Ecological pasture carbon storage dynamic prediction method and system based on meteorological data

Abstract

The application provides a kind of ecological pasture carbon storage dynamic prediction method and system based on meteorological data, relating to carbon storage prediction technical field, method includes: collecting multiple data, constructing carbon storage calculation model, according to pasture parameter data, calculate pasture basic carbon storage.Analysis of the influence of pasture parameter data on the target ecological pasture area obtains carbon cycle parameters, and the influence index of carbon storage is calculated according to the influence of dairy cattle breeding factors on the basic carbon storage of pasture.Construct a multi-task learning model based on secretary bird algorithm optimization, input carbon cycle parameters and meteorological data, output meteorological influence carbon data.According to the carbon storage influence index and combining meteorological influence carbon data, the carbon storage change rule is generated, and the carbon management strategy is formulated according to the carbon storage change rule.The application improves the accuracy and scientificity of ecological pasture carbon storage dynamic prediction, and provides strong support for sustainable management and carbon emission reduction of ecological pasture.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of carbon storage prediction, in particular to an ecological pasture carbon storage dynamic prediction method and system based on meteorological data. BACKGROUND

[0002] Under the background of the "double carbon" strategy and agricultural green transformation, as the core carrier of carbon cycle in the livestock industry, the accurate control of carbon storage in ecological pastures is the key to achieving agricultural emission reduction and carbon sequestration and ensuring the sustainable development of the dairy industry. According to IPCC data, global methane emissions from ruminant animals account for 28% of human activity methane emissions, and carbon emissions from dairy cattle breeding in China account for 20% of total livestock and poultry breeding emissions. The carbon footprint per unit of milk is significantly higher than that of New Zealand and other countries, and there is an urgent need for dynamic control of carbon storage.

[0003] Meteorology is a core driving factor of carbon cycle. Temperature affects the rate of soil carbon decomposition, and precipitation and sunlight directly determine the carbon sequestration efficiency of vegetation. However, existing technologies mostly use meteorological data as auxiliary reference or use only a single indicator for rough correction, ignoring the synergistic effects of factors and regional differences, relying on static emission coefficients, resulting in large deviations between carbon cycle simulation and reality, and failing to accurately reflect the driving effect of meteorology on carbon storage.

[0004] Carbon storage changes are the result of the combined effects of breeding (stocking, daily feed, and manure treatment) and meteorological factors. For example, excessive crude protein in daily feed can increase 15%-20% of nitrogen excretion in manure and urine, and yucca extract can reduce 10%-15% of methane emissions. However, existing technologies mostly focus on a single factor, ignore coupling effects, and are mainly qualitative descriptions, resulting in unclear emission reduction effects and no basis for strategy development.

[0005] Existing technologies mainly use static accounting, making it difficult to predict future trends. Some prediction models have single input features, do not integrate meteorological prediction data, and do not optimize algorithms for multi-objective prediction, resulting in poor adaptability, low precision, and inability to identify carbon loss risks in advance, restricting the initiative of pasture carbon management.

[0006] In summary, there is an urgent need for an ecological pasture carbon storage dynamic prediction method and system based on meteorological data to support accurate decision-making, quantify emission reduction effects, respond to climate risks, and promote low-carbon transformation of the dairy industry. SUMMARY

[0007] The present application provides an ecological pasture carbon storage dynamic prediction method and system based on meteorological data to solve the problem of lack of analysis of the impact of meteorology on carbon cycle and insufficient evaluation of the impact of carbon storage changes in existing technologies.

[0008] In one aspect, the present application provides an ecological pasture carbon storage dynamic prediction method based on meteorological data, comprising:

[0009] Meteorological data, cow breeding factors and pasture parameter data of the target ecological pasture area are collected, a carbon storage calculation model is constructed, and the basic carbon storage of the pasture is calculated according to the pasture parameter data.

[0010] The influence of the pasture parameter data on the target ecological pasture area is analyzed to obtain carbon cycle parameters, and the influence index of the carbon storage is calculated according to the influence of the cow breeding factors on the basic carbon storage of the pasture.

[0011] A multi-task learning model based on secretary bird algorithm optimization is constructed, meteorological data and carbon cycle parameters are input, and meteorological influence carbon data is output.

[0012] The carbon storage change rule is generated according to the carbon storage influence index and combined with the meteorological influence carbon data, and the carbon management strategy is formulated according to the carbon storage change rule.

[0013] The present application provides an ecological pasture carbon storage dynamic prediction method based on meteorological data, and the steps of constructing the carbon storage calculation model include:

[0014] For different forms of carbon storage, the corresponding influence factors and carbon conversion paths are determined as carbon characteristic parameters.

[0015] According to the carbon characteristic parameters, the vegetation sub-model is constructed in combination with the correlation of vegetation growth and meteorological factors, and the soil organic carbon storage is calculated according to the difference of soil layer and temperature decomposition depth to construct the soil sub-model.

