Determination method, application method and device of saline-alkali soil carbon income and expenditure evaluation and prediction model
By combining ecosystem process models and machine learning algorithms, the optimal integrated model is established, and the problem of uncertainty in the carbon revenue and expenditure simulation results of saline-alkali land is solved, and high-precision simulation of the carbon revenue and expenditure dynamics of saline-alkali land and the assessment of the impact of climate change are achieved, supporting the accurate prediction of carbon sinks and carbon reserves in saline-alkali land.
Patent Information
- Application Number
- CN202510478516.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-16
- Publication Date
- 2025-08-01
AI Technical Summary
The prior art cannot accurately simulate and predict the dynamics of carbon revenue and expenditure in saline-alkali land, especially because the impact of soil salinity and alkalinization on plant growth, respiration and soil organic matter decomposition processes is not considered, resulting in high uncertainty in the simulation results, and it is impossible to predict the dynamic changes in future carbon sinks and carbon storage.
Combining ecosystem process models, multi-source and multi-scale observation data and machine learning algorithms, the optimal integrated model is established through adaptive screening and weighted averaging. As the proxy model of the ecosystem process model, it uses historical and future parameters for training to achieve high-precision simulation and prediction of carbon revenue and expenditure in saline-alkali land.
The rapid, efficient and accurate simulation of carbon revenue and expenditure of different types of saline-alkali land can be achieved, and the carbon revenue and expenditure of saline-alkali land under climate change and its contribution to regional carbon neutrality can be evaluated, improving the reliability and accuracy of simulation and prediction.
Smart Images
Figure CN120409908A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of ecological carbon sink assessment and prediction, and in particular to a method for determining, an application method and a device for a saline-alkali land carbon budget assessment and prediction model. Background Art
[0002] The carbon balance of saline-alkali soils is a crucial component of the global carbon budget and has a significant impact on agricultural productivity. Understanding the cycling patterns of organic carbon in saline-alkali soils, revealing the strength, properties, and dynamics of their carbon sink function, as well as their response to climate change and human activities, and assessing the role and significance of the saline-alkali soil carbon cycle in the global carbon cycle are pressing scientific challenges.
[0003] Research on the carbon budget and carbon cycle dynamics of saline-alkali land includes field surveys, satellite remote sensing observations, and multi-scale simulations of ecosystem process models. Field survey observations can provide the most accurate data, but due to the diverse types of saline-alkali land and the influence of climate and human activities, the distribution of its carbon sinks and carbon reserves has a high degree of temporal and spatial variability. Monitoring carbon sinks and carbon reserves in saline-alkali land over large areas requires a lot of manpower, material resources, and time, and monitoring methods cannot predict the dynamic changes of carbon sinks and carbon reserves in saline-alkali land in the future. Satellite remote sensing observation technology provides dynamic and continuous monitoring data with multiple temporal and spatial resolutions for monitoring ecosystem structure and function. Currently, carbon cycle parameters based on remote sensing inversion include gross primary productivity (GPP) and net primary productivity (NPP). However, remote sensing monitoring cannot directly observe and invert ecosystem carbon sinks and carbon reserves, and remote sensing technology is also unable to achieve future predictions. The carbon sink refers to net ecosystem productivity (NEP), and the carbon reserves refer to soil organic carbon density (SOC) and vegetation carbon density (VC). Ecosystem process models, by comprehensively considering the relative contributions of ecosystem processes to the environment and human activities, can simulate the carbon cycle in saline-alkali soils at multiple scales. This helps to reveal the interactions and coupling relationships within the saline-alkali soil carbon cycle at different scales. However, due to limitations in parameter acquisition and model structure, the simulation results of ecosystem process models still suffer from significant uncertainty. Furthermore, most current ecosystem process models fail to account for the impact of soil salinity and alkalinity on plant growth, respiration, and soil organic matter decomposition in their simulations of the carbon budget for different saline-alkali soil types. Consequently, they are unable to provide accurate and reliable simulations and predictions of the carbon budget for saline-alkali soils. Summary of the Invention
[0004] The purpose of this application is to provide a method for determining, applying and predicting a saline-alkali land carbon balance assessment model, which can quickly, efficiently and accurately simulate the spatiotemporal dynamics of carbon balances of different types of saline-alkali lands and their responses to climate change, thereby accurately assessing the carbon balance of saline-alkali lands under climate change and their contribution to regional carbon neutrality.
