Straw mulching ecological effect dynamic prediction method and system based on multi-source data fusion

Abstract

The present application discloses a multi-source data fusion straw returning to field ecological effect dynamic prediction method and system, relates to the technical field of agricultural technology popularization and service technology, and comprises the following steps: obtaining the straw decomposition rate, the nutrient release rate time series, the soil temperature, the soil moisture content, the returning to field mode identifier and the baseline values of soil organic matter and available nutrients of a target field plot; determining the straw decomposition process index according to the straw decomposition rate and the nutrient release rate time series; obtaining the decomposition response lag coefficient through a lag evaluation model based on the straw decomposition process index, the soil temperature and the moisture content; correcting the changes of soil organic matter and available nutrients according to the decomposition response lag coefficient to obtain the ecological effect prediction result; generating the returning to field management strategy suggestion based on the prediction result and the returning to field mode identifier, and dynamically adjusting the next monitoring cycle. The problems that the straw returning to field ecological effect change is difficult to accurately evaluate in the prior art, resulting in insufficient accuracy of the straw returning to field effect evaluation and poor pertinence of the straw returning to field management strategy are solved.

Description

Technical Field

[0001] This invention relates to the field of agricultural technology extension services, specifically to a method and system for dynamic prediction of the ecological effects of straw returning to the field based on multi-source data fusion. Background Technology

[0002] Returning straw to the field is a common method of utilizing straw resources in agricultural production. After being returned to the field, straw gradually decomposes in the soil environment, releasing organic matter and nutrients. This process has a certain impact on soil properties such as organic matter and nutrients. Therefore, during the implementation of straw returning to the field, it is usually necessary to monitor the straw decomposition and related soil indicators to evaluate the ecological effects of straw returning to the field.

[0003] Current assessments of the ecological effects of straw return to the field typically analyze indicators such as straw decomposition, soil nutrient content, and soil organic matter content obtained within a certain monitoring period, and combine these with pre- and post-return testing results to evaluate changes in soil conditions. However, in practical applications, straw decomposition, nutrient release, and changes in soil properties all have a certain time frame. Factors such as soil environment, return methods, and climatic conditions in different fields can also cause variations in these indicators.

[0004] However, during the process of straw return to the field, the changes in the state of the straw itself and the changes in soil indicators are not always consistent. Relevant monitoring indicators may exhibit varying degrees of fluctuation at different time stages. This makes the evaluation of the ecological effects after straw return to the field based on monitoring data at limited time points easily affected by the monitoring time, field environment, and changes in indicators. At the same time, indicators such as soil organic matter and available nutrients have certain change cycles, and it is difficult to fully reflect the continuous changes during the straw return process based on only a single or limited monitoring result.

[0005] Furthermore, the state of straw in the field and its contact conditions with the soil environment vary under different straw return methods. Various environmental factors also change over time during the actual straw return process, resulting in differences in the ecological effects across different target fields. Current evaluation methods, when faced with the continuously changing straw return process, typically rely on periodic monitoring and manual analysis, making it difficult to promptly grasp the changes in the ecological effects of target fields. This, in turn, affects the evaluation of straw return effectiveness and the implementation of subsequent management work.

[0006] Therefore, how to more accurately and continuously evaluate the changes in the ecological effects of target fields based on monitoring data during the straw return process, and provide reliable data for the management of straw return process, has become a technical problem that needs to be solved in this field. Summary of the Invention

[0007] This invention provides a method and system for dynamic prediction of the ecological effects of straw returning to the field by integrating multi-source data. This addresses the technical problems in the prior art, such as the difficulty in accurately assessing changes in the ecological effects of straw returning to the field, resulting in insufficient accuracy in evaluating the effect of straw returning to the field and poor targeting of straw returning to the field management strategies.

[0008] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: In a first aspect, this invention provides a method for dynamically predicting the ecological effects of straw returning to the field based on multi-source data fusion, including: Obtain the time series of straw decomposition rate, straw nutrient release rate, soil temperature, soil moisture content, return method identification, soil organic matter baseline value, and soil available nutrient baseline value of the target field. Based on the time series of straw decomposition rate and straw nutrient release rate, the straw decomposition process index of the target field at the current monitoring time is analyzed and obtained; Based on the straw decomposition process index, soil temperature, and soil moisture content, the decomposition response lag coefficient of the target field is obtained through a pre-constructed lag assessment model. Based on the decomposition response hysteresis coefficient, and based on the baseline values ​​of soil organic matter and available soil nutrients, the corrected changes in soil organic matter and available soil nutrients in the target field at the current monitoring time are determined, and the corrected ecological effect prediction results are formed. Based on the revised ecological effect prediction results and the identification of the return-to-field method, the recommended return-to-field management strategy for the target field is generated, and the time interval of the next monitoring cycle is dynamically adjusted according to the decomposition response hysteresis coefficient.

[0009] Secondly, this invention provides a dynamic prediction system for the ecological effects of straw returning to the field based on multi-source data fusion, including: The data acquisition module is used to acquire the time series of straw decomposition rate, straw nutrient release rate, soil temperature, soil moisture content, return method identification, soil organic matter baseline value, and soil available nutrient baseline value of the target field. The decomposition process analysis module is used to analyze and obtain the straw decomposition process index of the target field at the current monitoring time based on the straw decomposition rate time series and the straw nutrient release rate time series. The hysteresis coefficient acquisition module is used to obtain the hysteresis coefficient of the decomposition response of the target field based on the straw decomposition process index, soil temperature and soil moisture content, through a pre-built hysteresis assessment model. The ecological effect prediction module is used to determine the corrected changes in soil organic matter and available soil nutrients in the target field at the current monitoring time based on the decomposition response hysteresis coefficient, the baseline values ​​of soil organic matter and available soil nutrients, and to form the corrected ecological effect prediction results. The field return management and monitoring adjustment module is used to generate field return management strategy suggestions for target fields based on the corrected ecological effect prediction results and field return method identification, and dynamically adjust the time interval of the next monitoring cycle according to the decomposition response hysteresis coefficient.

[0010] One or more technical solutions provided in this invention have at least the following technical effects or advantages: By integrating multi-source data such as straw decomposition rate, nutrient release rate, soil temperature, soil moisture content, and straw return method, the correlation between straw decomposition process and soil ecological response is constructed. The decomposition response lag coefficient is used to dynamically correct changes in soil organic matter and available nutrients, reducing prediction bias caused by the asynchronous response between the decomposition process and soil ecological indicators, and improving the accuracy and timeliness of predicting the ecological effects of straw return. At the same time, differentiated management strategies are generated based on the return method, and the monitoring cycle is dynamically adjusted according to the degree of decomposition response lag, so that the prediction results can be further used for the management of the return process and the optimization of monitoring resources, thereby improving the adaptability of monitoring and management of the ecological effects of straw return. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of the process for dynamically predicting the ecological effects of straw returning to the field based on multi-source data fusion provided in this embodiment of the invention. Figure 2 This is a schematic diagram illustrating the dynamic prediction principle of the ecological effects of straw returning to the field provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of the dynamic prediction system for the ecological effects of straw returning to the field based on multi-source data fusion provided in this embodiment of the invention.

[0012] In the attached diagram, the components represented by each number are as follows: Data acquisition module 11, decomposition process analysis module 12, hysteresis coefficient acquisition module 13, ecological effect prediction module 14, and return-to-field management and monitoring adjustment module 15. Detailed Implementation

[0013] This invention provides a method and system for dynamic prediction of the ecological effects of straw returning to the field by integrating multi-source data. It specifically solves the technical problems in the prior art, such as the difficulty in accurately assessing the changes in the ecological effects of straw returning to the field, resulting in insufficient accuracy in the evaluation of the returning effect and poor targeting of the returning management strategy.

[0014] It should be noted that the terms "comprising" and "having" are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to these processes, methods, products, or devices.

[0015] Example 1, as Figure 1 and Figure 2 As shown, this invention provides a method for dynamic prediction of the ecological effects of straw returning to the field based on multi-source data fusion, including: S100: Obtain the time series of straw decomposition rate, straw nutrient release rate, soil temperature, soil moisture content, return method identification, soil organic matter baseline value, and soil available nutrient baseline value of the target field. Step S100 in the method provided in this embodiment of the invention includes: Obtain the time series of straw decomposition rate of the target field. The straw decomposition rate time series is a sequence formed by arranging the straw decomposition rate values ​​obtained by sampling and measuring at multiple monitoring time points in chronological order. The straw decomposition rate is the percentage of decomposed straw mass to the initial straw mass. Obtain the time series of straw nutrient release rate of the target field. The straw nutrient release rate time series is a sequence formed by arranging the straw nutrient release rate values ​​obtained at multiple monitoring time points in chronological order. The straw nutrient release rate is the percentage of the amount of target nutrient element released in straw relative to the initial total amount of target nutrient element. Obtain the soil temperature and soil moisture content at the current monitoring time; Obtain the straw return method identifier for the target field. The straw return method identifier is a classification identifier for the straw return method adopted in the target field, including the straw crushing return method identifier, the straw covering return method identifier, and the straw turning and burying return method identifier. Obtain the baseline values ​​of soil organic matter and available nutrients in the target field. The baseline value of soil organic matter is the soil organic matter content of the target field measured before straw return, and the baseline value of available nutrients is the soil alkaline nitrogen, available phosphorus and available potassium content of the target field measured before straw return.

[0016] In this embodiment, the straw decomposition rate time series of the target field is first obtained. The straw decomposition rate time series is a sequence formed by arranging the straw decomposition rate values ​​obtained from sampling and measurement at multiple monitoring time points in chronological order, wherein the straw decomposition rate is the percentage of decomposed straw mass to the initial straw mass.

[0017] Specifically, the initial straw mass was obtained before straw was returned to the field, and straw samples were taken from the target field at multiple monitoring time points after straw return. The sampling area, sampling depth, and sampling method remained consistent across all monitoring time points, and the remaining straw mass at each monitoring time point was measured. The straw decomposition rate at each monitoring time point was determined as follows: Straw decomposition rate = (initial straw mass - remaining straw mass) ÷ initial straw mass × 100%.

[0018] For example, if the initial straw mass is 1000g, and the remaining straw mass measured at multiple monitoring time points after returning the straw to the field is 900g, 780g, 610g, and 480g respectively, then the corresponding straw decomposition rates are: 900g = (1000-900)÷1000×100%=10%, 780g = (1000-780)÷1000×100%=22%, 610g = (1000-610)÷1000×100%=39%, and 480g = (1000-480)÷1000×100%=52%. Arranging these straw decomposition rates in chronological order of the monitoring time points forms a straw decomposition rate time series.