[0016] According to the weight, the number of livestock and the temperature influence, the carbon storage in the cow is quantified to construct the cow carbon storage, and the carbon loss rule between manure and stubble is associated to construct the waste sub-model in combination with the influence of temperature and wind speed.

[0017] The vegetation sub-model, the soil sub-model, the cow carbon storage and the waste sub-model are coupled to form the carbon storage calculation model.

[0018] The present application provides an ecological pasture carbon storage dynamic prediction method based on meteorological data, and the steps of calculating the basic carbon storage of the pasture include:

[0019] The data exceeding the preset multiple standard deviation is removed from the pasture parameter data using the Laplace criterion, and the average value of the same type of pasture in the same region is used to fill the missing value to obtain the pasture processing parameter data.

[0020] The pasture processing parameter data is input into the carbon storage calculation model, the carbon storage of the corrected vegetation, soil, cow and waste is calculated, and the sum is obtained to obtain the basic carbon storage of the pasture.

[0021] The present application provides an ecological pasture carbon storage dynamic prediction method based on meteorological data, and the steps of obtaining the carbon cycle parameters include:

[0022] Based on the ecological pasture carbon flow law, the carbon cycle paths among different carbon storage carriers are analyzed, the pasture parameter data is classified, and the circulation parameter corresponding relationship is established.

[0023] The correlation analysis method is used to calculate the correlation degree of the pasture parameter data and the carbon cycle process index, so as to determine the influence weight.

[0024] According to the circulation parameter corresponding relationship and the influence weight, the process parameters of carbon input, carbon output and carbon conversion are calculated as carbon cycle parameters.

[0025] The present application provides an ecological pasture carbon storage dynamic prediction method based on meteorological data, and the steps of calculating the carbon storage influence index include:

[0026] Based on the actual breeding scene setting factor adjustment value, the total carbon storage is calculated according to the correlation logic between the dairy cattle breeding factors and the carbon storage carriers.

[0027] According to the basic carbon storage of the pasture and the total carbon storage, the carbon storage relative change range adjusted according to the factor adjustment value of different breeding factors is calculated to obtain the carbon storage influence index.

[0028] The present application provides an ecological pasture carbon storage dynamic prediction method based on meteorological data, and the steps of constructing a multi-task learning model include:

[0029] According to the feature extraction network shared by multiple tasks, the common features between different tasks are learned as shared layers, and the output layers of each task are learned as task-specific layers.

[0030] According to the shared layer and the task-specific layer, a model framework is constructed, and the to-be-optimized parameters of the model framework are encoded as secretary bird individual positions, and a comprehensive loss function is selected as the fitness function.

[0031] According to the upper and lower bounds of the to-be-optimized parameters, a plurality of secretary bird individuals are randomly generated.

[0032] The process of secretary birds searching for prey is simulated, and the secretary bird searching position is generated by differential mutation.

[0033] Local search is combined with Brownian motion, and the secretary bird digestion position and the secretary bird attack position are generated by simulating the precise attack of the secretary bird.

[0034] The process of secretary birds avoiding predators is simulated, and a dynamic disturbance strategy is adopted to balance exploration and development to obtain the secretary bird escape position.

[0035] When the preset iteration number is reached, the update is stopped, the fitness values of different secretary bird positions are calculated according to the fitness function, and the to-be-optimized parameters with the smallest fitness value are selected to construct a multi-task learning model.

[0036] The application provides an ecological pasture carbon storage dynamic prediction method based on meteorological data, and the output of meteorological influence carbon data includes the following steps:

[0037] The meteorological data and the time scale of the carbon cycle parameters are aligned and standardized for input into a multi-task learning model.

[0038] The common features of meteorological data are learned according to the shared layer, and the prediction accuracy index of each carbon cycle parameter is calculated according to the task-specific layer.

[0039] The sensitivity coefficient is obtained by quantifying the influence degree of a single meteorological factor on the carbon cycle parameter through the control variable method.

[0040] The relationship curve between each carbon cycle parameter and the meteorological factor is output, and the meteorological influence carbon data is sorted according to the absolute value of the sensitivity coefficient.

[0041] The application provides an ecological pasture carbon storage dynamic prediction method based on meteorological data, and the generation of the carbon storage change rule includes the following steps:

[0042] According to the preset time period as the link, the meteorological influence carbon data and the carbon storage influence index are associated with the pasture basic carbon storage to determine the influence path.

[0043] According to the influence path, the climate influence and the breeding influence are integrated to adjust the carbon storage calculation model to obtain the dynamic change formula.

[0044] According to the dynamic change formula, the dynamic changes of the vegetation carbon pool, the soil carbon pool, the cow carbon pool and the waste carbon pool are calculated and integrated to obtain the total carbon storage change.

[0045] By changing the meteorological conditions and breeding factors, the total carbon storage change under different scenarios is simulated, and the change trend with time is analyzed to identify the periodic and trend characteristics to obtain the carbon storage change rule.