[0005] To achieve the above object, the present application provides the following solutions:
[0006] In a first aspect, the present application provides a method for determining a carbon budget assessment and prediction model for saline-alkali land. The method for determining the carbon budget assessment and prediction model for saline-alkali land includes:
[0007] Obtain the first relevant parameters and the second relevant parameters of the target area; the first relevant parameters include: historical climate data, atmospheric CO2 concentration data, vegetation type, and soil particle composition data; the second relevant parameters include: the net primary productivity interpreted from satellite remote sensing data and the mean value of the relevant data within the 1m soil layer; the mean value of the relevant data within the 1m soil layer is the data calculated by weighted averaging the soil pH, electrical conductivity (ECE), and exchangeable sodium percentage (ESP) data of the surface layer (0-20cm) and the lower layer (20-100cm) with a 1km resolution according to the FAO World Soil Database (HWSD).
[0008] Input the first relevant parameters into an ecosystem process model to obtain the preliminary results of each component of the carbon budget in the target area; the preliminary results include: net ecosystem productivity, soil organic carbon density, and vegetation carbon density.
[0009] Construct a data set based on the first relevant parameters, the second relevant parameters, and the preliminary results.
[0010] Based on the data set, use the first relevant parameters and the second relevant parameters as inputs and the preliminary results as labels to train a machine learning model to obtain a carbon budget assessment and prediction model for saline-alkali land.
[0011] In a second aspect, the present application provides a method for applying a carbon budget assessment and prediction model for saline-alkali land. The method for applying the carbon budget assessment and prediction model for saline-alkali land includes:
[0012] Obtain the future first relevant parameter and the future second relevant parameter of the target area; the future first relevant parameter includes: future climate data, future atmospheric CO2 concentration data, future vegetation type, and future soil particle composition data; the future climate data and the future atmospheric CO2 concentration data are data obtained based on the Earth system model; the future vegetation type is historical vegetation type data; the future soil particle composition data is historical soil particle composition data; the second relevant parameter includes: the future net primary productivity interpreted based on satellite remote sensing data and the average value of relevant data within the future 1m soil layer; the future net primary productivity interpreted based on satellite remote sensing data is data predicted by using the time series autoregressive prediction method based on historical remotely sensed observed net primary productivity data; the average value of relevant data within the future 1m soil layer is data calculated by weighted average according to the surface layer and lower layer soil pH, conductivity, and alkalinity data with a 1km resolution of the historical FAO World Soil Database.
[0013] Input the future first relevant parameter and the future second relevant parameter into the saline-alkali land carbon budget assessment and prediction model to obtain the future net ecosystem productivity, the future soil organic carbon density, and the future vegetation carbon density of the target area; the saline-alkali land carbon budget assessment and prediction model is a model trained based on the above-mentioned method for determining the saline-alkali land carbon budget assessment and prediction model.
[0014] In a third aspect, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the computer program to implement the above-mentioned method for determining the saline-alkali land carbon budget assessment and prediction model or the application method of the saline-alkali land carbon budget assessment and prediction model.
[0015] In a fourth aspect, the present application provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the above-mentioned method for determining the saline-alkali land carbon budget assessment and prediction model or the application method of the saline-alkali land carbon budget assessment and prediction model.
[0016] In a fifth aspect, the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the above-mentioned method for determining the saline-alkali land carbon budget assessment and prediction model or the application method of the saline-alkali land carbon budget assessment and prediction model.
[0017] According to the specific embodiments provided by the present application, the following technical effects are disclosed in the present application:
[0018] The present application provides a method, an application method and a device for determining a carbon budget assessment and prediction model for saline-alkali land. First, first relevant parameters and second relevant parameters of a target area are obtained; the first relevant parameters include: historical climate data, atmospheric CO2 concentration data, vegetation type and soil particle composition data; the second relevant parameters include: net primary productivity interpreted from satellite remote sensing data and the mean value of relevant data within the 1m soil layer; the mean value of the relevant data within the 1m soil layer is data calculated by weighted averaging based on the surface and subsurface soil pH, electrical conductivity and alkalinity data with a 1km resolution from the FAO World Soil Database. Secondly, the first relevant parameters are input into an ecosystem process model to obtain preliminary results of each component of the carbon budget of the target area; the preliminary results include: net ecosystem productivity, soil organic carbon density and vegetation carbon density. Then, based on the first relevant parameters, the second relevant parameters and the preliminary results, a data set is constructed; based on the data set, with the first relevant parameters and the second relevant parameters as inputs and the preliminary results as labels, a machine learning model is trained to obtain a carbon budget assessment and prediction model for saline-alkali land. Through training with multiple machine learning models, the present application can obtain an optimal integrated model and use it as a proxy model for the ecosystem process model. The present application solves the defects in the prior art that the carbon sink and carbon storage dynamics of saline-alkali land cannot be directly obtained based on remote sensing observations, and the simulation results of the carbon budget of saline-alkali land based on the ecosystem process model have great uncertainties, and can quickly, efficiently and accurately simulate the spatio-temporal dynamics of the carbon budget of different types of saline-alkali land and their responses to climate change, so as to accurately evaluate the carbon budget of saline-alkali land under climate change and its contribution to regional carbon neutrality. Description of the Drawings
[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.