[0019] Furthermore, the time series of straw nutrient release rate for the target field was obtained. The straw nutrient release rate time series is a sequence formed by arranging the straw nutrient release rate values ​​obtained from multiple monitoring time points in chronological order. The straw nutrient release rate is the percentage of the amount of target nutrient element released from the straw relative to the initial total amount of the target nutrient element.

[0020] Specifically, the initial total amount of target nutrient elements is obtained, and the remaining straw mass and corresponding target nutrient element mass fraction are measured at each monitoring time point. The straw nutrient release rate is determined based on the initial total amount of target nutrient elements and the remaining amount of target nutrient elements at the current monitoring time point.

[0021] The initial total amount of target nutrient elements = initial straw mass × initial mass fraction of target nutrient elements.

[0022] The remaining amount of target nutrient elements at the current monitoring time point = the remaining straw mass × the mass fraction of target nutrient elements at the current monitoring time point.

[0023] Straw nutrient release rate = (initial total amount of target nutrient elements - remaining amount of target nutrient elements at the current monitoring time point) ÷ initial total amount of target nutrient elements × 100%.

[0024] For example, if the initial straw mass is 1000g, the target nutrient element is nitrogen, and the initial mass fraction of the target nutrient element is 1.20%, then the initial total amount of the target nutrient element = (1000 × 0.012) = 12g; if the remaining straw mass is measured to be 780g at one of the monitoring time points, and the mass fraction of the target nutrient element at the current monitoring time point is 1.05%, then the remaining amount of the target nutrient element at the current monitoring time point = (780 × 0.0105) = 8.19g, corresponding to a straw nutrient release rate of (12 - 8.19) ÷ 12 × 100% = 31.75%.

[0025] When the target nutrient elements include nitrogen, phosphorus, and potassium, the straw nutrient release rates corresponding to nitrogen, phosphorus, and potassium are obtained respectively. To obtain the comprehensive straw nutrient release rate, the corresponding weights are determined based on the proportion of the initial total amount of each target nutrient element to the sum of the initial total amounts of all target nutrient elements.

[0026] The weight of the target nutrient element = the initial total amount of the target nutrient element ÷ the sum of the initial total amounts of all target nutrient elements.

[0027] For example, if the initial total amounts of the target nutrients for nitrogen, phosphorus, and potassium are 12g, 3g, and 20g, respectively, then their corresponding weights are 12÷(12+3+20)≈34.29%, 3÷(12+3+20)≈8.57%, and 20÷(12+3+20)≈57.14%. When the straw nutrient release rates for nitrogen, phosphorus, and potassium at the current monitoring time point are 31.75%, 28%, and 35%, respectively, the comprehensive straw nutrient release rate = 31.75%×34.29%+28%×8.57%+35%×57.14%=32.92%.

[0028] The comprehensive straw nutrient release rate at multiple monitoring time points was obtained in the same manner, and a time series of straw nutrient release rate was formed in chronological order.

[0029] Furthermore, the soil temperature and soil moisture content at the current monitoring time are obtained. The soil temperature is the measured value of the target field at the current monitoring time, and the soil moisture content is the measured value of the target field at the current monitoring time. These data are obtained through soil temperature sensors and soil moisture sensors installed within the target field.

[0030] When multiple monitoring locations are set up for the target field, the average value of the effective soil temperature measurement and effective soil moisture measurement obtained from each monitoring location is calculated to obtain the soil temperature and soil moisture content at the current monitoring time.

[0031] For example, the soil temperatures measured at five monitoring locations were 22.1℃, 22.8℃, 22.5℃, 23.0℃, and 22.6℃, respectively. Therefore, the soil temperature = (22.1 + 22.8 + 22.5 + 23.0 + 22.6) ÷ 5 = 22.6℃. The corresponding soil moisture contents were 24.2%, 25.1%, 24.6%, 25.3%, and 24.8%, respectively. Therefore, the soil moisture content = (24.2% + 25.1% + 24.6% + 25.3% + 24.8%) ÷ 5 = 24.8%.

[0032] Furthermore, obtain the straw return method identifier for the target field. The straw return method identifier is a classification identifier for the straw return method used in the target field, including the straw crushing return method identifier, the straw covering return method identifier, and the straw plowing and burying return method identifier.

[0033] Specifically, when the target field uses a crushing method to process the straw before returning it to the field, the corresponding method of returning the straw to the field is identified as the crushing and returning method identifier; when the target field lays the straw on the soil surface, the corresponding method of returning the straw to the field is identified as the covering and returning method identifier; when the target field buries the straw in the soil through plowing, the corresponding method of returning the straw to the field is identified as the plowing and burying and returning method identifier.

[0034] Furthermore, baseline values ​​for soil organic matter and available nutrients in the target field were obtained. Specifically, the soil organic matter baseline value was the soil organic matter content measured before straw return to the field, and the soil available nutrient baseline value was the content of available nitrogen, available phosphorus, and available potassium measured before straw return to the field.

[0035] Specifically, before returning straw to the field, multiple soil sampling locations were set up in the target field, and soil samples were collected at the same soil depth. The contents of soil organic matter, soil available nitrogen, available phosphorus and available potassium were measured respectively. The average value of the effective measured values ​​at each sampling location was calculated to obtain the corresponding baseline values ​​of soil organic matter and soil available nutrients.

[0036] Baseline value of soil organic matter = sum of soil organic matter content values ​​at each valid sampling location ÷ number of valid sampling locations.

[0037] Baseline value of available nitrogen in soil = sum of available nitrogen content values ​​at each valid sampling location ÷ number of valid sampling locations.

[0038] Baseline value of available phosphorus in soil = sum of available phosphorus content values ​​at each valid sampling location ÷ number of valid sampling locations.

[0039] Baseline value of available potassium in soil = sum of available potassium content values ​​at each valid sampling location ÷ number of valid sampling locations.

[0040] For example, the soil organic matter content measured at the five sampling locations was 18.2 g / kg, 17.6 g / kg, 18.9 g / kg, 19.1 g / kg and 18.7 g / kg, respectively. Then the baseline value of soil organic matter = (18.2 + 17.6 + 18.9 + 19.1 + 18.7) ÷ 5 = 18.5 g / kg. The available nitrogen content in the soil measured at the five sampling locations were 92 mg / kg, 88 mg / kg, 95 mg / kg, 90 mg / kg, and 85 mg / kg, respectively. Therefore, the baseline value of available nitrogen in the soil is (92 + 88 + 95 + 90 + 85) ÷ 5 = 90 mg / kg. The available phosphorus content was 21 mg / kg, 19 mg / kg, 22 mg / kg, 20 mg / kg, and 18 mg / kg, respectively. Therefore, the baseline value of available phosphorus in the soil is (21 + 19 + 22 + 20 + 18) ÷ 5 = 20 mg / kg. The available potassium content was 125 mg / kg, 132 mg / kg, 128 mg / kg, 120 mg / kg, and 130 mg / kg, respectively. Therefore, the baseline value of available potassium in the soil is (125 + 132 + 128 + 120 + 130) ÷ 5 = 127 mg / kg.

[0041] In this embodiment, the number of soil sampling locations is determined based on the target field area and the degree of spatial variability of the soil. Sampling locations are gradually increased through pre-sampling, and the average values ​​of soil organic matter content, available nitrogen content, available phosphorus content, and available potassium content are calculated for different sampling numbers. When the maximum relative change rate of each average value is less than 5% after increasing the sampling locations, the corresponding sampling number is determined as the official sampling number.

[0042] For example, pre-sampling calculations were performed on the target field using 4, 5, and 6 sampling locations respectively. When the relative change rate of the average values ​​of soil organic matter content, soil available nitrogen content, available phosphorus content, and available potassium content was less than 5% after increasing the number of sampling locations from 5 to 6, the 5 sampling locations were determined as the formal sampling quantity.

[0043] In summary, this step obtains the time series of straw decomposition rate, straw nutrient release rate, soil temperature, soil moisture content, return method identification, soil organic matter baseline value, and soil available nutrient baseline value of the target field, forming the return process and basic soil data of the target field. By standardizing monitoring conditions and sampling quantity calibration, the consistency and representativeness of the data are improved, providing a data foundation for subsequent straw decomposition process analysis and ecological effect prediction.

[0044] S200: Based on the time series of straw decomposition rate and straw nutrient release rate, analyze and obtain the straw decomposition process index of the target field at the current monitoring time; Step S200 in the method provided in this embodiment of the invention includes: For every two adjacent monitoring time nodes in the straw decomposition rate time series, the straw decomposition rate value of the later monitoring time node is subtracted from the straw decomposition rate value of the earlier monitoring time node, and the difference is divided by the time interval between adjacent monitoring time nodes to obtain the decomposition rate of adjacent monitoring time nodes. The average rate of decay progression is obtained by taking the arithmetic mean of the decay progression rates of all adjacent monitoring time points. For every two adjacent monitoring time nodes in the straw nutrient release rate time series, the straw nutrient release rate value of the later monitoring time node is subtracted from the straw nutrient release rate value of the earlier monitoring time node, and the difference is divided by the time interval between adjacent monitoring time nodes to obtain the nutrient release advance rate of adjacent monitoring time nodes. The nutrient release propulsion rate is obtained by taking the arithmetic mean of the nutrient release propulsion rates of all adjacent monitoring time points. Divide the average decomposition rate by the preset decomposition rate reference value to obtain the decomposition rate normalization value, and divide the average nutrient release rate by the preset nutrient release rate reference value to obtain the nutrient release rate normalization value. The Pearson correlation coefficients of the time series of straw decomposition rate and straw nutrient release rate were calculated to obtain the synchronicity of decomposition and release. Calculate the arithmetic mean of the normalized values ​​of the decomposition rate and the nutrient release rate, and multiply it by the decomposition-release synchronicity to obtain the straw decomposition progress index of the target field at the current monitoring time.

[0045] In this embodiment, after obtaining the time series of straw decomposition rate and straw nutrient release rate of the target field, the straw decomposition advancement rate and nutrient release advancement rate are determined according to the changes between adjacent monitoring time nodes, and further combined with the degree of synchronization of the changes of the two, the straw decomposition process index of the target field at the current monitoring time is obtained.