[0046] The application provides an ecological pasture carbon storage dynamic prediction method based on meteorological data, and the steps of formulating a carbon management strategy include:

[0047] The carbon storage change rule is disassembled from the time dimension, the factor contribution and the scene threshold to determine the problems in different links of the carbon cycle.

[0048] According to the influence cost, the problems in different links are sorted to set the overall goal, which is disassembled into phased goals, and matched with the time change rule of carbon storage.

[0049] According to the change characteristics of the preset time period of carbon storage, carbon loss prevention and control measures and carbon accumulation strengthening measures are formulated, and differential measures are formulated according to the factor contribution, and boundary management schemes are formulated according to the scene threshold, which are integrated to form a carbon management strategy.

[0050] In another aspect, the present application provides a meteorological data-based ecological pasture carbon storage dynamic prediction system, comprising:

[0051] A carbon storage base algorithm module is configured to collect meteorological data, dairy cattle breeding factors and pasture parameter data of a target ecological pasture area, construct a carbon storage calculation model, and calculate a basic carbon storage of the pasture according to the pasture parameter data.

[0052] A carbon cycle influence calculation module is configured to analyze the influence of the pasture parameter data on the target ecological pasture area to obtain carbon cycle parameters, and calculate a carbon storage influence index according to the influence of the dairy cattle breeding factors on the basic carbon storage of the pasture.

[0053] A carbon influence module is configured to construct a multi-task learning model optimized based on a secretary bird algorithm, input the carbon cycle parameters and meteorological data, and output meteorological influence carbon data.

[0054] A carbon management strategy formulation module is configured to generate a carbon storage change rule according to the carbon storage influence index and in combination with the meteorological influence carbon data, and formulate a carbon management strategy according to the carbon storage change rule.

[0055] The meteorological data-based ecological pasture carbon storage dynamic prediction method and system provided by the present application comprehensively consider multiple factors to construct a model, accurately calculate a basic carbon storage of a pasture, and realize accurate prediction of the dynamic change of carbon storage by analyzing the influence of meteorological and breeding factors on carbon storage, thereby providing reliable data support for carbon management of the pasture. The influence of dairy cattle breeding factors and meteorological factors on carbon storage is quantified, and the key influencing factors and their influence degree are determined. For example, the specific influence of factors such as stocking rate, crude protein content of daily feed, and meteorological factors such as temperature and precipitation on carbon storage is determined, thereby providing a basis for taking targeted measures. The secretary bird algorithm is used to optimize the multi-task learning model, thereby improving the analysis capability of the model for meteorological influence carbon data, enabling the model to better adapt to complex ecological pasture environments, and improving the accuracy and reliability of the prediction. A scientific carbon management strategy is formulated according to the carbon storage change rule, which covers multiple aspects such as time dimension, factor contribution and scene threshold, thereby effectively guiding carbon resource management of the pasture, improving carbon sink capacity, reducing carbon emissions, and helping the pasture to achieve low-carbon sustainable development. BRIEF DESCRIPTION OF DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the present application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without any creative effort.

[0057] Figure 1This is one of the flowcharts illustrating the dynamic prediction method for carbon storage in ecological pastures based on meteorological data provided in this embodiment of the invention.

[0058] Figure 2 This is the second flowchart of the dynamic prediction method for carbon storage in ecological pastures based on meteorological data provided in this embodiment of the invention.

[0059] Figure 3 This is a flowchart illustrating the impact of dairy cow manure on carbon storage in the dynamic prediction method for carbon storage in ecological pastures based on meteorological data provided in this embodiment of the invention.

[0060] Figure 4 This is a flowchart illustrating the dynamic prediction system for carbon storage in ecological pastures based on meteorological data provided in an embodiment of the present invention. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0062] The following is combined Figures 1-4 This invention describes a method and system for dynamic prediction of carbon storage in ecological pastures based on meteorological data.

[0063] like Figure 1 As shown in the embodiment of the present invention, the method for dynamic prediction of carbon storage in ecological pastures based on meteorological data includes:

[0064] Meteorological data, dairy farming factors, and pasture parameter data of the target ecological pasture area are collected to construct a carbon storage calculation model, and the basic carbon storage of the pasture is calculated based on the pasture parameter data. Pasture parameter data may include carbon pools from vegetation, soil, dairy cows, and waste.

[0065] The steps involved in constructing a carbon storage calculation model include:

[0066] For different forms of carbon storage, corresponding influencing factors and carbon conversion pathways are determined as carbon characteristic parameters.

[0067] Based on carbon characteristic parameters, a vegetation sub-model was constructed by combining the correlation between vegetation growth and meteorological factors. A soil sub-model was constructed by calculating soil organic carbon storage based on soil layer differences and temperature decomposition depth.

[0068] The carbon storage in the dairy cows is quantified according to the weight, the stock and the temperature influence, the carbon storage of the dairy cows is constructed, and the carbon loss law between the manure and the stubble is combined, and the temperature and the wind speed influence are associated to construct the waste sub-model.