[0020] Figure 1 It is an application environment diagram of a method for determining a carbon budget assessment and prediction model for saline-alkali land in an embodiment of the present application.
[0021] Figure 2 It is a flowchart of a method for determining a carbon budget assessment and prediction model for saline-alkali land provided in an embodiment of the present application.
[0022] Figure 3 It is a flowchart of an application method of a carbon budget assessment and prediction model for saline-alkali land provided in an embodiment of the present application.
[0023] Figure 4Schematic diagram of the overall process of a method for evaluating and predicting the carbon budget of saline-alkali land provided by an embodiment of the present application.
[0024] Figure 5 Schematic diagram of the structure of a system for evaluating and predicting the carbon budget of saline-alkali land provided by an embodiment of the present application.
[0025] Figure 6 Schematic diagram of the structure of a computer device provided by an embodiment of the present application. Detailed implementation manners
[0026] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.
[0027] To make the above objects, features, and advantages of the present application more obvious and understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.
[0028] Machine learning can achieve higher accuracy and simulation efficiency than process models at large scales. The large amount of observational data accumulated from the leaf to the global scale in the past few decades has also provided a large amount of high-quality data for the training of machine learning algorithms. By combining the prior knowledge implicit in the process model with the learning ability of machine learning and using a large amount of multi-source and multi-scale observational data, rapid simulation and prediction at high-resolution spatio-temporal scales can be achieved, while improving the reliability and accuracy of the simulation and prediction of the carbon budget of saline-alkali land.
[0029] At present, satellite remote sensing technology provides GPP and NPP data products with high spatio-temporal resolution, but it cannot directly observe and invert the carbon sink (NEP) and carbon storage (SOC and VC) of ecosystems. Although ecosystem process models can provide simulation outputs and future predictions of carbon fluxes (GPP, Ra, NPP, Rh, RE, NEP) and carbon storage (SOC, VC) with various spatio-temporal resolutions for different regions, due to reasons such as parameter acquisition and model structure, there are still large uncertainties in their simulation results. Especially in the simulation of carbon budgets in different types of saline-alkali lands, because most models fail to consider the effects of soil salt content, alkalinity, etc. on processes such as plant growth, respiration, and soil organic matter decomposition, there are also large uncertainties in the simulation of saline-alkali lands. The purpose of this application is to overcome the above technical defects and provide a high-precision spatio-temporal dynamic simulation and prediction method for carbon budgets in saline-alkali lands that couples ecosystem process models, multi-source multi-scale observation data, and machine learning algorithms. This application couples ecosystem process models, multi-source multi-scale observation data, and various machine learning algorithms, and establishes an optimal integrated model through adaptive screening and weighted averaging as a proxy model for ecosystem process models, so as to provide technical methods for accurately simulating and predicting the dynamics of carbon budgets in different types of saline-alkali lands, averaging the contribution of saline-alkali land carbon sinks to regional carbon neutrality, and analyzing the impact of climate change on saline-alkali land carbon sinks and carbon storage.