[0046] Specifically, for adjacent monitoring time nodes in the straw decomposition rate time series, the corresponding decomposition rate is calculated: Decomposition rate = (straw decomposition rate at the next monitoring time point - straw decomposition rate at the previous monitoring time point) ÷ time interval between adjacent monitoring time points.

[0047] For example, the straw decomposition rate time series of the target field includes four consecutive monitoring time nodes, with corresponding straw decomposition rates of 10%, 22%, 39% and 52%, respectively. The time interval between adjacent monitoring time nodes is 15 days, and the corresponding decomposition progress rates are (22%-10%)÷15=0.80% / day, (39%-22%)÷15=1.13% / day and (52%-39%)÷15=0.87% / day, respectively.

[0048] Furthermore, the average decay rate is calculated based on the decay rate at each adjacent monitoring time point: Average rate of decay = Sum of decay rates at all adjacent monitoring time points ÷ Number of adjacent monitoring time points.

[0049] For example, for the target field mentioned above, the average rate of decomposition is (0.80 + 1.13 + 0.87) ÷ 3 = 0.93% / day.

[0050] Furthermore, for adjacent monitoring time points in the straw nutrient release rate time series, the corresponding nutrient release propulsion rate is calculated: Nutrient release rate = (Straw nutrient release rate at the next monitoring time point - Straw nutrient release rate at the previous monitoring time point) ÷ Time interval between adjacent monitoring time points.

[0051] For example, if the straw nutrient release rates corresponding to the above monitoring time points are 8%, 20%, 35% and 50%, then the corresponding nutrient release promotion rates are (20%-8%)÷15=0.80% / day, (35%-20%)÷15=1.00% / day and (50%-35%)÷15=1.00% / day, respectively.

[0052] Furthermore, the average nutrient release propulsion rate is calculated based on the nutrient release propulsion rates at each adjacent monitoring time point: Average nutrient release advance rate = Sum of nutrient release advance rates of all adjacent monitoring time points ÷ Number of adjacent monitoring time points.

[0053] In the example above, the average nutrient release rate is (0.80 + 1.00 + 1.00) ÷ 3 = 0.93% / day.

[0054] Furthermore, in order to eliminate the order of magnitude difference between the average rate of decomposition and the average rate of nutrient release, normalization was performed using preset reference values ​​for decomposition rate and nutrient release rate, respectively, to obtain normalized values ​​for decomposition rate and nutrient release rate.

[0055] Normalized corrosion rate = Average corrosion rate ÷ Preset corrosion rate reference value.

[0056] Normalized nutrient release rate = Average nutrient release rate ÷ Preset nutrient release rate reference value.

[0057] In this embodiment, both the decomposition rate reference value and the nutrient release rate reference value are calibrated based on historical monitoring data. Specifically, monitoring data from historical fields with the same or similar crop type, straw type, and return method to the target field are selected. The average decomposition rate and average nutrient release rate corresponding to each historical field are calculated, and the median of the corresponding historical data is determined as the decomposition rate reference value and the nutrient release rate reference value, respectively.

[0058] For example, based on historical monitoring data, the median average rate of decomposition is 1.163% / day, and the median average rate of nutrient release is 1.163% / day. Therefore, 1.163% / day and 1.08% / day are determined as reference values ​​for decomposition rate and nutrient release rate, respectively. For the target field mentioned above, the average rate of decomposition is 0.93% / day, and the average rate of nutrient release is 0.93% / day. Therefore, the normalized value of decomposition rate = 0.93 ÷ 1.163 = 0.80, and the normalized value of nutrient release rate = 0.93 ÷ 1.08 ≈ 0.86.

[0059] Furthermore, the Pearson correlation coefficient between the time series of straw decomposition rate and the time series of straw nutrient release rate was calculated, and the synchronicity of decomposition and release was determined based on the Pearson correlation coefficient.

[0060] Specifically, the straw decomposition rate and straw nutrient release rate values ​​corresponding to the same monitoring time node are time-paired, and the Pearson correlation coefficient between the two time series is calculated. When the Pearson correlation coefficient is greater than or equal to 0, the Pearson correlation coefficient is determined as the decomposition and release synchronization degree; when the Pearson correlation coefficient is less than 0, the decomposition and release synchronization degree is set to 0.

[0061] For example, for the above four monitoring time points, the straw decomposition rates are 10%, 22%, 39% and 52%, respectively, and the corresponding straw nutrient release rates are 8%, 20%, 35% and 50%, respectively. The calculated Pearson correlation coefficient is 0.998, so the decomposition and release synchronization degree is 0.998; if the calculated Pearson correlation coefficient is -0.20, then the decomposition and release synchronization degree is set to 0.

[0062] Furthermore, the straw decomposition process index was calculated based on the normalized values ​​of the decomposition rate, the normalized values ​​of the nutrient release rate, and the synchronicity of decomposition and release. Straw decomposition process index = (normalized value of decomposition rate + normalized value of nutrient release rate) ÷ 2 × decomposition and release synchronicity.

[0063] For example, taking the above-mentioned normalized value of decomposition rate of 0.80, normalized value of nutrient release rate of 0.86, and decomposition-release synchronization degree of 0.998 as examples, the straw decomposition process index = (0.80 + 0.86) ÷ 2 × 0.998 ≈ 0.83.

[0064] In this embodiment, when the normalized value of the decomposition rate or the normalized value of the nutrient release rate is greater than 1, the corresponding normalized value is set to 1 to keep the straw decomposition progress index within the range of 0 to 1. The closer the straw decomposition progress index is to 1, the higher the degree of decomposition and nutrient release progress in the target field, and the higher the degree of synchronization between the two changes over time; the lower the straw decomposition progress index, the lower the degree of progress of decomposition and nutrient release, or the lower the degree of synchronization between the two changes.

[0065] In summary, this step calculates the decomposition and nutrient release rates at adjacent monitoring time points, normalizes them using reference values, and then combines the Pearson correlation coefficient between the straw decomposition rate time series and the straw nutrient release rate time series to obtain the straw decomposition process index, thereby comprehensively characterizing the degree of advancement and synchronization of straw decomposition and nutrient release.

[0066] S300: Based on the straw decomposition process index, soil temperature and soil moisture content, the decomposition response lag coefficient of the target field is obtained through a pre-constructed lag assessment model. Step S300 in the method provided in this embodiment of the invention includes: Obtain a pre-constructed hysteresis assessment model. The hysteresis assessment model is a regression model that takes straw decomposition process index, soil temperature and soil moisture content as inputs and outputs decomposition response hysteresis coefficient. Input the straw decomposition process index, soil temperature, and soil moisture content of the target field at the current monitoring time into the hysteresis assessment model, and obtain the decomposition response hysteresis coefficient of the target field at the current monitoring time from the output of the hysteresis assessment model.

[0067] In this embodiment, after obtaining the straw decomposition process index, soil temperature and soil moisture content of the target field at the current monitoring time, a pre-constructed hysteresis evaluation model is obtained, and the above data is input into the hysteresis evaluation model to obtain the decomposition response hysteresis coefficient of the target field at the current monitoring time.

[0068] Specifically, a pre-constructed hysteresis assessment model is obtained. The hysteresis assessment model is a regression model that takes straw decomposition process index, soil temperature, and soil moisture content as input and outputs a hysteresis coefficient of decomposition response. The hysteresis assessment model is used to establish the mapping relationship between the straw decomposition process index, soil temperature, and soil moisture content and the hysteresis coefficient of decomposition response.

[0069] Furthermore, the straw decomposition process index, soil temperature, and soil moisture content of the target field at the current monitoring time are input into the hysteresis assessment model to obtain the decomposition response hysteresis coefficient output by the hysteresis assessment model.

[0070] For example, the straw decomposition process index obtained at the current monitoring time is 0.83, the soil temperature is 22.6℃, and the soil moisture content is 24.8%. The straw decomposition process index of 0.83, the soil temperature of 22.6℃, and the soil moisture content of 24.8% are input into the hysteresis evaluation model. The decomposition response hysteresis coefficient output by the hysteresis evaluation model is 1.35. Then, 1.35 is determined as the decomposition response hysteresis coefficient of the target field at the current monitoring time.

[0071] In this embodiment, the decomposition response hysteresis coefficient is a dimensionless value, greater than zero, used to reflect the degree of time lag in soil property changes caused by straw decomposition products entering the soil. The larger the decomposition response hysteresis coefficient, the more obvious the time lag in soil property changes caused by straw decomposition products entering the soil; the closer the decomposition response hysteresis coefficient is to zero, the lower the corresponding time lag.

[0072] Furthermore, the corresponding straw decomposition process index, soil temperature, and soil moisture content were obtained at different monitoring times and input into the hysteresis assessment model to obtain the decomposition response hysteresis coefficient at the corresponding monitoring time, thereby obtaining the change in decomposition response hysteresis of the target field during the straw return process.

[0073] Step S300 in the method provided in this embodiment of the invention further includes: Historical monitoring data of multiple historical monitoring plots were obtained. The historical monitoring data included the straw decomposition process index, soil temperature, soil moisture content, and decomposition response hysteresis coefficient labels of each historical monitoring plot at each historical monitoring time. A hysteresis evaluation model is constructed based on a backpropagation neural network model. The hysteresis evaluation model includes an input layer, a hidden layer, and an output layer. The input layer contains three input nodes, which correspond to the straw decomposition process index, soil temperature, and soil moisture content, respectively. The output layer contains one output node, which outputs the decomposition response hysteresis coefficient. Using the straw decomposition process index, soil temperature, and soil moisture content at each historical monitoring time as input features, and the corresponding decomposition response hysteresis coefficient label as output label, the training objective is to minimize the mean square error between the decomposition response hysteresis coefficient output by the hysteresis evaluation model and the decomposition response hysteresis coefficient label. The parameters of the hysteresis evaluation model are iteratively updated until the mean square error converges, thus obtaining the trained hysteresis evaluation model.

[0074] In this embodiment, historical monitoring data of multiple historical monitoring plots are obtained. The historical monitoring data includes the straw decomposition process index, soil temperature, soil moisture content, and corresponding decomposition response hysteresis coefficient labels of each historical monitoring plot at each historical monitoring time.

[0075] Among them, the straw decomposition process index, soil temperature, soil moisture content and decomposition response hysteresis coefficient labels at the same historical monitoring time are used as a set of corresponding historical monitoring data.