[0069] The vegetation sub-model, the soil sub-model, the dairy cow carbon storage and the waste sub-model are coupled to form the carbon storage calculation model.

[0070] The steps of calculating the basic carbon storage of the pasture include:

[0071] The data exceeding the preset multiple standard deviation is removed from the pasture parameter data by using the Relyda criterion, and the average value of the same type of pasture in the same region is used to fill the missing value to obtain the processed pasture parameter data.

[0072] The processed pasture parameter data is input into the carbon storage calculation model, the carbon storage of the corrected vegetation, soil, dairy cow and waste is calculated, and the sum is obtained to obtain the basic carbon storage of the pasture.

[0073] For example: the total carbon storage is 29,427 tons, and the last two decimal places are retained.

[0074] The proportion of the library is shown in the following table:

[0075] Carbon pool Carbon stock (tonnes) Proportion Vegetation 4,214.02 14.32% Soil 11,969.30 40.67% Dairy cows 1,851.00 6.29% Waste 11,393.00 38.72%

[0076] The carbon cycle parameters are obtained by analyzing the influence of the pasture parameter data on the target ecological pasture area, and the carbon storage influence index is calculated according to the influence of the dairy farming factors on the basic carbon storage of the pasture.

[0077] The steps of obtaining the carbon cycle parameters include:

[0078] Based on the carbon flow law of the ecological pasture, the carbon cycle path among different carbon storage carriers is analyzed, the pasture parameter data is classified, and the corresponding relationship of the cycle parameters is established.

[0079] The corresponding relationship of the cycle parameters can include: vegetation carbon pool parameters: biomass, carbon content, growth cycle→associated carbon input, carbon output, carbon conversion.

[0080] Soil carbon pool parameters: organic carbon content, bulk density, soil thickness→associated carbon input, carbon output.

[0081] Dairy cow carbon pool parameters: stock, weight, feed intake, carbon content→associated carbon input, carbon output, carbon conversion.

[0082] Waste carbon pool parameters: manure yield, stubble yield, carbon content→associated carbon input, carbon output, carbon conversion.

[0083] The correlation analysis method is used to calculate the correlation degree of the pasture parameter data and the carbon cycle process index, so as to determine the influence weight.

[0084] According to the circulation parameter correspondence and the influence weight, process parameters of carbon input, carbon output and carbon conversion are calculated as carbon circulation parameters.

[0085] The carbon input parameters can include vegetation photosynthetic carbon fixation rate, external feed carbon input amount, residual stubble carbon input amount into soil, and fecal carbon input amount into soil, etc.

[0086] The carbon output parameters can include vegetation respiration carbon release rate, soil respiration carbon release rate, methane emission rate of dairy cows, and waste decomposition carbon loss rate, etc.

[0087] The carbon conversion parameters can include vegetation-dairy cow carbon conversion rate, vegetation-waste carbon transfer rate, dairy cow-waste carbon transfer rate, and waste-soil carbon conversion rate, etc.

[0088] The step of calculating the carbon storage influence index comprises:

[0089] Based on the actual breeding scene setting factor adjustment value, the total carbon storage is calculated according to the association logic between the dairy cow breeding factors and the carbon storage carriers.

[0090] The dairy cow breeding factors can include stock number, feed, additive, etc.

[0091] The stock number directly adjusts the dairy cow carbon pool and the waste carbon pool, and indirectly affects the soil and vegetation carbon pools. The feed directly affects the vegetation carbon pool, the dairy cow carbon pool and the waste carbon pool. The additive directly affects the dairy cow carbon pool and the waste carbon pool, causing the increase of the body weight of the dairy cow and the decrease of the fecal discharge rate.

[0092] For example: the addition of nano bubble water can destroy the cellulose structure and reduce the crystallinity of cellulose, which is beneficial for microorganisms to decompose cellulose into VFA, and then VFA can be used to produce methane compounds to produce more methane. From the economic point of view, the micro-nano bubble generator can be used for pretreatment, and the micro-nano bubble water prepared by the generator can be used for anaerobic fermentation of fecal water to improve the fermentation efficiency. The gas required for bubble generation can be air, etc., which has low cost. Considering the characteristics of the wastewater of the dairy farm and the economy, the project takes the wastewater of the milking hall and the cleaning wastewater of the drinking trough as the water source to prepare air nano bubble water and carry out related research.