[0030] The method for determining the saline-alkali land carbon budget assessment and prediction model provided by the embodiments of this application can be applied to, for example Figure 1In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be set separately, integrated on the server 104, placed on the cloud or other servers. The terminal 102 can send the first relevant parameters and the second relevant parameters of the target area to the server 104. The first relevant parameters include: historical climate data, atmospheric CO2 concentration data, vegetation type, and soil particle composition data. The second relevant parameters include: net primary productivity interpreted from satellite remote sensing data and the mean value of relevant data within 1 m soil layer. The mean value of the relevant data within 1 m soil layer is the data calculated by weighted average based on the surface and subsurface soil pH, conductivity, and alkalinity data with a resolution of 1 km in the FAO World Soil Database. After receiving the first relevant parameters and the second relevant parameters, for the first relevant parameters and the second relevant parameters, the server 104 inputs the first relevant parameters into the ecosystem process model to obtain the preliminary results of each component of the carbon budget in the target area. The preliminary results include: net ecosystem productivity, soil organic carbon density, and vegetation carbon density. Based on the first relevant parameters, the second relevant parameters, and the preliminary results, a data set is constructed. Based on the data set, with the first relevant parameters and the second relevant parameters as inputs and the preliminary results as labels, a machine learning model is trained to obtain a saline-alkali land carbon budget assessment and prediction model. The server 104 can feedback the obtained saline-alkali land carbon budget assessment and prediction model to the terminal 102. In addition, in some embodiments, the method for determining the saline-alkali land carbon budget assessment and prediction model can also be implemented by the server 104 or the terminal 102 alone. For example, the terminal 102 can directly perform the saline-alkali land carbon budget assessment and prediction for the first relevant parameters and the second relevant parameters of the target area, or the server 104 can obtain the first relevant parameters and the second relevant parameters of the target area from the data storage system and perform the saline-alkali land carbon budget assessment and prediction for the first relevant parameters and the second relevant parameters of the target area.
[0031] Among them, the terminal 102 can be, but is not limited to, various desktop computers, laptop computers, smart phones, and tablet computers. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.
[0032] In an exemplary embodiment, as Figure 2 shown, a method for determining a saline-alkali land carbon budget assessment and prediction model is provided. This method is executed by a computer device, and can be specifically executed by a computer device such as a terminal or a server alone, or jointly executed by a terminal and a server. In the embodiments of the present application, taking this method applied to Figure 1 the server 104 in
[0033] A1: Obtain the first relevant parameter and the second relevant parameter of the target area; the first relevant parameter includes: historical climate data, atmospheric CO2 concentration data, vegetation type, and soil particle composition data; the second relevant parameter includes: the net primary productivity interpreted from satellite remote sensing data and the mean value of relevant data within the 1m soil layer; the mean value of relevant data within the 1m soil layer is the data calculated by weighted averaging based on the surface and subsurface soil pH, conductivity, and alkalinity data with a 1km resolution from the FAO World Soil Database.
[0034] A2: Input the first relevant parameter into the ecosystem process model to obtain the preliminary results of each component of the carbon budget in the target area; the preliminary results include: net ecosystem productivity, soil organic carbon density, and vegetation carbon density.
[0035] A3: Construct a dataset based on the first relevant parameter, the second relevant parameter, and the preliminary results.
[0036] A4: Based on the dataset, use the first relevant parameter and the second relevant parameter as inputs and the preliminary results as labels to train a machine learning model to obtain a carbon budget assessment and prediction model for saline-alkali land.
[0037] Implementing the above steps A1 to A4 can solve the defects in the prior art that it is impossible to directly obtain the dynamic changes of the carbon sink and carbon storage in saline-alkali land based on remote sensing observations, and there are great uncertainties in the simulation results of the carbon budget in saline-alkali land based on the ecosystem process model. It can quickly, efficiently, and accurately simulate the spatio-temporal dynamics of the carbon budget in different types of saline-alkali land and their responses to climate change, so as to accurately evaluate the carbon budget of saline-alkali land under climate change and its contribution to regional carbon neutrality.
[0038] As an optional implementation manner, in step A1, the historical climate data includes temperature, precipitation, relative humidity, total radiation, and net radiation at dekadal scale; the vegetation type is the land use / land cover data interpreted from satellite remote sensing data, including 13 vegetation types such as evergreen coniferous forest, evergreen broad-leaved forest, deciduous coniferous forest, deciduous broad-leaved forest, coniferous and broad-leaved mixed forest, forest land, closed shrubland, open shrubland, grassland, sparse grassland, farmland, wetland, desert, and other land cover types without vegetation. The soil particle composition data includes the composition of soil clay, silt, and sand.
[0039] As an alternative implementation, in step A2, the first relevant parameters are input into the ecosystem process model - the CEVSA2 model to obtain preliminary results of the components of the carbon budget in the target area. Among them, in addition to net ecosystem productivity, soil organic carbon density, and vegetation carbon density, the outputs of CEVSA2 also include gross primary productivity, net primary productivity, and soil heterotrophic respiration (Rh) on an annual scale.
