to achieve more complex local data analysis and decision support functions.

[0136] The technical scope of this invention is not limited to the content described above. Those skilled in the art can make various modifications and variations to the above embodiments without departing from the technical concept of this invention, and all such modifications and variations should fall within the protection scope of this invention.

Claims

1. A method for monitoring cold disasters in black soil based on multi-source data fusion, characterized in that, Specifically, it includes: Acquire air temperature data, layered soil temperature data, and radar / satellite data for the monitored area; The air temperature data is accumulated and processed to calculate the effective active accumulated temperature deficit. Frontal identification processing is performed on the layered soil temperature data to calculate the freeze-thaw front depth difference; The radar satellite data is subjected to polarization decomposition processing to calculate the topsoil structure compaction coefficient; Based on the cumulative value of the freeze-thaw front depth difference within a preset time window, a cumulative freeze-thaw damage amount is generated. Based on the cumulative freeze-thaw damage amount, a freeze-thaw correction coefficient is determined according to a preset monotonically increasing mapping relationship. The topsoil structure compactness coefficient is multiplied by the freeze-thaw correction coefficient to obtain the corrected compactness coefficient. The effective active accumulated temperature deficit value and the modified compaction coefficient are weighted and fused together to generate the black soil cold disaster risk index. The system acquires the crop phenological stages of the monitored area, matches preset early warning rules with the black soil cold disaster risk index and the crop phenological stages, and outputs graded early warning information.

2. The method for monitoring cold disasters in black soil based on multi-source data fusion according to claim 1, characterized in that: The calculation of the effective active accumulated temperature deficit is as follows: Determine the start date of spring for the current year. The start date of spring is the first day when the average daily temperature is greater than or equal to a preset temperature threshold for a consecutive preset number of days in spring. Starting from the start date of spring for the current year, accumulate the accumulated temperature above zero degrees Celsius day by day to obtain the current actual accumulated temperature. For each historical year within a preset number of years, determine the start date of spring for that historical year, and obtain the cumulative accumulated temperature of activities greater than zero degrees from the start date of spring of that historical year to the same month and day as the current date. Calculate the average of the cumulative accumulated temperature of activities greater than zero degrees obtained for all historical years as the baseline accumulated temperature for the same historical period. Calculate the difference between the historical baseline accumulated temperature and the current actual accumulated temperature; If the difference is greater than 0, then the difference is taken as the effective active accumulated temperature deficit value; If the difference is less than or equal to 0, the effective active accumulated temperature deficit is determined to be 0.

3. The method for monitoring cold disasters in black soil based on multi-source data fusion according to claim 1, characterized in that: The calculation of the depth difference between the freeze-thaw fronts is as follows: Acquire temperature values ​​collected by soil temperature sensors at at least three different depth layers, including shallow, middle and deep layers, which correspond to preset depth ranges of the black soil tillage layer, plow layer and upper subsoil layer, respectively. Determine whether there is a first-type depth layer with a temperature greater than zero degrees Celsius and a second-type depth layer with a temperature less than zero degrees Celsius among all depth layers. If there is a layer with a temperature equal to 0 degrees Celsius, it can be classified into either side. If both the first type of depth layer and the second type of depth layer exist simultaneously, the position with the greatest depth in the first type of depth layer is identified as the first front, and the position with the least depth in the second type of depth layer is identified as the second front. The vertical distance between the first front and the second front is calculated, and this vertical distance is taken as the freeze-thaw front depth difference. If the first type of depth layer and the second type of depth layer do not exist simultaneously, the depth difference of the freeze-thaw front is determined to be 0.

4. The method for monitoring cold disasters in black soil based on multi-source data fusion according to claim 1, characterized in that: The calculation of the topsoil structure compaction coefficient is as follows: Extract polarimetric synthetic aperture radar data from the radar satellite data; Polarization decomposition parameters are extracted from the polarization synthetic aperture radar data, and the polarization decomposition parameters include at least the surface scattering ratio and the secondary scattering ratio; The surface roughness index is calculated based on the ratio of the surface scattering ratio to the secondary scattering ratio. Based on the surface roughness index, the topsoil structure compaction coefficient is determined according to a preset monotonically increasing mapping relationship.