[0093] For example, the addition of nano bubble water can destroy the cellulose structure and reduce the crystallinity of cellulose, which is beneficial for microorganisms to decompose cellulose into VFA, and then VFA can be used to produce methane compounds to produce more methane. From the economic point of view, the micro-nano bubble generator can be used for pretreatment, and the micro-nano bubble water prepared by the generator can be used for anaerobic fermentation of fecal water to improve the fermentation efficiency. The gas required for bubble generation can be air, etc., which has low cost. Considering the characteristics of the wastewater of the dairy farm and the economy, the project takes the wastewater of the milking hall and the cleaning wastewater of the drinking trough as the water source to prepare air nano bubble water and carry out related research. Figure 3As shown, the largest amount of dairy farming wastewater is produced, and the storage process will cause a large amount of CH4, NH3 and other gas emissions, which not only pollutes the atmospheric environment, but also causes nutrient resource loss. Although large-scale dairy manure and water storage facilities have gradually adopted black film covering to reduce carbon and nitrogen gas emissions, open-type manure and water storage facilities are still common in dairy farms, and emission reduction is still in great demand. Adding acidifiers is considered to promote NH3 emission reduction and not to increase GHG emissions, which is a good technical measure with comprehensive effects, but often requires a matching mixing device to fully mix, which does not have such conditions in many scenarios. In view of the above characteristics, the project selects three types of additives that can be industrialized, such as citric acid, wood vinegar and microbial ammonia fixation bacteria, and uses surface spraying to explore carbon and nitrogen gas emissions during the storage process of dairy manure and water.

[0094] According to the basic carbon storage and total carbon storage of the pasture, the carbon storage influence index is calculated according to the relative change amplitude of the carbon storage adjusted by the adjustment value of different breeding factors, and the formula is expressed as:

[0095]

[0096] In the formula, is the carbon storage influence index, is the basic carbon storage of the pasture, is the total carbon storage. When , the current factor adjustment increases the carbon storage. , the current factor adjustment reduces the carbon storage. The greater the influence degree is, the higher the influence degree is.

[0097] A multi-task learning model based on secretary bird algorithm optimization is constructed, inputting carbon cycle parameters and meteorological data, and outputting meteorological influence carbon data.

[0098] As shown in Figure 2 , the steps of constructing the multi-task learning model include:

[0099] According to the feature extraction network shared by multiple tasks, learn the common features between different tasks as shared layers, and learn the task-specific features of each task as task-specific layers.

[0100] According to the shared layer and the task-specific layer, a model framework is constructed, and the to-be-optimized parameters of the model framework are coded as secretary bird individual positions, and a comprehensive loss function is selected as the fitness function, and the formula is expressed as:

[0101]

[0102] In the formula, is the number of tasks, is the weight of the task , It's a loss for the mission. It is the fitness function.

[0103] Based on the upper and lower bounds of the parameters to be optimized, multiple individual Secretary Birds are randomly generated, as expressed by the formula:

[0104]

[0105] In the formula, It is the first The location of each secretary bird individual It is the first An upper bound for the parameters to be optimized. It is the first The lower bound of the parameter to be optimized. It is a random number between [0, 1].

[0106] Simulating the process of a secretary bird searching for prey, the location of the secretary bird is generated through differential mutation, expressed by the formula:

[0107]

[0108] In the formula, They are randomly selected individuals. It is a scaling factor. The secretary bird is searching for its location.

[0109] By combining Brownian motion for local search and simulating the secretary bird's precise attack to generate the secretary bird's digestion location and attack location, the formula is expressed as:

[0110]

[0111] In the formula, These are standard normally distributed random numbers. It is the current globally optimal solution. It is the secretary bird's digestive location. This is the Secretary Bird's attack position. It is a nonlinear perturbation factor.

[0112] in

[0113] In the formula, It is a fixed constant. yes index, It is a random number between [0,1]. yes The power of.

[0114] The process of a secretary bird evading predators is simulated. A dynamic perturbation strategy is used to balance exploration and development to determine the secretary bird's escape location. The formula is expressed as:

[0115]

[0116] In the formula, is a random candidate solution, is a dynamic disturbance factor, is a secretary bird escape position.

[0117] After reaching the preset number of iterations, stop updating, calculate the fitness values of different secretary bird positions according to the fitness function, and select the to-be-optimized parameter with the minimum fitness value to construct a multi-task learning model.

[0118] The step of outputting the meteorological influence carbon data includes:

[0119] Align the time scales of meteorological data and carbon cycle parameters, and perform standardization processing to input the multi-task learning model.

[0120] According to the shared layer, learn the common features of meteorological data, and according to the task-specific layer, calculate the prediction accuracy index of each carbon cycle parameter.

[0121] Quantify the influence degree of a single meteorological factor on the carbon cycle parameter to obtain a sensitivity coefficient through a control variable method.

[0122] The step of obtaining the sensitivity coefficient can include: fixing other meteorological factors as the mean value, and only changing the target meteorological factor.

[0123] Use the trained multi-task learning model to predict the changes of the corresponding carbon cycle parameters.

[0124] Calculate the sensitivity coefficient (the change amount of the carbon cycle parameter when the meteorological factor changes by 1 unit), which is expressed by the formula:

[0125]

[0126] In the formula, is a meteorological factor, is a carbon cycle parameter, is the change amount of the meteorological factor, is the change amount of the parameter.