[0040] It should be noted that the method for determining the saline-alkali land carbon budget assessment and prediction model further includes: uniformly interpolating the first relevant parameters and the second relevant parameters to the same spatial resolution.
[0041] As an alternative implementation, in step A3, the first relevant parameters, the second relevant parameters, and the preliminary results are integrated to construct a data set; among them, 80% is the training data set and 20% is the validation data set, which are used to train different types of machine learning models to enable the machine learning models to fully learn and simulate the changes in state variables simulated by the ecosystem process model; evaluate the performance of machine learning, adaptively eliminate the machine learning models with poor performance, and obtain the optimal integrated model according to the weighted evaluation method with different model weights.
[0042] As an alternative implementation, in step A4, it specifically includes:
[0043] A41: Input historical climate data, atmospheric CO2 concentration data, vegetation type, soil particle composition data, net primary productivity interpreted from satellite remote sensing data, and the mean value of relevant data within the 1m soil layer into several machine learning models to obtain the outputs of several machine learning models.
[0044] A42: Based on the output of each machine learning model and the preliminary results, calculate the root mean square error and coefficient of determination of each machine learning model respectively.
[0045] A43: Remove the models with a coefficient of determination lower than the average level of the corresponding machine learning model to obtain the selected models.
[0046] A44: Based on the root mean square error, determine the weights of the corresponding selected models.
[0047] A45: Construct a weighted average integrated model based on the weights of the selected models to obtain the saline-alkali land carbon budget assessment and prediction model.
[0048] Specifically, the machine learning models include support vector machine, random forest, XGBoost, convolutional neural network (CNN), and long short-term memory network (LSTM). Calculate the root mean square error RMSE and coefficient of determination R respectively 2 to evaluate the performance of each machine learning model in predicting the output of the process model. By setting screening criteria, R2 Models with performance below the average of all models will be deprecated, and several models with performance above the average will be selected; based on the model performance, weights will be calculated for each selected model, with the weight being 1 / RMSE, and a weighted average ensemble model will be constructed to obtain the optimal ensemble model, which is the saline-alkali land carbon budget assessment prediction model.
[0049] The calculation formula for the root mean square error is as follows:
[0050]
[0051] where RMSE is the root mean square error; y i is the measured value, that is, the corresponding preliminary result; is the predicted value, that is, the output of the corresponding machine learning model; n is the number of samples.
[0052] The calculation formula for the coefficient of determination is as follows:
[0053]
[0054] where R 2 is the coefficient of determination; y i is the measured value, that is, the corresponding preliminary result; is the predicted value, that is, the output of the corresponding machine learning model; is the average value of the measured values; n is the number of samples.
[0055] In an exemplary embodiment, historical climate data, atmospheric CO2 concentration data, vegetation types, soil particle composition data, net primary productivity interpreted from satellite remote sensing data, and the mean value of relevant data within the 1m soil layer of the target area are input into the saline-alkali land carbon budget assessment prediction model, and through parallel computing, the carbon budget (net ecosystem productivity) and carbon storage (soil organic carbon density and vegetation carbon density) of the saline-alkali land in the target area can be simulated and output.
[0056] In another exemplary embodiment of the present application, as Figure 3 shown, a method for applying the saline-alkali land carbon budget assessment prediction model is provided, and the method for applying the saline-alkali land carbon budget assessment prediction model includes:
[0057] B1: Obtain the future first relevant parameters and future second relevant parameters of the target area; the future first relevant parameters include: future climate data, future atmospheric CO2 concentration data, future vegetation type, and future soil particle composition data; the future climate data and the future atmospheric CO2 concentration data are data obtained based on an Earth system model; the future vegetation type is historical vegetation type data; the future soil particle composition data is historical soil particle composition data; the second relevant parameters include: future net primary productivity interpreted from satellite remote sensing data and the mean value of relevant data within the future 1m soil layer; the future net primary productivity interpreted from satellite remote sensing data is data predicted using the time series autoregressive prediction method based on historical remote sensing observed net primary productivity data; the mean value of relevant data within the future 1m soil layer is data obtained by weighted average calculation based on the surface and subsurface soil pH, conductivity, and alkalinity data with a 1km resolution from the historical FAO World Soil Database.