[0076] For example, historical monitoring data from 20 historical monitoring plots at multiple historical monitoring times were obtained, resulting in 1000 sets of valid historical monitoring data. 80% of the historical monitoring data was randomly selected as training data, and the remaining 20% ​​was used as validation data. The training data was used to update the model parameters of the hysteresis evaluation model, while the validation data was used to determine the network structure and training parameters of the hysteresis evaluation model and to evaluate the model prediction error after training.

[0077] Furthermore, a hysteresis evaluation model is constructed based on a backpropagation neural network model. The hysteresis evaluation model includes an input layer, a hidden layer, and an output layer. The input layer contains three input nodes, corresponding to the straw decomposition process index, soil temperature, and soil moisture content, respectively; the hidden layer is used to nonlinearly combine the above three input features; the output layer contains one output node, which is used to output the decomposition response hysteresis coefficient.

[0078] In this embodiment, the number of hidden layer nodes is calibrated based on the mean squared error of the validation data. With the number of input layer nodes remaining constant at 3 and the number of output layer nodes remaining constant at 1, the number of hidden layer nodes is set to 4–10, and corresponding hysteresis evaluation models are constructed using 4, 5, 6, 7, 8, 9, and 10 hidden nodes, respectively. The above hysteresis evaluation models are trained using the same training data, and the mean squared error of each hysteresis evaluation model is calculated using the validation data. The number of hidden layer nodes with the smallest mean squared error of the validation data is determined as the final number of hidden layer nodes.

[0079] For example, when the number of hidden layer nodes is set to 4, 5, 6, 7, 8, 9 and 10 respectively, the corresponding mean square errors of the validation data are 0.083, 0.071, 0.061, 0.058, 0.064, 0.067 and 0.072 respectively. The mean square error corresponding to 7 hidden nodes is the smallest. Therefore, the number of hidden layer nodes is determined to be 7, thus obtaining the 3-7-1 hysteresis evaluation model structure.

[0080] Furthermore, the straw decomposition process index, soil temperature, and soil moisture content from the training data are used as input features, and the corresponding decomposition response hysteresis coefficient labels are used as output labels to train a hysteresis evaluation model with a defined network structure. Specifically, the connection weights and bias parameters in the backpropagation neural network serve as model parameters iteratively updated during training. Based on the error between the decomposition response hysteresis coefficient output by the model and its corresponding label, the model parameters are updated via backpropagation.

[0081] For each set of training data, the corresponding straw decomposition process index, soil temperature, and soil moisture content are input into the hysteresis evaluation model. After processing through the input layer, hidden layer, and output layer, the predicted value of the corresponding decomposition response hysteresis coefficient is obtained. The mean square error between the predicted value of the decomposition response hysteresis coefficient and the corresponding decomposition response hysteresis coefficient label is used as the model training objective. Mean square error = .

[0082] For example, in a certain training session, four sets of training data are obtained, with predicted values ​​of the corrosion response hysteresis coefficients of 1.18, 1.43, 0.92, and 1.67, respectively, and corresponding labels of the corrosion response hysteresis coefficients of 1.20, 1.50, 0.85, and 1.60, respectively. Then, the mean squared error = [(1.18-1.20)]. 2 +(1.43-1.50) 2 +(0.92-0.85) 2 +(1.67-1.60) 2 ] ÷ 4 = 0.00318.

[0083] Furthermore, based on the mean square error, the model parameters of the hysteresis evaluation model are iteratively updated through backpropagation, so that the corrosion response hysteresis coefficient output by the model gradually approaches the corresponding corrosion response hysteresis coefficient label.

[0084] In this embodiment, the learning rate used for model training is calibrated based on the mean squared error of the validation data. Specifically, 0.001, 0.005, 0.01, 0.05, and 0.1 are used as candidate learning rates, respectively. The same training data is used for training, and the mean squared error corresponding to each candidate learning rate is calculated using the validation data. The candidate learning rate with the lowest mean squared error of the validation data is determined as the learning rate used for model training.

[0085] For example, the mean squared errors of the validation data corresponding to the above 5 candidate learning rates are 0.072, 0.065, 0.061, 0.068 and 0.093, respectively. Therefore, 0.01 is determined as the learning rate of the hysteresis evaluation model.

[0086] Furthermore, the training data is iteratively trained using a defined network structure and training parameters until the mean square error of the hysteresis evaluation model reaches a convergent state, thus obtaining the trained hysteresis evaluation model.

[0087] In this embodiment, the mean square error change rate of 10 consecutive training iterations is used as the convergence criterion. When the mean square error change rate of 10 consecutive training iterations is lower than 0.001, the mean square error is determined to be converged and training is stopped.

[0088] The rate of change of mean square error = |mean square error of the current training round - mean square error of the previous training round| ÷ mean square error of the previous training round.

[0089] The mean squared error rate of change threshold of 0.001 was calibrated based on historical model training data. Specifically, the mean squared error rate of change after entering the stable phase during the training of multiple historical hysteresis evaluation models was statistically analyzed. The 95th percentile of the mean squared error rate of change was used as the convergence reference value, and the 95th percentile was rounded up to the thousandth percentile as the mean squared error rate of change convergence threshold.

[0090] For example, the 95th percentile value of the mean square error change rate in the stable phase obtained by statistics is 0.00082. After rounding up to the thousandth, it is 0.001. Therefore, 0.001 is determined as the convergence threshold of the mean square error change rate in this embodiment.

[0091] Furthermore, the trained hysteresis assessment model is input into the straw decomposition process index, soil temperature, and soil moisture content in the validation data to obtain the corresponding predicted values ​​of the decomposition response hysteresis coefficients. These values ​​are then compared with the corresponding decomposition response hysteresis coefficient labels to determine the prediction error of the trained hysteresis assessment model on data that have not been updated with model parameters.

[0092] For example, after training, the mean square error between the predicted value of the decomposition response hysteresis coefficient corresponding to the validation data and the label of the decomposition response hysteresis coefficient is 0.063, indicating that the trained hysteresis evaluation model can output the corresponding decomposition response hysteresis coefficient according to the straw decomposition process index, soil temperature and soil moisture content.

[0093] After training is complete, the network structure of the hysteresis evaluation model and the trained model parameters are saved to obtain the trained hysteresis evaluation model.

[0094] Step S300 in the method provided in this embodiment of the invention further includes: For each historical monitoring field, obtain its straw decomposition rate time series and soil organic matter content time series; From the time series of straw decomposition rate, the moment when the straw decomposition rate first reaches the preset intermediate decomposition rate is determined as the intermediate decomposition moment; Obtain the soil organic matter content at the mid-decomposition stage as the baseline value for mid-decomposition organic matter; Starting from the mid-decomposition period, the soil organic matter content at each subsequent monitoring time point is checked sequentially. When the soil organic matter content at a consecutive preset number of monitoring time points is higher than the mid-decomposition organic matter baseline value, the first monitoring time point among the consecutive preset number of monitoring time points is taken as the soil response start time. The difference between the soil response start time and the mid-decomposition time is used as the response lag time for historical monitoring plots. Divide the response lag time by the preset reference decomposition cycle to obtain the decomposition response lag coefficient label of the historical monitoring plot.

[0095] In this embodiment, in order to make the historical training data of the hysteresis assessment model more applicable to the target field to be predicted, historical monitoring fields with similar straw return conditions and soil environmental conditions to the target field are selected from multiple candidate historical monitoring fields as historical sample fields for constructing the hysteresis assessment model.

[0096] For each candidate historical monitoring plot, the corresponding straw return conditions and soil environmental conditions were obtained and compared with the corresponding conditions of the target plot. When the candidate historical monitoring plot and the target plot have the same straw type and straw return method, the amount of straw returned is within the preset similarity range, and the soil type is the same or belongs to the preset similar soil type, and the soil organic matter content, soil basic nutrient level, soil temperature and soil moisture content before straw return are within the corresponding preset similarity range, the candidate historical monitoring plot is determined as a historical sample plot.

[0097] For each selected historical sample field, time series of straw decomposition rate and soil organic matter content were obtained. The time series of straw decomposition rate and soil organic matter content were recorded using the same historical monitoring time points, establishing a time correspondence between the straw decomposition rate and the corresponding soil organic matter content at each historical monitoring time point.

[0098] Specifically, the monitoring time point at which the straw decomposition rate first reaches the preset intermediate decomposition rate is determined from the straw decomposition rate time series, and this monitoring time point is defined as the intermediate decomposition moment. The preset intermediate decomposition rate is used to divide the intermediate stage of the straw decomposition process, so as to avoid directly using the rapid decomposition fluctuations at the beginning of returning straw to the field as the initial judgment benchmark for soil response.

[0099] In this embodiment, the preset mid-term decomposition rate is calibrated based on historical monitoring data. Specifically, multiple candidate mid-term decomposition rates are set within the preset candidate decomposition rate range. The mid-term decomposition time of each historical monitoring plot is determined based on each candidate mid-term decomposition rate, and the corresponding response lag duration is further determined. The coefficient of variation of the response lag duration corresponding to each candidate mid-term decomposition rate is used as an evaluation index, and the candidate mid-term decomposition rate with the smallest coefficient of variation of the response lag duration is determined as the preset mid-term decomposition rate.

[0100] For example, within the range of 30% to 70% of the candidate decomposition rates, candidate intermediate decomposition rates of 40%, 50% and 60% are set in increments of 10 percentage points for historical data verification. The corresponding response lag time variation coefficients are 0.31, 0.18 and 0.24, respectively. Then, 50% is determined as the preset intermediate decomposition rate.

[0101] Furthermore, the soil organic matter content at the mid-decomposition stage was obtained, and this soil organic matter content was determined as the baseline value of mid-decomposition organic matter for the corresponding historical monitoring plot. This baseline value was used as a comparison benchmark for judging the subsequent soil organic matter response.

[0102] Subsequently, starting from the mid-decomposition stage, the soil organic matter content at each subsequent monitoring time point was checked sequentially. When the soil organic matter content at a consecutive preset number of monitoring time points is higher than the mid-decomposition organic matter baseline value, the first monitoring time point among these consecutive preset number of monitoring time points is determined as the soil response start time.

[0103] In this embodiment, the preset quantity is determined based on historical monitoring data. Specifically, multiple candidate quantities are set within the preset candidate quantity range, and the coefficient of variation of response lag time for each candidate quantity corresponding to the historical monitoring plots is calculated. The candidate quantity with the smallest coefficient of variation of response lag time is determined as the preset quantity.