[0127] Output the relationship curve between each carbon cycle parameter and the meteorological factor, and sort according to the absolute value of the sensitivity coefficient to obtain the meteorological influence carbon data.

[0128] When the carbon cycle parameter is methane emissions, the most sensitive meteorological factor is temperature, with a sensitivity coefficient of 0.025 kgC / (head·d·℃). When the carbon cycle parameter is soil respiration, the most sensitive meteorological factor is temperature, with a sensitivity coefficient of 0.01 kgC / (hm²·d·℃). For example, for every 1℃ increase in temperature, the methane emission rate of dairy cows increases by 0.025 kgC / (head·d) (sensitivity coefficient). For every 100 hours of additional sunshine, the vegetation photosynthetic carbon sequestration rate (GPP) increases by 3 kgC / (hm²·d).

[0129] Based on the carbon storage impact index and combined with meteorological carbon data, a carbon storage change pattern is generated, and a carbon management strategy is formulated based on the carbon storage change pattern.

[0130] The steps involved in understanding the changing patterns of carbon reserves include:

[0131] Based on a preset time period, the impact path is determined by correlation analysis between meteorological carbon data and carbon storage impact index and the basic carbon storage of pastures.

[0132] By integrating climate and aquaculture impacts based on the impact pathways, a dynamic change formula is obtained by adjusting the carbon storage calculation model.

[0133] The dynamic change formula includes the vegetation dynamic change formula, which is expressed as:

[0134]

[0135] In the formula, It is the baseline value of vegetation carbon pool. Carbon data based on meteorological impacts Change yes Benchmark value It is the first The impact index of dietary crude protein on vegetation carbon pool over a time period. It is a change in vegetation carbon.

[0136] The formula for soil dynamics is expressed as:

[0137]

[0138] In the formula, It is the baseline value of the soil carbon pool. Carbon data based on meteorological impacts Change yes The baseline value, It is the first The impact index of time-dependent micro-nano bubble technology on soil carbon pool It is a change in soil carbon.

[0139] The formula for the dynamic changes of dairy cows is expressed as:

[0140]

[0141] In the formula, This is the baseline value for dairy cow carbon stocks. It is the change in body weight of a single dairy cow based on the carbon data affected by meteorological conditions. yes The baseline value, It is the first The impact of time-based stocking on dairy cow carbon pool index. It refers to the carbon changes in dairy cows.

[0142] The formula for the dynamic change of waste is expressed as:

[0143]

[0144] In the formula, It is the baseline value for the waste carbon pool. Carbon data based on meteorological impacts Change yes The baseline value, It is the first The impact index of yucca extract on waste carbon pool. It refers to carbon changes in waste.

[0145] The dynamic changes of vegetation carbon pool, soil carbon pool, dairy cow carbon pool and waste carbon pool are calculated based on the dynamic change formula and integrated to obtain the change in total carbon storage.

[0146] By altering meteorological conditions and aquaculture factors, the study simulates changes in total carbon storage under different scenarios, analyzes the trends over time, and identifies periodic and trend characteristics to obtain the patterns of carbon storage change.

[0147] The steps involved in developing a carbon management strategy include:

[0148] By analyzing the patterns of carbon storage changes over time, considering factors of contribution, and considering scenario thresholds, we can identify problems at different stages of the carbon cycle.

[0149] Based on the impact on costs, issues at different stages are prioritized to set overall goals, which are then broken down into phased goals to match the time-varying patterns of carbon reserves.

[0150] Based on the changing characteristics of carbon storage over a preset time period, carbon loss prevention and control measures and carbon accumulation enhancement measures are formulated, and differentiated measures are developed according to the contribution of factors. Boundary management schemes are developed based on scenario thresholds, and these are integrated to form a carbon management strategy.

[0151] Carbon loss prevention and control measures can include weather adaptability measures: install sunshade nets + spray cooling systems in the cowshed to control the temperature in the shed within 25℃. Adjust the irrigation period of the vegetation to alleviate the decline in GPP caused by drought.

[0152] Carbon accumulation strengthening measures can include soil carbon pool improvement: after winter forage harvesting, all residues are returned to the field, combined with microbial carbon fixation agents to improve soil carbon conversion rate.

[0153] Control the soil irrigation amount to maintain the soil water content at 60% of the field water holding capacity.

[0154] Stable carbon pool of dairy cows: add 1% betaine to the diet to reduce body weight loss caused by low temperature.

[0155] The number of animals is adjusted quarterly (increased to 7800 in winter and decreased to 7500 in spring) to match the forage supply capacity of the vegetation growth period.

[0156] The boundary management scheme can include threshold control of the number of animals: establish a "stock-crop processing capacity" linkage mechanism: when the number of animals reaches 7800, start the expansion of the manure treatment system.

[0157] Implement "precise stocking": combined with monthly milk production, eliminate low-yielding cows (annual milk production <8 tons per head) to maintain the optimal range of 7800-8000 head.