[0058] B2: Input the future first relevant parameters and the future second relevant parameters into the saline-alkali land carbon budget assessment and prediction model to obtain the future net ecosystem productivity, future soil organic carbon density, and future vegetation carbon density of the target area; the saline-alkali land carbon budget assessment and prediction model is a model trained based on the determination method of the saline-alkali land carbon budget assessment and prediction model described in any one of the above.
[0059] As an alternative implementation, in step B1, future NPP data is predicted using the time series autoregressive prediction method based on historical remote sensing observed NPP data. MODIS data has advantages such as wide coverage and easy access, and the quality of the data products is highly accurate and reliable. Download the NPP data product of MOD173AH since 2001 on the GEE platform, with an annual time resolution and a spatial resolution consistent with the above spatial data. First, use the S-G filtering method for processing to eliminate the noise of the original time series data. Use the ADF test to determine whether the data is stationary. If it is not stationary, it is necessary to perform differencing (such as first-order differencing) until it is stationary. Plot the original data curve to observe the long-term trend and interannual volatility. Plot the autocorrelation function (ACF) and partial autocorrelation function (PACF) for model identification, determine the order p, and construct the autoregressive (AR) model AR(p):
[0060]
[0061] where, y t is the NPP observation value at time t; c is the constant term; is the autoregressive coefficient; ∈ t is the white noise error term. By minimizing the sum of squared residuals, solve for the coefficients and c.
[0062] After the model is constructed, model diagnosis is carried out: draw the residual ACF graph. If there is no significant correlation (p-value > significance level), the residuals are approximately white noise. Use the QQ plot or Jarque-Bera test to judge whether the residuals conform to the normal distribution. Check whether the mean is close to 0 and whether the variance is stable (no obvious heteroscedasticity). If the diagnosis fails, the p-value needs to be readjusted.
[0063] Based on the diagnosed autoregressive time series model AR(p), predict the future NPP. Based on the NPP in the most recent year, substitute it into the model to calculate the predicted value for the next year, and recursively use the predicted results as input to gradually generate the NPP values for future years.
[0064] Future climate data comes from the outputs of 11 Earth system models in CMIP6. These Earth system models include Access-CM2, Access-ESM1-5, CanESM5, CMCC-ESM2, GFDL-ESM4, INM-CM4-8, INM-CM5-0, MIROC, MPI-ESM1-2-LR, MRI-ESM2-0, TaiESM1. It includes atmospheric CO2 data, air temperature, precipitation, relative humidity, total radiation, and net radiation data under four scenarios: SSP126, SSP245, SSP370, and SSP585. The climate data is spatially downscaled using Anuspline software, and the spatial resolution is consistent with the spatial data of the first relevant parameter and the second relevant parameter, and is calculated from daily means to dekadal means and totals.
[0065] Therefore, in step B2, input the future first relevant parameter and the future second relevant parameter into the saline-alkali land carbon budget assessment and prediction model, and the final results of the dynamic simulation of the saline-alkali land carbon budget under different future climate change scenarios in the target area can be obtained. The overall process of the saline-alkali land carbon budget assessment and prediction method is as Figure 4 shown.
[0066] The present application also provides an application scenario, which applies the above-mentioned application method of the saline-alkali land carbon budget assessment and prediction model. Specifically: The application method of the saline-alkali land carbon budget assessment and prediction model provided in this embodiment can be applied in the ecological carbon sink assessment and prediction scenario. The ecological carbon sink assessment and prediction scenario includes: a data acquisition link, a model training link, and a prediction link; First, obtain the first relevant parameters and the second relevant parameters of the target area; The first relevant parameters include: historical climate data, atmospheric CO2 concentration data, vegetation type, and soil particle composition data; The second relevant parameters include: the net primary productivity interpreted based on satellite remote sensing data and the mean value of the relevant data within the 1m soil layer; The mean value of the relevant data within the 1m soil layer is the data calculated by weighted averaging the surface and subsurface soil pH, electrical conductivity, and alkalinity data with a 1km resolution from the FAO World Soil Database; Input the first relevant parameters into the ecosystem process model to obtain the preliminary results of each component of the carbon budget in the target area; The preliminary results include: net ecosystem productivity, soil organic carbon density, and vegetation carbon density; Based on the first relevant parameters, the second relevant parameters, and the preliminary results, construct a data set; Use the first relevant parameters and the second relevant parameters as inputs, and the preliminary results as labels to train a machine learning model to obtain a saline-alkali land carbon budget assessment and prediction model; Then, obtain the future first relevant parameters and the future second relevant parameters of the target area; Input the future first relevant parameters and the future second relevant parameters into the saline-alkali land carbon budget assessment and prediction model to obtain the future net ecosystem productivity, future soil organic carbon density, and future vegetation carbon density of the target area.