[0104] For example, within the candidate number range of 2 to 4 monitoring time nodes, verification is performed with 1 monitoring time node as the step size. The coefficients of variation of response lag time corresponding to 2, 3 and 4 consecutive monitoring time nodes are 0.23, 0.16 and 0.17, respectively. Then, 3 monitoring time nodes are determined as the preset number.

[0105] Furthermore, the time interval between the soil response initiation time and the mid-decomposition time is defined as the response lag time: Response lag time = Soil response start time - mid-decomposition time.

[0106] For example, the straw decomposition rate of a historically monitored field first reached 50% at a certain monitoring time point. Therefore, this monitoring time point was determined as the mid-decomposition point, and the corresponding soil organic matter content was 18.6 g / kg. Subsequent monitoring time points were checked chronologically. When the soil organic matter content at three consecutive monitoring time points was 18.7 g / kg, 18.9 g / kg, and 19.1 g / kg, respectively, all higher than 18.6 g / kg, the first of these monitoring time points was determined as the soil response initiation point. If the time interval between the mid-decomposition point and the soil response initiation point is 15 days, then the response lag time is 15 days.

[0107] Furthermore, by dividing the response lag time by a preset reference decomposition cycle, the decomposition response lag coefficient label for the corresponding historical monitoring plot is obtained: Corrosion response hysteresis coefficient label = response hysteresis duration ÷ preset reference corrosion cycle.

[0108] In this embodiment, the preset reference decomposition period is calibrated based on the time series of straw decomposition rates of historical monitoring plots. Specifically, the time elapsed from the start of straw return to the first time the straw decomposition rate reaches the preset intermediate decomposition rate for each historical monitoring plot is statistically analyzed. These times are arranged in ascending order, and the median is taken as the preset reference decomposition period.

[0109] For example, the time taken for historically monitored fields to reach a 50% straw decomposition rate for the first time from the start of straw return to the field was 40 days, 42 days, 45 days, 48 ​​days, and 50 days, respectively. The median was 45 days, so 45 days was determined as the preset reference decomposition period. For the aforementioned historically monitored fields, the response lag time was 15 days, and the preset reference decomposition period was 45 days. Therefore, the decomposition response lag coefficient label = 15 ÷ 45 = 0.33.

[0110] The larger the decomposition response hysteresis coefficient label, the longer the time from the mid-stage of straw decomposition to the sustained response of soil organic matter compared to the reference decomposition cycle; the smaller the decomposition response hysteresis coefficient label, the faster the soil organic matter response.

[0111] In this embodiment, the hysteresis coefficient label of each historical monitoring field is associated with the straw decomposition process index, soil temperature and soil moisture content of the corresponding historical monitoring time node, and a data correspondence is established to form the historical training data of the hysteresis assessment model.

[0112] In summary, this step constructs a label for the lag coefficient of decomposition response and trains a lag assessment model using historical monitoring data. This establishes a correlation between the straw decomposition process, soil temperature, soil moisture content, and the degree of lag in the soil ecological response, enabling dynamic quantification of the current degree of lag in the decomposition response. On one hand, by determining the soil response start time through continuous monitoring nodes, the impact of fluctuations in single monitoring sessions on the determination of the lag label is reduced. On the other hand, a dimensionless lag coefficient is formed by normalizing the response lag time, and a backpropagation neural network is used to learn the nonlinear relationship between multi-source environmental factors and the degree of lag. This improves the accuracy of obtaining the decomposition response lag coefficient and provides a reliable basis for subsequent lag correction of changes in soil organic matter and available nutrients.

[0113] S400: Based on the decomposition response hysteresis coefficient, and using the baseline values ​​of soil organic matter and available soil nutrients, determine the corrected changes in soil organic matter and available soil nutrients in the target field at the current monitoring time, and form the corrected ecological effect prediction results. Step S400 in the method provided in this embodiment of the invention includes: Obtain the measured value of soil organic matter at the current monitoring time, and subtract the baseline value of soil organic matter from the measured value of soil organic matter to obtain the original change value of soil organic matter; Obtain the measured values ​​of available nitrogen, available phosphorus, and available potassium in the soil at the current monitoring time. Subtract the corresponding baseline values ​​of available nutrients from the measured values ​​of available nitrogen, available phosphorus, and available potassium in the soil, and then sum them to obtain the original change values ​​of available nutrients in the soil. Obtain the decomposition response hysteresis coefficient of the target field at the current monitoring time, and obtain the preset organic matter hysteresis sensitivity coefficient and available nutrient hysteresis sensitivity coefficient; Calculate the organic matter hysteresis correction factor based on the decomposition response hysteresis coefficient and the organic matter hysteresis sensitivity coefficient; Multiply the original change value of soil organic matter by the organic matter lag correction factor to obtain the corrected change value of soil organic matter. Calculate the lag correction factor for available nutrients based on the lag coefficient of decomposition response and the lag sensitivity coefficient of available nutrients. Multiply the original change value of available soil nutrients by the available nutrient lag correction factor, and add the corrected change value of soil organic matter by the preset interaction coefficient to obtain the corrected change value of available soil nutrients. The revised changes in soil organic matter and available soil nutrients constitute the revised ecological effect prediction results.

[0114] In this embodiment, based on obtaining the decomposition response hysteresis coefficient, soil organic matter baseline value, and soil available nutrient baseline value of the target field at the current monitoring time, the measured values ​​of soil organic matter, soil available nitrogen, available phosphorus, and available potassium at the current monitoring time are further obtained, and the changes in soil organic matter and soil available nutrient at the current monitoring time are corrected according to the decomposition response hysteresis coefficient.

[0115] Specifically, the original change in soil organic matter is obtained by subtracting the baseline value from the measured value at the current monitoring time: Original change value of soil organic matter = measured value of soil organic matter at the current monitoring time - baseline value of soil organic matter.

[0116] Subtract the corresponding baseline values ​​of available soil nutrients from the measured values ​​of soil available nitrogen, available phosphorus, and available potassium at the current monitoring time, and sum the differences to obtain the original changes in available soil nutrients: Original change value of available soil nutrients = (measured value of available soil nitrogen at the current monitoring time - baseline value of available soil nitrogen) + (measured value of available phosphorus at the current monitoring time - baseline value of available soil phosphorus) + (measured value of available potassium at the current monitoring time - baseline value of available soil potassium).

[0117] Furthermore, the decomposition response lag coefficient of the target field at the current monitoring time is obtained, along with preset organic matter lag sensitivity coefficients and available nutrient lag sensitivity coefficients. The organic matter lag sensitivity coefficient is used to characterize the degree of influence of decomposition response lag on soil organic matter changes, while the available nutrient lag sensitivity coefficient is used to characterize the degree of influence of decomposition response lag on soil available nutrient changes.

[0118] The lag sensitivity coefficients for organic matter and available nutrients were calibrated based on historical monitoring data. Specifically, candidate value ranges and candidate step sizes were set for the lag sensitivity coefficients for organic matter and available nutrients, respectively. Within the corresponding candidate value ranges, the candidate coefficients were gradually changed according to the candidate step size, and the mean absolute error corresponding to each candidate coefficient was calculated based on historical monitoring data. The candidate coefficient with the smallest mean absolute error was determined as the corresponding lag sensitivity coefficient.

[0119] In this embodiment, the candidate value ranges for both the organic matter lag sensitivity coefficient and the readily available nutrient lag sensitivity coefficient are set to 0.1–1.0, and the candidate step size is set to 0.01. After calibration using historical monitoring data, the organic matter lag sensitivity coefficient is determined to be 0.60, and the readily available nutrient lag sensitivity coefficient is determined to be 0.58.

[0120] Furthermore, the organic matter hysteresis correction factor is calculated based on the decomposition response hysteresis coefficient and the organic matter hysteresis sensitivity coefficient: Organic matter hysteresis correction factor = 1 ÷ (1 + decomposition response hysteresis coefficient × organic matter hysteresis sensitivity coefficient).

[0121] Among them, the organic matter lag correction factor decreases as the lag coefficient of the decomposition response increases, so that when the lag in the decomposition response is large, the original changes in soil organic matter are more strongly corrected by lag.

[0122] Furthermore, the original change value of soil organic matter is multiplied by the organic matter lag correction factor to obtain the corrected change value of soil organic matter: Corrected soil organic matter change value = original soil organic matter change value × organic matter lag correction factor.

[0123] For example, if the baseline soil organic matter value of the target field is 18.5 g / kg and the measured soil organic matter value at the current monitoring time is 19.3 g / kg, then the original change in soil organic matter is 0.8 g / kg. When the hysteresis response lag coefficient at the current monitoring time is 0.33 and the organic matter lag sensitivity coefficient is 0.6, the organic matter lag correction factor = 1 ÷ (1 + 0.33 × 0.6) ≈ 0.835. Therefore, the corrected change in soil organic matter = 0.8 × 0.835 ≈ 0.668 g / kg.

[0124] Furthermore, the available nutrient hysteresis correction factor is calculated based on the decomposition response hysteresis coefficient and the available nutrient hysteresis sensitivity coefficient: The correction factor for the lag in readily available nutrients is 1 ÷ (1 + lag coefficient for decomposition response × lag sensitivity coefficient for readily available nutrients).

[0125] The original change value of available soil nutrients is multiplied by the available nutrient lag correction factor, and then the product of the corrected change value of soil organic matter and the preset interaction coefficient is added to obtain the corrected change value of available soil nutrients. Corrected change in available soil nutrients = Original change in available soil nutrients × Available nutrient lag correction factor + Corrected change in soil organic matter × Interaction coefficient.

[0126] The interaction coefficient, measured in mg / g, is used to convert the corrected soil organic matter change value into the same unit of measurement as the soil available nutrient change value. The interaction coefficient is calibrated based on historical monitoring data. Specifically, a candidate value range and candidate step size are set for the interaction coefficient. Within the candidate value range, the candidate interaction coefficient is gradually changed according to the candidate step size. The mean absolute error of the soil available nutrient change corresponding to each candidate interaction coefficient is calculated based on historical monitoring data. The candidate interaction coefficient with the smallest mean absolute error is determined as the final interaction coefficient. In this embodiment, the candidate value range for the interaction coefficient is set to 0–0.20 mg / g, and the candidate step size is set to 0.01 mg / g. After calibration using historical monitoring data, 0.10 mg / g is determined as the interaction coefficient.