[0158] Diet CP threshold control: establish a "CP-dairy cow performance" monitoring system: when CP falls to 14.5%, test the milk protein rate (≥3.1%) and body weight change (weekly increase ≥0.5 kg per head) every week, based on the "CP <14%→\(C_{animal}\) decreases" rule.

[0159] As shown in Figure 3 Based on the same overall inventive concept, the present application also protects an ecological ranch carbon storage dynamic prediction system based on weather data, the dynamic prediction system comprising:

[0160] A carbon storage calculation module for collecting weather data, dairy farming factors and ranch parameter data in the target ecological ranch area, constructing a carbon storage calculation model, and calculating the basic carbon storage of the ranch according to the ranch parameter data.

[0161] A carbon cycle impact estimation module for analyzing the impact of ranch parameter data on the target ecological ranch area to obtain carbon cycle parameters, and calculating the carbon storage impact index according to the impact of dairy farming factors on the basic carbon storage of the ranch.

[0162] A carbon cycle impact module for constructing a multi-task learning model optimized based on the secretary bird algorithm, inputting carbon cycle parameters and weather data, and outputting weather-affected carbon data.

[0163] The carbon management strategy module is configured to generate a carbon storage change rule according to the carbon storage influence index and meteorological influence carbon data, and to formulate a carbon management strategy according to the carbon storage change rule.

[0164] The ecological pasture carbon storage dynamic prediction method and system based on meteorological data provided by the embodiment effectively improves data quality and is more in line with the actual carbon storage situation of ecological pastures. Based on the carbon flow rule of ecological pastures, the carbon circulation paths among different carbon storage carriers (vegetation, soil, cows, and waste) are analyzed, the correlation degree of pasture parameter data and carbon circulation process indicators is calculated through correlation analysis, the influence weight is determined, and the process parameters of carbon input, carbon output, and carbon conversion are quantified. This method makes the carbon circulation parameters more in line with the actual situation of pastures, and provides a reliable basis for subsequent carbon storage dynamic prediction. Through the carbon storage influence index, the breeding factors are included in the carbon storage dynamic prediction framework. By combining meteorological influence carbon data with the carbon storage influence index, the "breeding-meteorological coordination" carbon storage change rule is generated. This method is more in line with the actual operation scenario of ecological pastures, and the prediction result is more practical. The carbon storage change rule is also disassembled to identify problems at different links of the carbon cycle, and a boundary management scheme is formulated in combination with the scene threshold. This strategy effectively improves the efficiency and effectiveness of carbon management.

[0165] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments can be implemented by means of software and the necessary general hardware platform, and of course, can also be implemented by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes a plurality of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute the method described in each embodiment or some parts of the embodiment.