[0067] Based on the same inventive concept, the embodiment of the present application also provides a saline-alkali land carbon budget assessment and prediction system. The implementation solutions provided by this system to solve problems are similar to the implementation solutions described in the above method. Therefore, the specific limitations in one or more embodiments of the saline-alkali land carbon budget assessment and prediction system provided below can refer to the limitations of the above method in the foregoing text, and will not be elaborated here.
[0068] As Figure 5 shown, a saline-alkali land carbon budget assessment and prediction system is provided. The system includes the following modules:
[0069] Module 1: Data acquisition and processing module, used to obtain the first relevant parameters and the second relevant parameters of the target area; The first relevant parameters include: historical climate data, atmospheric CO2 concentration data, vegetation type, and soil particle composition data; The second relevant parameters include: the net primary productivity interpreted based on satellite remote sensing data and the mean value of the relevant data within the 1m soil layer; The mean value of the relevant data within the 1m soil layer is the data calculated by weighted averaging the surface and subsurface soil pH, electrical conductivity, and alkalinity data with a 1km resolution from the FAO World Soil Database.
[0070] Module 2: Process model simulation module, which is used to input the first relevant parameters into CEVSA2 to obtain the preliminary results of each component of the carbon budget in the target area.
[0071] Module 3: Optimal combination model training module, which is used to integrate and train different types of machine learning models with remote sensing observations, ground observations and model output data to obtain the optimal combination model.
[0072] Module 4: Prediction module, which is used to predict the future NPP data of the target area based on the historical NPP data of remote sensing observations by using the time series autoregressive prediction method; and predict the future first relevant parameter and the future second relevant parameter according to the first relevant parameter and the second relevant parameter.
[0073] Module 5: Carbon budget evaluation and prediction module, which is used to input the future first relevant parameter and the future second relevant parameter into the optimal integration model to obtain the historical evaluation results of the dynamic carbon budget of saline-alkali land in the target area output by the proxy model, and the final prediction results under different future climate change scenarios.
[0074] In an exemplary embodiment, a computer device is provided. The computer device can be a server or a terminal, and its internal structure diagram can be as Figure 6 shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the first relevant parameter and the second relevant parameter. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it is used to implement a method for determining a saline-alkali land carbon budget evaluation and prediction model or a method for applying a saline-alkali land carbon budget evaluation and prediction model.
[0075] Those skilled in the art can understand that Figure 6 the structure shown in merely represents the block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.
[0076] In an exemplary embodiment, a computer device is further provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the above method embodiments are implemented.
[0077] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the above method embodiments are implemented.
[0078] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the above method embodiments are implemented.
[0079] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.
[0080] Those of ordinary skill in the art can understand that all or part of the processes of implementing the above method embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above method embodiments. Among them, any reference to a memory, database, or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memories. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.
[0081] In each of the embodiments provided in the present application, the databases involved may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on a blockchain, etc., without limitation. The processors involved in each of the embodiments provided in the present application may be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without limitation.
[0082] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.
[0083] In this article, specific examples are used to elaborate on the principles and implementation manners of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those of ordinary skill in the art, according to the idea of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.
Claims
1. A method for determining a carbon budget assessment and prediction model for saline-alkali land, characterized in that The method for determining the saline-alkali land carbon budget assessment and prediction model includes: Obtaining the first relevant parameters and the second relevant parameters of the target area; the first relevant parameters include: historical climate data, atmospheric CO2 concentration data, vegetation type, and soil particle composition data; the second relevant parameters include: the net primary productivity interpreted from satellite remote sensing data and the mean value of relevant data within the 1m soil layer; the mean value of relevant data within the 1m soil layer is the data calculated by weighted averaging based on the surface layer and subsurface soil pH, electrical conductivity, and alkalinity data with a 1km resolution from the FAO World Soil Database. Inputting the first relevant parameters into the ecosystem process model to obtain the preliminary results of each component of the carbon budget in the target area; the preliminary results include: net ecosystem productivity, soil organic carbon density, and vegetation carbon density. Based on the first relevant parameters, the second relevant parameters, and the preliminary results, constructing a data set. Based on the data set, using the first relevant parameters and the second relevant parameters as inputs and the preliminary results as labels, training a machine learning model to obtain the saline-alkali land carbon budget assessment and prediction model.