[0127] For example, at the current monitoring time, the measured values ​​of available nitrogen, available phosphorus, and available potassium in the target field are 96 mg / kg, 23 mg / kg, and 134 mg / kg, respectively; the corresponding baseline values ​​of available nutrients are 90 mg / kg, 20 mg / kg, and 127 mg / kg, respectively. Therefore, the original change in available nutrients is (96-90) + (23-20) + (134-127) = 16 mg / kg. When the hysteresis response lag coefficient is 0.33 and the available nutrient lag sensitivity coefficient is 0.58, the available nutrient lag correction factor is 1 ÷ (1 + 0.33 × 0.58) ≈ 0.839. Therefore, with the corrected change in soil organic matter being 0.668 g / kg and the interaction coefficient being 0.10 mg / g, the corrected change in available soil nutrients is approximately 16 × 0.839 + 0.668 × 0.10 ≈ 13.49 mg / kg.

[0128] Furthermore, the revised changes in soil organic matter and soil available nutrients are combined to form the revised ecological effect prediction results.

[0129] In summary, this step obtains the original changes in soil organic matter and available nutrients, introduces a hysteresis coefficient for decomposition response correction, and uses an interaction coefficient to characterize the coupling effect between changes in soil organic matter and available nutrients. This yields corrected changes in soil organic matter and available nutrients, forming an ecological effect prediction result that considers the hysteresis response and carbon nutrient coupling effect, thus improving the accuracy and adaptability of predicting the ecological effects of straw return to the field.

[0130] S500: Based on the revised ecological effect prediction results and the identification of the return-to-field method, generate recommendations for the return-to-field management strategy of the target field, and dynamically adjust the time interval of the next monitoring cycle according to the decomposition response hysteresis coefficient.

[0131] Step S500 in the method provided in this embodiment of the invention includes: When the corrected soil organic matter change value is less than the preset organic matter increase threshold, a management strategy suggestion for returning more organic materials to the field is generated. When the corrected change in available soil nutrients is less than the preset threshold for available nutrients, a suggested management strategy for supplementing available nutrients by returning them to the field is generated. When the corrected soil organic matter change value is less than the preset organic matter increase threshold and the corrected soil available nutrient change value is less than the preset available nutrient threshold, a combined return-to-field management strategy recommendation of increasing organic matter application and supplementing available nutrients is generated. When the corrected soil organic matter change value is greater than or equal to the preset organic matter increase threshold, and the corrected soil available nutrient change value is greater than or equal to the preset available nutrient threshold, a strategy recommendation to maintain the current return-to-field method is generated.

[0132] In this embodiment, after obtaining the corrected ecological effect prediction results and the return-to-field method identifier of the target field at the current monitoring time, a return-to-field management strategy suggestion for the target field is generated based on the comparison results between the corrected soil organic matter change value, the corrected soil available nutrient change value and the corresponding preset threshold.

[0133] Specifically, preset thresholds for organic matter enhancement and available nutrient enhancement are established. The organic matter enhancement threshold is used to characterize whether the change in soil organic matter in the target field after modification has reached the expected level of enhancement, and the available nutrient threshold is used to characterize whether the change in available nutrients in the soil in the target field after modification has reached the expected level of enhancement.

[0134] In this embodiment, the organic matter enhancement threshold and the available nutrient threshold are calibrated based on historical monitoring data. Specifically, the corrected soil organic matter change value and the corrected soil available nutrient change value for multiple historical monitoring plots are obtained, and the historical monitoring data that achieve the expected soil return effect are determined based on the actual soil return management effect of the historical monitoring plots. The distribution of the corrected soil organic matter change value and the corrected soil available nutrient change value are statistically analyzed in the historical monitoring data that achieve the expected soil return effect, and the corresponding preset quantile values ​​are used as candidate organic matter enhancement thresholds and candidate available nutrient thresholds. The values ​​are then corrected based on the historical soil return management results, and the determined candidate values ​​are used as the organic matter enhancement threshold and the available nutrient threshold, respectively.

[0135] For example, based on historical monitoring data, the threshold for increasing organic matter was determined to be 0.50 g / kg, and the threshold for readily available nutrients was determined to be 10 mg / kg.

[0136] Furthermore, the revised soil organic matter change value is compared with the organic matter enhancement threshold, and the revised soil available nutrient change value is compared with the available nutrient threshold. Based on the comparison results, corresponding return-to-field management strategy recommendations are generated.

[0137] When the revised soil organic matter change value is less than the organic matter increase threshold, a management strategy recommendation for applying additional organic materials to the field is generated; when the revised soil available nutrient change value is less than the available nutrient threshold, a management strategy recommendation for supplementing available nutrients to the field is generated.

[0138] When the corrected soil organic matter change value is less than the organic matter increase threshold and the corrected soil available nutrient change value is less than the available nutrient threshold, a combined management strategy of applying additional organic materials and supplementing available nutrients is recommended.

[0139] When the revised soil organic matter change value is greater than or equal to the organic matter increase threshold, and the revised soil available nutrient change value is greater than or equal to the available nutrient threshold, a strategy recommendation to maintain the current return-to-field method is generated.

[0140] For example, when the corrected soil organic matter change value at the current monitoring time of the target field is 0.668 g / kg, the corrected soil available nutrient change value is 13.49 mg / kg, the organic matter enhancement threshold is 0.50 g / kg, and the available nutrient threshold is 10 mg / kg, the corrected soil organic matter change value is greater than or equal to the organic matter enhancement threshold, and the corrected soil available nutrient change value is greater than or equal to the available nutrient threshold. Therefore, a strategy recommendation to maintain the current return-to-field method is generated.

[0141] Furthermore, the generated field return management strategy suggestions correspond to the field return method identifier of the target field and are used as the field return management strategy output at the current monitoring time.

[0142] Step S500 in the method provided in this embodiment of the invention further includes: Obtain the preset baseline monitoring period, which is the monitoring time interval of the target field under normal conditions; The baseline monitoring cycle is corrected using the corrosion response hysteresis coefficient to obtain the time interval of the next monitoring cycle. The time interval of the next monitoring cycle is positively correlated with the corrosion response hysteresis coefficient.

[0143] In this embodiment, based on the revised ecological effect prediction results generated according to the current monitoring time of the target field, the time interval of the next monitoring cycle is further dynamically determined according to the decomposition response lag coefficient at the current monitoring time, so that the monitoring cycle can be adjusted according to the lag degree of straw decomposition response.

[0144] Specifically, a preset baseline monitoring period is obtained, which is the preset time interval between two consecutive monitoring sessions of the target field under normal conditions. In this embodiment, the baseline monitoring period is set based on the normal monitoring period in the historical monitoring data of the target field and determined in combination with the actual monitoring frequency. For example, when the target field uses a 15-day monitoring interval under normal conditions, the baseline monitoring period is set to 15 days.

[0145] Furthermore, the baseline monitoring cycle is corrected based on the corrosion response hysteresis coefficient at the current monitoring moment, so that the larger the corrosion response hysteresis coefficient, the longer the time interval of the next monitoring cycle. Specifically, the time correction is determined by multiplying the baseline monitoring cycle by the corrosion response hysteresis coefficient, and then the time correction is added to the baseline monitoring cycle to obtain the time interval of the next monitoring cycle. The time interval of the next monitoring cycle = baseline monitoring cycle × (1 + cycle adjustment coefficient × corrosion response hysteresis coefficient).

[0146] The period adjustment coefficient is used to control the degree to which the corrosion response hysteresis coefficient adjusts the monitoring period, and is a preset coefficient that is greater than zero.

[0147] In this embodiment, the cycle adjustment coefficient is calibrated based on historical monitoring data. Specifically, the decomposition response hysteresis coefficients and their corresponding actual effective monitoring cycles of multiple historical monitoring plots are obtained. Within a preset candidate value range, the cycle adjustment coefficient is gradually changed according to a preset step size. Based on the average absolute error between the monitoring cycle calculated from each candidate cycle adjustment coefficient and the historical actual effective monitoring cycle, the candidate cycle adjustment coefficient with the smallest average absolute error is determined as the final cycle adjustment coefficient.

[0148] For example, the candidate value range for the periodic adjustment coefficient is set to 0.1–1.0, and the candidate step size is set to 0.01. After calibration with historical monitoring data, 0.5 is determined as the periodic adjustment coefficient. When the decomposition response hysteresis coefficient of the target field at the current monitoring time is 0.33, the baseline monitoring period is 15 days, and the periodic adjustment coefficient is 0.5, the time interval for the next monitoring period is 15 × (1 + 0.5 × 0.33) = 15.99 days. Therefore, the time interval for the next monitoring period can be set to 16 days, and the next monitoring can be conducted 16 days after the current monitoring ends.

[0149] Furthermore, to avoid excessively prolonged monitoring cycles due to a large hysteresis coefficient in the decomposition response, this embodiment can preset an upper limit for the monitoring cycle. If the time interval for the next monitoring cycle calculated as described above exceeds this upper limit, the upper limit will be used as the final time interval for the next monitoring cycle. The upper limit for the monitoring cycle is determined based on the maximum permissible monitoring interval that has enabled effective monitoring of the straw decomposition process throughout historical monitoring processes.

[0150] In summary, this step generates corresponding suggestions for land return management strategies based on the comparison between the revised ecological effect prediction results and the preset thresholds. It also dynamically adjusts the baseline monitoring cycle by combining the decomposition response lag coefficient, so that target fields with a large degree of decomposition response lag adopt a relatively long monitoring interval. This achieves coordinated adjustment of land return management strategies and monitoring frequency, improving the targeting of land return management and the efficiency of monitoring resource utilization.

[0151] Example 2, as Figure 3 As shown, this invention provides a dynamic prediction system for the ecological effects of straw returning to the field based on multi-source data fusion, comprising: Data acquisition module 11 is used to acquire the time series of straw decomposition rate, straw nutrient release rate, soil temperature, soil moisture content, return method identification, soil organic matter baseline value and soil available nutrient baseline value of the target field. The decomposition process analysis module 12 is used to analyze and obtain the straw decomposition process index of the target field at the current monitoring time based on the straw decomposition rate time series and the straw nutrient release rate time series. The hysteresis coefficient acquisition module 13 is used to obtain the hysteresis coefficient of the decomposition response of the target field based on the straw decomposition process index, soil temperature and soil moisture content, through a pre-constructed hysteresis assessment model. The ecological effect prediction module 14 is used to determine the corrected soil organic matter change value and the corrected soil available nutrient change value of the target field at the current monitoring time based on the decomposition response hysteresis coefficient, the soil organic matter baseline value and the soil available nutrient baseline value, and to form the corrected ecological effect prediction result. The field return management and monitoring adjustment module 15 is used to generate field return management strategy suggestions for target fields based on the corrected ecological effect prediction results and field return method identification, and dynamically adjust the time interval of the next monitoring cycle according to the decomposition response hysteresis coefficient.