5. The method for monitoring cold disasters in black soil based on multi-source data fusion according to claim 1, characterized in that: The calculation of the cumulative freeze-thaw damage is as follows: Obtain the depth difference of the freeze-thaw front at each sampling moment within the preset time window; For each sampling moment, if the depth difference between the freeze-thaw front at that moment is greater than a preset depth threshold, then the difference between the depth difference between the freeze-thaw front at that moment and the preset depth threshold is calculated as the instantaneous excess depth at that moment. The cumulative damage amount is obtained by integrating the instantaneous excess depth at all sampling moments within the preset time window, and is used as the cumulative freeze-thaw damage amount.

6. The method for monitoring cold disasters in black soil based on multi-source data fusion according to claim 5, characterized in that: The freeze-thaw correction factor is calculated as follows: An exponential saturation mapping relationship is adopted, and the minimum value of the freeze-thaw correction coefficient is set to 1, and the maximum value is set to a preset upper limit value; When the cumulative amount of freeze-thaw damage is 0, the freeze-thaw correction coefficient takes the minimum value of 1; When the cumulative amount of freeze-thaw damage is greater than 0, the freeze-thaw correction coefficient gradually increases from 1 as the cumulative amount of freeze-thaw damage increases, and the rate of increase gradually slows down as the cumulative amount of freeze-thaw damage increases, eventually approaching the preset upper limit value. The rate of increase is controlled by a preset saturation rate constant.

7. The method for monitoring cold disasters in black soil based on multi-source data fusion according to claim 1, characterized in that: The calculation of the black soil cold disaster risk index is as follows: The effective active accumulated temperature deficit value is normalized to obtain the heat deficit sub-index; The modified compactness coefficient is normalized to obtain the structural damage sub-index; Obtain the preset first weight and second weight, calculate the first product of the heat deficit sub-index and the first weight, and the second product of the structural damage sub-index and the second weight, and use the sum of the first product and the second product as the black soil cold disaster risk index.

8. A method for monitoring cold disasters in black soil based on multi-source data fusion according to claim 7, characterized in that: The first weight and the second weight are obtained as follows: Acquire historical monitoring data of the monitoring area recorded during the same phenological stage as the current date within a preset number of years in the past. The historical monitoring data includes historical heat deficit sub-index sequences and historical structural damage sub-index sequences. Calculate the variance of the historical heat deficit sub-index sequence and the variance of the historical structural damage sub-index sequence, respectively. The ratio of the variance of the historical heat deficit sub-index sequence to the sum of the variances of the two is used as the first weight; the ratio of the variance of the historical structural damage sub-index sequence to the sum of the variances of the two is used as the second weight.

9. A method for monitoring cold disasters in black soil based on multi-source data fusion according to claim 1, characterized in that: The generation of the tiered early warning information is specifically as follows: A table of cold disaster tolerance thresholds for major crops in black soil at different growth stages is pre-stored. The table of cold disaster tolerance thresholds includes at least the upper limit of the tolerance index of black soil cold disaster risk index for corn, soybean, and rice at the three-leaf stage, jointing stage, and pod-filling stage, respectively. Based on the crop type and growth stage of the current monitoring area, the corresponding tolerance upper limit is queried from the cold disaster tolerance threshold table; The black soil cold disaster risk index is compared with the tolerance limit. If it exceeds the tolerance limit, a preset level is matched according to the extent of the exceedance, and the corresponding level of level warning information is output.

10. A method for monitoring cold disasters in black soil based on multi-source data fusion according to claim 1, characterized in that: This also includes agricultural-friendly terminals, specifically: At the IoT gateway in the monitoring area, the acquired air temperature data, layered soil temperature data, and radar satellite data are cached to local storage according to timestamps. The local storage retains the original data within the most recent preset number of days to support offline caching in the absence of network access. The IoT gateway acquires the automatic location information of the monitored plots and stores the location information in association with cached data. When a network interruption is detected, the calculation of the effective active accumulated temperature deficit, the freeze-thaw front depth difference, the topsoil structure compaction coefficient, the corrected compaction coefficient, and the black soil cold disaster risk index is continued using locally cached data, and the calculation results are cached. When the network is restored, the cached calculation results will be synchronized to the cloud server; The agricultural-friendly terminal supports zero-code voice interaction, receives user voice input, parses the plot identification and query command in the voice input, retrieves the corresponding black soil cold disaster risk index or graded early warning information from the local cache or cloud server according to the query command, and outputs it through voice broadcast.

Images