2. The method for determining the saline-alkali land carbon budget assessment and prediction model according to claim 1, characterized in that Using the first relevant parameters and the second relevant parameters as inputs and the preliminary results as labels, training a machine learning model to obtain the saline-alkali land carbon budget assessment and prediction model, specifically including: Inputting the historical climate data, atmospheric CO2 concentration data, vegetation type, soil particle composition data, the net primary productivity interpreted from satellite remote sensing data, and the mean value of relevant data within the 1m soil layer into several machine learning models to obtain the outputs of several machine learning models. Based on the output of each machine learning model and the preliminary results, calculating the root mean square error and the coefficient of determination of each machine learning model respectively. Removing the models with a coefficient of determination lower than the average level of the corresponding machine learning model to obtain the selected models. Based on the root mean square error, determining the weights of the corresponding selected models. Based on the weights of the selected models, constructing a weighted average ensemble model to obtain the saline-alkali land carbon budget assessment and prediction model.
3. The method for determining the saline-alkali land carbon budget assessment and prediction model according to claim 2, wherein The calculation formula for the root mean square error is: Among them, RMSE is the root mean square error; y i is the measured value, that is, the corresponding preliminary result; is the predicted value, that is, the output of the corresponding machine learning model; n is the number of samples.
4. The method for determining the saline-alkali land carbon budget assessment and prediction model according to claim 2, characterized in that The calculation formula for the coefficient of determination is: Among them, R 2 is the coefficient of determination; y i is the measured value, that is, the corresponding preliminary result; is the predicted value, that is, the output of the corresponding machine learning model; is the average value of the measured values; n is the number of samples.
5. The method for determining the saline-alkali land carbon budget assessment and prediction model according to claim 1, wherein The method for determining the saline-alkali land carbon budget assessment and prediction model further includes: uniformly interpolating the first relevant parameters and the second relevant parameters to the same spatial resolution.
6. The method for determining the saline-alkali land carbon budget assessment and prediction model according to claim 1, wherein The machine learning models include support vector machine, random forest, XGBoost, convolutional neural network, and long short-term memory network.
7. A method for applying a carbon budget assessment and prediction model for saline-alkali land, characterized in that, The application method of the saline-alkali land carbon budget assessment and prediction model includes: Obtain the future first relevant parameters and future second relevant parameters of the target area; the future first relevant parameters include: future climate data, future atmospheric CO2 concentration data, future vegetation type, and future soil particle composition data; the future climate data and the future atmospheric CO2 concentration data are data obtained based on an Earth system model; the future vegetation type is historical vegetation type data; the future soil particle composition data is historical soil particle composition data; the second relevant parameters include: future net primary productivity interpreted based on satellite remote sensing data and the mean value of relevant data within the future 1m soil layer; the future net primary productivity interpreted based on satellite remote sensing data is data predicted using a time series autoregressive prediction method based on historical remotely sensed observed net primary productivity data; the mean value of relevant data within the future 1m soil layer is data calculated by weighted averaging based on the surface and subsurface soil pH, conductivity, and alkalinity data at a 1km resolution of the historical FAO World Soil Database. Input the future first relevant parameters and the future second relevant parameters into the saline-alkali land carbon budget assessment and prediction model to obtain the future net ecosystem productivity, future soil organic carbon density, and future vegetation carbon density of the target area; the saline-alkali land carbon budget assessment and prediction model is a model trained based on the determination method of the saline-alkali land carbon budget assessment and prediction model according to any one of claims 1-6.
8. A computer device, comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the computer program to implement the determination method of the saline-alkali land carbon budget assessment and prediction model according to any one of claims 1-6 or the application method of the saline-alkali land carbon budget assessment and prediction model according to claim 7.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the determination method of the saline-alkali land carbon budget assessment and prediction model according to any one of claims 1-6 or the application method of the saline-alkali land carbon budget assessment and prediction model according to claim 7.
10. A computer program product comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the determination method of the saline-alkali land carbon budget assessment and prediction model according to any one of claims 1-6 or the application method of the saline-alkali land carbon budget assessment and prediction model according to claim 7.