[0152] In this embodiment of the invention, the data acquisition module 11 is further configured to: Obtain the time series of straw decomposition rate of the target field. The straw decomposition rate time series is a sequence formed by arranging the straw decomposition rate values ​​obtained by sampling and measuring at multiple monitoring time points in chronological order. The straw decomposition rate is the percentage of decomposed straw mass to the initial straw mass. Obtain the time series of straw nutrient release rate of the target field. The straw nutrient release rate time series is a sequence formed by arranging the straw nutrient release rate values ​​obtained at multiple monitoring time points in chronological order. The straw nutrient release rate is the percentage of the amount of target nutrient element released in straw relative to the initial total amount of target nutrient element. Obtain the soil temperature and soil moisture content at the current monitoring time; Obtain the straw return method identifier for the target field. The straw return method identifier is a classification identifier for the straw return method adopted in the target field, including the straw crushing return method identifier, the straw covering return method identifier, and the straw turning and burying return method identifier. Obtain the baseline values ​​of soil organic matter and available nutrients in the target field. The baseline value of soil organic matter is the soil organic matter content of the target field measured before straw return, and the baseline value of available nutrients is the soil alkaline nitrogen, available phosphorus and available potassium content of the target field measured before straw return.

[0153] In this embodiment of the invention, the corrosion process analysis module 12 is further used for: For every two adjacent monitoring time nodes in the straw decomposition rate time series, the straw decomposition rate value of the later monitoring time node is subtracted from the straw decomposition rate value of the earlier monitoring time node, and the difference is divided by the time interval between adjacent monitoring time nodes to obtain the decomposition rate of adjacent monitoring time nodes. The average rate of decay progression is obtained by taking the arithmetic mean of the decay progression rates of all adjacent monitoring time points. For every two adjacent monitoring time nodes in the straw nutrient release rate time series, the straw nutrient release rate value of the later monitoring time node is subtracted from the straw nutrient release rate value of the earlier monitoring time node, and the difference is divided by the time interval between adjacent monitoring time nodes to obtain the nutrient release advance rate of adjacent monitoring time nodes. The nutrient release propulsion rate is obtained by taking the arithmetic mean of the nutrient release propulsion rates of all adjacent monitoring time points. Divide the average decomposition rate by the preset decomposition rate reference value to obtain the decomposition rate normalization value, and divide the average nutrient release rate by the preset nutrient release rate reference value to obtain the nutrient release rate normalization value. The Pearson correlation coefficients of the time series of straw decomposition rate and straw nutrient release rate were calculated to obtain the synchronicity of decomposition and release. Calculate the arithmetic mean of the normalized values ​​of the decomposition rate and the nutrient release rate, and multiply it by the decomposition-release synchronicity to obtain the straw decomposition progress index of the target field at the current monitoring time.

[0154] In this embodiment of the invention, the hysteresis coefficient acquisition module 13 is further configured to: Obtain a pre-constructed hysteresis assessment model. The hysteresis assessment model is a regression model that takes straw decomposition process index, soil temperature and soil moisture content as inputs and outputs decomposition response hysteresis coefficient. Input the straw decomposition process index, soil temperature, and soil moisture content of the target field at the current monitoring time into the hysteresis assessment model, and obtain the decomposition response hysteresis coefficient of the target field at the current monitoring time from the output of the hysteresis assessment model.

[0155] The pre-construction of the hysteresis assessment model includes: Historical monitoring data of multiple historical monitoring plots were obtained. The historical monitoring data included the straw decomposition process index, soil temperature, soil moisture content, and decomposition response hysteresis coefficient labels of each historical monitoring plot at each historical monitoring time. A hysteresis evaluation model is constructed based on a backpropagation neural network model. The hysteresis evaluation model includes an input layer, a hidden layer, and an output layer. The input layer contains three input nodes, which correspond to the straw decomposition process index, soil temperature, and soil moisture content, respectively. The output layer contains one output node, which outputs the decomposition response hysteresis coefficient. Using the straw decomposition process index, soil temperature, and soil moisture content at each historical monitoring time as input features, and the corresponding decomposition response hysteresis coefficient label as output label, the training objective is to minimize the mean square error between the decomposition response hysteresis coefficient output by the hysteresis evaluation model and the decomposition response hysteresis coefficient label. The parameters of the hysteresis evaluation model are iteratively updated until the mean square error converges, thus obtaining the trained hysteresis evaluation model.

[0156] The acquisition of the corrosion response hysteresis coefficient label includes: For each historical monitoring field, obtain its straw decomposition rate time series and soil organic matter content time series; From the time series of straw decomposition rate, the moment when the straw decomposition rate first reaches the preset intermediate decomposition rate is determined as the intermediate decomposition moment; Obtain the soil organic matter content at the mid-decomposition stage as the baseline value for mid-decomposition organic matter; Starting from the mid-decomposition period, the soil organic matter content at each subsequent monitoring time point is checked sequentially. When the soil organic matter content at a consecutive preset number of monitoring time points is higher than the mid-decomposition organic matter baseline value, the first monitoring time point among the consecutive preset number of monitoring time points is taken as the soil response start time. The difference between the soil response start time and the mid-decomposition time is used as the response lag time for historical monitoring plots. Divide the response lag time by the preset reference decomposition cycle to obtain the decomposition response lag coefficient label of the historical monitoring plot.

[0157] In this embodiment of the invention, the ecological effect prediction module 14 is further used for: Obtain the measured value of soil organic matter at the current monitoring time, and subtract the baseline value of soil organic matter from the measured value of soil organic matter to obtain the original change value of soil organic matter; Obtain the measured values ​​of available nitrogen, available phosphorus, and available potassium in the soil at the current monitoring time. Subtract the corresponding baseline values ​​of available nutrients from the measured values ​​of available nitrogen, available phosphorus, and available potassium in the soil, and then sum them to obtain the original change values ​​of available nutrients in the soil. Obtain the decomposition response hysteresis coefficient of the target field at the current monitoring time, and obtain the preset organic matter hysteresis sensitivity coefficient and available nutrient hysteresis sensitivity coefficient; Calculate the organic matter hysteresis correction factor based on the decomposition response hysteresis coefficient and the organic matter hysteresis sensitivity coefficient; Multiply the original change value of soil organic matter by the organic matter lag correction factor to obtain the corrected change value of soil organic matter. Calculate the lag correction factor for available nutrients based on the lag coefficient of decomposition response and the lag sensitivity coefficient of available nutrients. Multiply the original change value of available soil nutrients by the available nutrient lag correction factor, and add the corrected change value of soil organic matter by the preset interaction coefficient to obtain the corrected change value of available soil nutrients. The revised changes in soil organic matter and available soil nutrients constitute the revised ecological effect prediction results.

[0158] In this embodiment of the invention, the field return management and monitoring adjustment module 15 is further used for: When the corrected soil organic matter change value is less than the preset organic matter increase threshold, a management strategy suggestion for returning more organic materials to the field is generated. When the corrected change in available soil nutrients is less than the preset threshold for available nutrients, a suggested management strategy for supplementing available nutrients by returning them to the field is generated. When the corrected soil organic matter change value is less than the preset organic matter increase threshold and the corrected soil available nutrient change value is less than the preset available nutrient threshold, a combined return-to-field management strategy recommendation of increasing organic matter application and supplementing available nutrients is generated. When the corrected soil organic matter change value is greater than or equal to the preset organic matter increase threshold, and the corrected soil available nutrient change value is greater than or equal to the preset available nutrient threshold, a strategy recommendation to maintain the current return-to-field method is generated.

[0159] The time interval for dynamically adjusting the next monitoring cycle based on the corrosion response hysteresis coefficient includes: Obtain the preset baseline monitoring period, which is the monitoring time interval of the target field under normal conditions; The baseline monitoring cycle is corrected using the corrosion response hysteresis coefficient to obtain the time interval of the next monitoring cycle. The time interval of the next monitoring cycle is positively correlated with the corrosion response hysteresis coefficient.

[0160] In summary, this invention achieves coordinated analysis of the straw decomposition process, nutrient release process, and soil ecological response by constructing modules for data acquisition, decomposition process analysis, lag coefficient acquisition, ecological effect prediction, and straw return management and monitoring adjustment. By training a lag assessment model using historical monitoring data and dynamically correcting changes in soil organic matter and available nutrients using the decomposition response lag coefficient, the accuracy of ecological effect prediction results is improved. Simultaneously, differentiated straw return management strategies are generated based on the prediction results, and the monitoring cycle is dynamically adjusted according to the degree of decomposition response lag. This integrates straw return ecological effect prediction, management decision-making, and monitoring scheduling, improving the targeting and dynamic adaptability of straw return management.

[0161] It should be noted that the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0162] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0163] This specification and accompanying drawings are merely illustrative examples of the invention and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its scope. Therefore, if such modifications and modifications fall within the scope of the invention and its equivalents, the invention is intended to include these modifications and modifications.

Claims

1. A method for dynamic prediction of the ecological effects of straw returning to the field based on multi-source data fusion, characterized in that, include: Obtain the time series of straw decomposition rate, straw nutrient release rate, soil temperature, soil moisture content, return method identification, soil organic matter baseline value, and soil available nutrient baseline value of the target field. Based on the straw decomposition rate time series and the straw nutrient release rate time series, the straw decomposition process index of the target field at the current monitoring time is analyzed and obtained; Based on the straw decomposition process index, the soil temperature, and the soil moisture content, the decomposition response hysteresis coefficient of the target field is obtained through a pre-constructed hysteresis assessment model. Based on the decomposition response hysteresis coefficient, and based on the baseline values ​​of soil organic matter and available soil nutrients, the corrected changes in soil organic matter and available soil nutrients in the target field at the current monitoring time are determined, forming the corrected ecological effect prediction results. Based on the revised ecological effect prediction results and the field return method identifier, a field return management strategy suggestion for the target field is generated, and the time interval of the next monitoring cycle is dynamically adjusted according to the decomposition response hysteresis coefficient.