[0166] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for dynamically predicting carbon storage of an ecological pasture based on meteorological data, characterized in that, The method comprises the following steps: Collecting meteorological data, cow breeding factors and pasture parameter data of a target ecological pasture area, constructing a carbon storage calculation model, and calculating the basic carbon storage of the pasture according to the pasture parameter data; The step of constructing the carbon storage calculation model comprises: For different forms of carbon storage, determine the corresponding influence factors and carbon conversion paths as carbon characteristic parameters; According to the carbon characteristic parameters, combine the correlation of vegetation growth and meteorological factors to construct a vegetation sub-model, and calculate the soil organic carbon storage according to the differences in soil layers and temperature decomposition depth to construct a soil sub-model; According to the influence of body weight, inventory and temperature on the carbon storage in the body of a cow, construct a cow carbon storage, and combine the carbon loss law between manure and stubble to correlate temperature and wind speed to construct a waste sub-model; Couple the vegetation sub-model, the soil sub-model, the cow carbon storage and the waste sub-model to form the carbon storage calculation model; Analyze the influence of the pasture parameter data on the target ecological pasture area to obtain carbon cycle parameters, and calculate the carbon storage influence index according to the influence of the cow breeding factors on the basic carbon storage of the pasture; Construct a multi-task learning model optimized based on the secretary bird algorithm, input the carbon cycle parameters and the meteorological data, and output meteorological influence carbon data; According to the carbon storage influence index and combining the meteorological influence carbon data, generate a carbon storage change rule, and formulate a carbon management strategy according to the carbon storage change rule. 2.The method of claim 1, wherein, The step of calculating the basic carbon storage of the pasture comprises: Using the Laplace criterion to remove data exceeding the preset multiple standard deviation from the pasture parameter data, and using the average value of the same type of pasture in the same region to fill in the missing values to obtain pasture processing parameter data; Input the pasture processing parameter data into the carbon storage calculation model, calculate the corrected carbon storage of vegetation, soil, cows and waste, and sum up to obtain the basic carbon storage of the pasture. 3.The method of claim 2, wherein, The step of obtaining the carbon cycle parameters comprises: Based on the carbon flow rule of ecological pasture, analyze the carbon cycle path between different carbon storage carriers, classify the pasture parameter data, and establish a circulation parameter correspondence; Using correlation analysis method to calculate the correlation degree of the pasture parameter data and the carbon cycle process index, so as to determine the influence weight; According to the circulation parameter correspondence and the influence weight, calculate the process parameters of carbon input, carbon output and carbon conversion as the carbon cycle parameters. 4.The method of claim 3, wherein, The step of calculating the carbon storage influence index comprises: Based on the factor adjustment value of the actual breeding scene, calculating the total carbon storage according to the association logic between the cow breeding factors and the carbon storage carriers; According to the basic carbon storage of the pasture and the total carbon storage, calculate the relative change amplitude of the carbon storage adjusted according to the factor adjustment value for different breeding factors to obtain the carbon storage influence index. 5.The method of claim 1, wherein, The step of constructing the multi-task learning model comprises: According to the feature extraction network shared by multiple tasks, learn the common features between different tasks as shared layers, and learn the task features of each task independent output layer as task specific layers; According to the shared layer and the task-specific layer, a model framework is constructed, and to-be-optimized parameters of the model framework are encoded as secretary bird individual positions, and a comprehensive loss function is selected as a fitness function; According to the upper and lower bounds of the to-be-optimized parameters, a plurality of secretary bird individuals are randomly generated; The process of secretary birds searching for prey is simulated, and secretary bird searching positions are generated through differential mutation; Local search is combined with Brownian motion to simulate the precise attack of secretary birds and generate secretary bird digestion positions and secretary bird attack positions; The process of secretary birds avoiding predators is simulated, and a dynamic disturbance strategy is adopted to balance exploration and development to obtain secretary bird escape positions; When a preset number of iterations is reached, the update is stopped, the fitness values of different secretary bird positions are calculated according to the fitness function, and to-be-optimized parameters with the smallest fitness value are selected to construct the multi-task learning model. 6.The method of claim 5, wherein, The step of outputting the meteorological influence carbon data includes: Aligning the time scales of the meteorological data and the carbon cycle parameters, and performing standardization processing to input the multi-task learning model; Learning the common features of the meteorological data according to the shared layer, and calculating the prediction accuracy index of each carbon cycle parameter according to the task-specific layer; Quantifying the influence degree of a single meteorological factor on the carbon cycle parameter to obtain a sensitivity coefficient through the control variable method; For each carbon cycle parameter, a relationship curve with meteorological factors is output, and the sensitivity coefficients are sorted according to the absolute values to obtain the meteorological influence carbon data. 7.The method of claim 1, wherein, The step of generating the carbon storage change rule includes: According to a preset time period as a link, the meteorological influence carbon data and the carbon storage influence index are associated with the pasture basic carbon storage to determine the influence path; According to the influence path, the climate influence and the breeding influence are integrated to adjust the carbon storage calculation model to obtain a dynamic change formula; According to the dynamic change formula, the dynamic changes of the vegetation carbon pool, the soil carbon pool, the dairy cow carbon pool and the waste carbon pool are calculated and integrated to obtain the total carbon storage change; By changing the meteorological conditions and the breeding factors, the total carbon storage change under different scenarios is simulated, and the change trend over time is analyzed to identify periodic and trend characteristics to obtain the carbon storage change rule. 8.The method of claim 1, wherein, The step of formulating the carbon management strategy includes: Decompose the carbon storage change rule from the time dimension, factor contribution and scenario threshold to determine the problems in different links of the carbon cycle; According to the influence cost, the problems in different links are sorted to set overall goals and decompose them into phased goals, matching the time variation rule of carbon storage; For the change characteristics of carbon storage in a preset time period, carbon loss prevention and control measures and carbon accumulation strengthening measures are formulated, differentiated measures are formulated according to the factor contribution, boundary management schemes are formulated according to the scenario threshold, and the carbon management strategy is integrated.

9. A system for dynamically predicting carbon stock of an ecological pasture based on meteorological data, which is applied to the method for dynamically predicting carbon stock of an ecological pasture based on meteorological data according to any one of claims 1 to 8, characterized in that, The dynamic prediction system includes: A carbon storage calculation module is configured to collect meteorological data, dairy cow breeding factors and pasture parameter data of a target ecological pasture area, construct a carbon storage calculation model, and calculate a pasture basic carbon storage according to the pasture parameter data; The carbon cycle influence estimation module is configured to analyze the influence of the pasture parameter data on the target ecological pasture area to obtain carbon cycle parameters, and calculate a carbon storage influence index according to the influence of the dairy cattle breeding factors on the pasture base carbon storage; The gas carbon influence module is configured to construct a multi-task learning model optimized based on a secretary bird algorithm, input the carbon cycle parameters and the meteorological data, and output meteorological influence carbon data; The carbon regulation strategy formulation module is configured to generate a carbon storage change rule according to the carbon storage influence index and in combination with the meteorological influence carbon data, and formulate a carbon management strategy according to the carbon storage change rule.

Images