2. The method according to claim 1, wherein, The acquisition of the time series of straw decomposition rate, straw nutrient release rate, soil temperature, soil moisture content, return method identification, soil organic matter baseline value, and soil available nutrient baseline value of the target field includes: Obtain the straw decomposition rate time series of the target field, wherein the straw decomposition rate time series is a sequence formed by arranging the straw decomposition rate values ​​obtained by sampling and measuring at multiple monitoring time nodes in chronological order, and the straw decomposition rate is the percentage of decomposed straw mass to the initial straw mass; Obtain the time series of straw nutrient release rate of the target field. The time series of straw nutrient release rate is a sequence formed by arranging the straw nutrient release rate values ​​obtained at multiple monitoring time nodes in chronological order. The straw nutrient release rate is the percentage of the amount of target nutrient element released in the straw relative to the initial total amount of the target nutrient element. Obtain the soil temperature and soil moisture content at the current monitoring time; Obtain the straw return method identifier of the target field, which is a classification identifier of the straw return method adopted by the target field, including the straw crushing return method identifier, the straw covering return method identifier, and the straw turning and burying return method identifier. Obtain the baseline values ​​of soil organic matter and available soil nutrients for the target field, wherein the baseline value of soil organic matter is the soil organic matter content of the target field measured before straw return, and the baseline value of available soil nutrients is the soil alkaline nitrogen, available phosphorus and available potassium content of the target field measured before straw return.

3. The method for dynamic prediction of the ecological effects of straw returning to the field based on multi-source data fusion according to claim 1, characterized in that, The step of analyzing and obtaining the straw decomposition process index of the target field at the current monitoring time based on the straw decomposition rate time series and the straw nutrient release rate time series includes: For every two adjacent monitoring time nodes in the straw decomposition rate time series, the straw decomposition rate value of the later monitoring time node is subtracted from the straw decomposition rate value of the earlier monitoring time node, and the difference is divided by the time interval between adjacent monitoring time nodes to obtain the decomposition advancement rate of the adjacent monitoring time nodes. The average rate of decay progression is obtained by taking the arithmetic mean of the decay progression rates of all adjacent monitoring time points. For every two adjacent monitoring time nodes in the straw nutrient release rate time series, the straw nutrient release rate value of the later monitoring time node is subtracted from the straw nutrient release rate value of the earlier monitoring time node, and the difference is divided by the time interval between adjacent monitoring time nodes to obtain the nutrient release propulsion rate of the adjacent monitoring time nodes. The nutrient release propulsion rate is obtained by taking the arithmetic mean of the nutrient release propulsion rates of all adjacent monitoring time points. Divide the average decomposition rate by a preset decomposition rate reference value to obtain a normalized decomposition rate value, and divide the average nutrient release rate by a preset nutrient release rate reference value to obtain a normalized nutrient release rate value. The Pearson correlation coefficients of the straw decomposition rate time series and the straw nutrient release rate time series were calculated to obtain the synchronicity of decomposition and release. Calculate the arithmetic mean of the normalized value of the decomposition rate and the normalized value of the nutrient release rate, and multiply it by the decomposition-release synchronicity to obtain the straw decomposition process index of the target field at the current monitoring time.

4. The method for dynamic prediction of the ecological effects of straw return to the field based on multi-source data fusion according to claim 1, characterized in that, The step of obtaining the decomposition response hysteresis coefficient of the target field based on the straw decomposition process index, the soil temperature, and the soil moisture content using a pre-constructed hysteresis assessment model includes: Obtain a pre-constructed hysteresis assessment model, wherein the hysteresis assessment model is a regression model that takes straw decomposition process index, soil temperature and soil moisture content as input and outputs decomposition response hysteresis coefficient. The straw decomposition process index, soil temperature, and soil moisture content of the target field at the current monitoring time are input into the hysteresis evaluation model to obtain the decomposition response hysteresis coefficient of the target field at the current monitoring time.

5. The method for dynamic prediction of the ecological effects of straw return to the field based on multi-source data fusion according to claim 4, characterized in that, The pre-construction of the hysteresis assessment model includes: Acquire historical monitoring data from multiple historical monitoring plots, wherein the historical monitoring data includes the straw decomposition process index, soil temperature, soil moisture content, and decomposition response hysteresis coefficient label of each historical monitoring plot at each historical monitoring time. A hysteresis evaluation model is constructed based on a backpropagation neural network model. The hysteresis evaluation model includes an input layer, a hidden layer, and an output layer. The input layer contains three input nodes, which correspond to the straw decomposition process index, the soil temperature, and the soil moisture content, respectively. The output layer contains one output node, which outputs the decomposition response hysteresis coefficient. Using the straw decomposition process index, soil temperature, and soil moisture content at each historical monitoring time as input features, and the corresponding decomposition response hysteresis coefficient label as output label, the training objective is to minimize the mean square error between the decomposition response hysteresis coefficient output by the hysteresis evaluation model and the decomposition response hysteresis coefficient label. The parameters of the hysteresis evaluation model are iteratively updated until the mean square error converges, thus obtaining the trained hysteresis evaluation model.

6. The method for dynamic prediction of the ecological effects of straw return to the field based on multi-source data fusion according to claim 5, characterized in that, The acquisition of the corrosion response hysteresis coefficient label includes: For each historical monitoring field, obtain its straw decomposition rate time series and soil organic matter content time series; From the straw decomposition rate time series, the moment when the straw decomposition rate first reaches the preset intermediate decomposition rate is determined as the intermediate decomposition moment; The soil organic matter content at the mid-decomposition time point is obtained as the mid-decomposition organic matter baseline value; Starting from the mid-decomposition time, the soil organic matter content of each subsequent monitoring time node is checked sequentially along the time sequence. When the soil organic matter content of a consecutive preset number of monitoring time nodes is higher than the mid-decomposition organic matter benchmark value, the first monitoring time node among the consecutive preset number of monitoring time nodes is taken as the soil response start time. The difference between the soil response start time and the decomposition mid-term time is taken as the response lag time of the historical monitoring plot. Divide the response lag time by a preset reference decomposition cycle to obtain the decomposition response lag coefficient label of the historical monitoring field.

7. The method for dynamic prediction of the ecological effects of straw return to the field based on multi-source data fusion according to claim 1, characterized in that, The process involves determining the corrected changes in soil organic matter and available soil nutrients at the current monitoring time for the target field based on the decomposition response hysteresis coefficient, the baseline values ​​of soil organic matter and available soil nutrients, and constructing the corrected ecological effect prediction results, including: Obtain the measured value of soil organic matter at the current monitoring time, and subtract the baseline value of soil organic matter from the measured value of soil organic matter to obtain the original change value of soil organic matter; Obtain the measured values ​​of available nitrogen, available phosphorus, and available potassium in the soil at the current monitoring time. Subtract the corresponding baseline values ​​of available nutrients in the soil from the measured values ​​of available nitrogen, available phosphorus, and available potassium, and then sum them to obtain the original change values ​​of available nutrients in the soil. Obtain the decomposition response hysteresis coefficient of the target field at the current monitoring time, and obtain the preset organic matter hysteresis sensitivity coefficient and available nutrient hysteresis sensitivity coefficient; Calculate the organic matter hysteresis correction factor based on the decomposition response hysteresis coefficient and the organic matter hysteresis sensitivity coefficient; Multiply the original change value of soil organic matter by the organic matter lag correction factor to obtain the corrected change value of soil organic matter. Calculate the hysteresis correction factor for readily available nutrients based on the decomposition response hysteresis coefficient and the readily available nutrient hysteresis sensitivity coefficient. Multiply the original change value of available soil nutrients by the available nutrient lag correction factor, and add the corrected change value of soil organic matter by the preset interaction coefficient to obtain the corrected change value of available soil nutrients. The corrected changes in soil organic matter and soil available nutrients are combined to form the corrected ecological effect prediction result.

8. The method for dynamic prediction of the ecological effects of straw return to the field based on multi-source data fusion according to claim 1, characterized in that, The process of generating a suggested management strategy for returning farmland to the target field based on the revised ecological effect prediction results and the farmland return method identifier includes: When the modified soil organic matter change value is less than the preset organic matter increase threshold, a management strategy suggestion for returning more organic materials to the field is generated. When the modified value of the change in available soil nutrients is less than the preset threshold for available nutrients, a suggested management strategy for supplementing available nutrients by returning them to the field is generated. When the modified soil organic matter change value is less than the preset organic matter increase threshold and the modified soil available nutrient change value is less than the preset available nutrient threshold, a combined return-to-field management strategy recommendation of increasing organic matter application and supplementing available nutrients is generated. When the modified soil organic matter change value is greater than or equal to the preset organic matter increase threshold, and the modified soil available nutrient change value is greater than or equal to the preset available nutrient threshold, a strategy suggestion to maintain the current return-to-field method is generated.

9. The method for dynamic prediction of the ecological effects of straw return to the field based on multi-source data fusion according to claim 1, characterized in that, The step of dynamically adjusting the time interval for the next monitoring cycle based on the corrosion response hysteresis coefficient includes: Obtain a preset baseline monitoring period, which is the monitoring time interval of the target field under normal conditions; The baseline monitoring cycle is corrected using the corrosion response hysteresis coefficient to obtain the time interval of the next monitoring cycle, wherein the time interval of the next monitoring cycle is positively correlated with the corrosion response hysteresis coefficient.

10. A dynamic prediction system for the ecological effects of straw returning to the field based on multi-source data fusion, characterized in that, The method for dynamically predicting the ecological effects of straw return to the field by implementing multi-source data fusion as described in any one of claims 1 to 9 includes: The data acquisition module is used to acquire the time series of straw decomposition rate, straw nutrient release rate, soil temperature, soil moisture content, return method identification, soil organic matter baseline value, and soil available nutrient baseline value of the target field. The decomposition process analysis module is used to analyze and obtain the straw decomposition process index of the target field at the current monitoring time based on the straw decomposition rate time series and the straw nutrient release rate time series. The hysteresis coefficient acquisition module is used to obtain the decomposition response hysteresis coefficient of the target field based on the straw decomposition process index, the soil temperature and the soil moisture content, through a pre-constructed hysteresis evaluation model. The ecological effect prediction module is used to determine the corrected changes in soil organic matter and soil available nutrients of the target field at the current monitoring time based on the decomposition response hysteresis coefficient, the baseline values ​​of soil organic matter and soil available nutrients, and to form the corrected ecological effect prediction result. The field return management and monitoring adjustment module is used to generate field return management strategy suggestions for the target field based on the corrected ecological effect prediction results and the field return method identifier, and dynamically adjust the time interval of the next monitoring cycle according to the decomposition response hysteresis coefficient.

Images